API

Top level user functions:

all(a[, axis, out, keepdims]) Test whether all array elements along a given axis evaluate to True.
allclose(arr1, arr2[, rtol, atol, equal_nan]) Returns True if two arrays are element-wise equal within a tolerance.
angle(x[, deg]) Return the angle of the complex argument.
any(a[, axis, out, keepdims]) Test whether any array element along a given axis evaluates to True.
apply_along_axis(func1d, axis, arr, *args, …) Apply a function to 1-D slices along the given axis.
apply_over_axes(func, a, axes) Apply a function repeatedly over multiple axes.
arange(*args, **kwargs) Return evenly spaced values from start to stop with step size step.
arccos(x, /[, out, where, casting, order, …]) Trigonometric inverse cosine, element-wise.
arccosh(x, /[, out, where, casting, order, …]) Inverse hyperbolic cosine, element-wise.
arcsin(x, /[, out, where, casting, order, …]) Inverse sine, element-wise.
arcsinh(x, /[, out, where, casting, order, …]) Inverse hyperbolic sine element-wise.
arctan(x, /[, out, where, casting, order, …]) Trigonometric inverse tangent, element-wise.
arctan2(x1, x2, /[, out, where, casting, …]) Element-wise arc tangent of x1/x2 choosing the quadrant correctly.
arctanh(x, /[, out, where, casting, order, …]) Inverse hyperbolic tangent element-wise.
argmax(a[, axis, out]) Returns the indices of the maximum values along an axis.
argmin(a[, axis, out]) Returns the indices of the minimum values along an axis.
argtopk(a, k[, axis, split_every]) Extract the indices of the k largest elements from a on the given axis, and return them sorted from largest to smallest.
argwhere(a) Find the indices of array elements that are non-zero, grouped by element.
around(x[, decimals]) Evenly round to the given number of decimals.
array(object[, dtype, copy, order, subok, ndmin]) This docstring was copied from numpy.array.
asanyarray(a) Convert the input to a dask array.
asarray(a, **kwargs) Convert the input to a dask array.
atleast_1d(*arys) Convert inputs to arrays with at least one dimension.
atleast_2d(*arys) View inputs as arrays with at least two dimensions.
atleast_3d(*arys) View inputs as arrays with at least three dimensions.
average(a[, axis, weights, returned]) Compute the weighted average along the specified axis.
bincount(x[, weights, minlength]) This docstring was copied from numpy.bincount.
bitwise_and(x1, x2, /[, out, where, …]) Compute the bit-wise AND of two arrays element-wise.
bitwise_not(x, /[, out, where, casting, …]) Compute bit-wise inversion, or bit-wise NOT, element-wise.
bitwise_or(x1, x2, /[, out, where, casting, …]) Compute the bit-wise OR of two arrays element-wise.
bitwise_xor(x1, x2, /[, out, where, …]) Compute the bit-wise XOR of two arrays element-wise.
block(arrays[, allow_unknown_chunksizes]) Assemble an nd-array from nested lists of blocks.
broadcast_arrays(*args, **kwargs) Broadcast any number of arrays against each other.
broadcast_to(x, shape[, chunks]) Broadcast an array to a new shape.
coarsen(reduction, x, axes[, trim_excess]) Coarsen array by applying reduction to fixed size neighborhoods
ceil(x, /[, out, where, casting, order, …]) Return the ceiling of the input, element-wise.
choose(a, choices) Construct an array from an index array and a set of arrays to choose from.
clip(*args, **kwargs) Clip (limit) the values in an array.
compress(condition, a[, axis]) Return selected slices of an array along given axis.
concatenate(seq[, axis, …]) Concatenate arrays along an existing axis
conj(x, /[, out, where, casting, order, …]) Return the complex conjugate, element-wise.
copysign(x1, x2, /[, out, where, casting, …]) Change the sign of x1 to that of x2, element-wise.
corrcoef(x[, y, rowvar]) Return Pearson product-moment correlation coefficients.
cos(x, /[, out, where, casting, order, …]) Cosine element-wise.
cosh(x, /[, out, where, casting, order, …]) Hyperbolic cosine, element-wise.
count_nonzero(a[, axis]) Counts the number of non-zero values in the array a.
cov(m[, y, rowvar, bias, ddof]) Estimate a covariance matrix, given data and weights.
cumprod(a[, axis, dtype, out]) Return the cumulative product of elements along a given axis.
cumsum(a[, axis, dtype, out]) Return the cumulative sum of the elements along a given axis.
deg2rad(x, /[, out, where, casting, order, …]) Convert angles from degrees to radians.
degrees(x, /[, out, where, casting, order, …]) Convert angles from radians to degrees.
diag(v) Extract a diagonal or construct a diagonal array.
diagonal(a[, offset, axis1, axis2]) Return specified diagonals.
diff(a[, n, axis]) Calculate the n-th discrete difference along the given axis.
digitize(a, bins[, right]) Return the indices of the bins to which each value in input array belongs.
dot(a, b[, out]) This docstring was copied from numpy.dot.
dstack(tup[, allow_unknown_chunksizes]) Stack arrays in sequence depth wise (along third axis).
ediff1d(ary[, to_end, to_begin]) The differences between consecutive elements of an array.
einsum(subscripts, *operands[, out, dtype, …]) This docstring was copied from numpy.einsum.
empty(*args, **kwargs) Blocked variant of empty
empty_like(a[, dtype, chunks]) Return a new array with the same shape and type as a given array.
exp(x, /[, out, where, casting, order, …]) Calculate the exponential of all elements in the input array.
expm1(x, /[, out, where, casting, order, …]) Calculate exp(x) - 1 for all elements in the array.
eye(N[, chunks, M, k, dtype]) Return a 2-D Array with ones on the diagonal and zeros elsewhere.
fabs(x, /[, out, where, casting, order, …]) Compute the absolute values element-wise.
fix(*args, **kwargs) Round to nearest integer towards zero.
flatnonzero(a) Return indices that are non-zero in the flattened version of a.
flip(m, axis) Reverse element order along axis.
flipud(m) Flip array in the up/down direction.
fliplr(m) Flip array in the left/right direction.
floor(x, /[, out, where, casting, order, …]) Return the floor of the input, element-wise.
fmax(x1, x2, /[, out, where, casting, …]) Element-wise maximum of array elements.
fmin(x1, x2, /[, out, where, casting, …]) Element-wise minimum of array elements.
fmod(x1, x2, /[, out, where, casting, …]) Return the element-wise remainder of division.
frexp(x[, out1, out2], / [[, out, where, …]) Decompose the elements of x into mantissa and twos exponent.
fromfunction(func[, chunks, shape, dtype]) Construct an array by executing a function over each coordinate.
frompyfunc(func, nin, nout) This docstring was copied from numpy.frompyfunc.
full(*args, **kwargs) Blocked variant of full
full_like(a, fill_value[, dtype, chunks]) Return a full array with the same shape and type as a given array.
gradient(f, *varargs, **kwargs) Return the gradient of an N-dimensional array.
histogram(a[, bins, range, normed, weights, …]) Blocked variant of numpy.histogram().
hstack(tup[, allow_unknown_chunksizes]) Stack arrays in sequence horizontally (column wise).
hypot(x1, x2, /[, out, where, casting, …]) Given the “legs” of a right triangle, return its hypotenuse.
imag(*args, **kwargs) Return the imaginary part of the complex argument.
indices(dimensions[, dtype, chunks]) Implements NumPy’s indices for Dask Arrays.
insert(arr, obj, values, axis) Insert values along the given axis before the given indices.
invert(x, /[, out, where, casting, order, …]) Compute bit-wise inversion, or bit-wise NOT, element-wise.
isclose(arr1, arr2[, rtol, atol, equal_nan]) Returns a boolean array where two arrays are element-wise equal within a tolerance.
iscomplex(*args, **kwargs) Returns a bool array, where True if input element is complex.
isfinite(x, /[, out, where, casting, order, …]) Test element-wise for finiteness (not infinity or not Not a Number).
isin(element, test_elements[, …]) Calculates element in test_elements, broadcasting over element only.
isinf(x, /[, out, where, casting, order, …]) Test element-wise for positive or negative infinity.
isneginf Return (x1 == x2) element-wise.
isnan(x, /[, out, where, casting, order, …]) Test element-wise for NaN and return result as a boolean array.
isnull(values) pandas.isnull for dask arrays
isposinf Return (x1 == x2) element-wise.
isreal(*args, **kwargs) Returns a bool array, where True if input element is real.
ldexp(x1, x2, /[, out, where, casting, …]) Returns x1 * 2**x2, element-wise.
linspace(start, stop[, num, endpoint, …]) Return num evenly spaced values over the closed interval [start, stop].
log(x, /[, out, where, casting, order, …]) Natural logarithm, element-wise.
log10(x, /[, out, where, casting, order, …]) Return the base 10 logarithm of the input array, element-wise.
log1p(x, /[, out, where, casting, order, …]) Return the natural logarithm of one plus the input array, element-wise.
log2(x, /[, out, where, casting, order, …]) Base-2 logarithm of x.
logaddexp(x1, x2, /[, out, where, casting, …]) Logarithm of the sum of exponentiations of the inputs.
logaddexp2(x1, x2, /[, out, where, casting, …]) Logarithm of the sum of exponentiations of the inputs in base-2.
logical_and(x1, x2, /[, out, where, …]) Compute the truth value of x1 AND x2 element-wise.
logical_not(x, /[, out, where, casting, …]) Compute the truth value of NOT x element-wise.
logical_or(x1, x2, /[, out, where, casting, …]) Compute the truth value of x1 OR x2 element-wise.
logical_xor(x1, x2, /[, out, where, …]) Compute the truth value of x1 XOR x2, element-wise.
map_blocks(func, *args, **kwargs) Map a function across all blocks of a dask array.
map_overlap(x, func, depth[, boundary, trim]) Map a function over blocks of the array with some overlap
matmul(x1, x2, /[, out, casting, order, …]) This docstring was copied from numpy.matmul.
max(a[, axis, out, keepdims, initial]) Return the maximum of an array or maximum along an axis.
maximum(x1, x2, /[, out, where, casting, …]) Element-wise maximum of array elements.
mean(a[, axis, dtype, out, keepdims]) Compute the arithmetic mean along the specified axis.
meshgrid(*xi, **kwargs) Return coordinate matrices from coordinate vectors.
min(a[, axis, out, keepdims, initial]) Return the minimum of an array or minimum along an axis.
minimum(x1, x2, /[, out, where, casting, …]) Element-wise minimum of array elements.
modf(x[, out1, out2], / [[, out, where, …]) Return the fractional and integral parts of an array, element-wise.
moment(a, order[, axis, dtype, keepdims, …])
nanargmax(x, axis, **kwargs)
nanargmin(x, axis, **kwargs)
nancumprod(a[, axis, dtype, out]) Return the cumulative product of array elements over a given axis treating Not a Numbers (NaNs) as one.
nancumsum(a[, axis, dtype, out]) Return the cumulative sum of array elements over a given axis treating Not a Numbers (NaNs) as zero.
nanmax(a[, axis, out, keepdims]) Return the maximum of an array or maximum along an axis, ignoring any NaNs.
nanmean(a[, axis, dtype, out, keepdims]) Compute the arithmetic mean along the specified axis, ignoring NaNs.
nanmin(a[, axis, out, keepdims]) Return minimum of an array or minimum along an axis, ignoring any NaNs.
nanprod(a[, axis, dtype, out, keepdims]) Return the product of array elements over a given axis treating Not a Numbers (NaNs) as ones.
nanstd(a[, axis, dtype, out, ddof, keepdims]) Compute the standard deviation along the specified axis, while ignoring NaNs.
nansum(a[, axis, dtype, out, keepdims]) Return the sum of array elements over a given axis treating Not a Numbers (NaNs) as zero.
nanvar(a[, axis, dtype, out, ddof, keepdims]) Compute the variance along the specified axis, while ignoring NaNs.
nan_to_num(*args, **kwargs) Replace NaN with zero and infinity with large finite numbers.
nextafter(x1, x2, /[, out, where, casting, …]) Return the next floating-point value after x1 towards x2, element-wise.
nonzero(a) Return the indices of the elements that are non-zero.
notnull(values) pandas.notnull for dask arrays
ones(*args, **kwargs) Blocked variant of ones
ones_like(a[, dtype, chunks]) Return an array of ones with the same shape and type as a given array.
outer(a, b) Compute the outer product of two vectors.
pad(array, pad_width, mode, **kwargs) Pads an array.
percentile(a, q[, interpolation, method]) Approximate percentile of 1-D array
PerformanceWarning A warning given when bad chunking may cause poor performance
piecewise(x, condlist, funclist, *args, **kw) Evaluate a piecewise-defined function.
prod(a[, axis, dtype, out, keepdims, initial]) Return the product of array elements over a given axis.
ptp(a[, axis]) Range of values (maximum - minimum) along an axis.
rad2deg(x, /[, out, where, casting, order, …]) Convert angles from radians to degrees.
radians(x, /[, out, where, casting, order, …]) Convert angles from degrees to radians.
ravel(array) Return a contiguous flattened array.
real(*args, **kwargs) Return the real part of the complex argument.
rechunk(x, chunks[, threshold, block_size_limit]) Convert blocks in dask array x for new chunks.
repeat(a, repeats[, axis]) Repeat elements of an array.
reshape(x, shape) Reshape array to new shape
result_type(*arrays_and_dtypes) This docstring was copied from numpy.result_type.
rint(x, /[, out, where, casting, order, …]) Round elements of the array to the nearest integer.
roll(array, shift[, axis]) Roll array elements along a given axis.
round(a[, decimals]) Round an array to the given number of decimals.
sign(x, /[, out, where, casting, order, …]) Returns an element-wise indication of the sign of a number.
signbit(x, /[, out, where, casting, order, …]) Returns element-wise True where signbit is set (less than zero).
sin(x, /[, out, where, casting, order, …]) Trigonometric sine, element-wise.
sinh(x, /[, out, where, casting, order, …]) Hyperbolic sine, element-wise.
sqrt(x, /[, out, where, casting, order, …]) Return the non-negative square-root of an array, element-wise.
square(x, /[, out, where, casting, order, …]) Return the element-wise square of the input.
squeeze(a[, axis]) Remove single-dimensional entries from the shape of an array.
stack(seq[, axis]) Stack arrays along a new axis
std(a[, axis, dtype, out, ddof, keepdims]) Compute the standard deviation along the specified axis.
sum(a[, axis, dtype, out, keepdims, initial]) Sum of array elements over a given axis.
take(a, indices[, axis]) Take elements from an array along an axis.
tan(x, /[, out, where, casting, order, …]) Compute tangent element-wise.
tanh(x, /[, out, where, casting, order, …]) Compute hyperbolic tangent element-wise.
tensordot(lhs, rhs[, axes]) Compute tensor dot product along specified axes for arrays >= 1-D.
tile(A, reps) Construct an array by repeating A the number of times given by reps.
topk(a, k[, axis, split_every]) Extract the k largest elements from a on the given axis, and return them sorted from largest to smallest.
trace(a[, offset, axis1, axis2, dtype, out]) Return the sum along diagonals of the array.
transpose(a[, axes]) Permute the dimensions of an array.
tril(m[, k]) Lower triangle of an array with elements above the k-th diagonal zeroed.
triu(m[, k]) Upper triangle of an array with elements above the k-th diagonal zeroed.
trunc(x, /[, out, where, casting, order, …]) Return the truncated value of the input, element-wise.
unique(ar[, return_index, return_inverse, …]) Find the unique elements of an array.
unravel_index(indices, shape[, order]) This docstring was copied from numpy.unravel_index.
var(a[, axis, dtype, out, ddof, keepdims]) Compute the variance along the specified axis.
vdot(a, b) This docstring was copied from numpy.vdot.
vstack(tup[, allow_unknown_chunksizes]) Stack arrays in sequence vertically (row wise).
where(condition, [x, y]) This docstring was copied from numpy.where.
zeros(*args, **kwargs) Blocked variant of zeros
zeros_like(a[, dtype, chunks]) Return an array of zeros with the same shape and type as a given array.

Fast Fourier Transforms

fft.fft_wrap(fft_func[, kind, dtype]) Wrap 1D, 2D, and ND real and complex FFT functions
fft.fft(a[, n, axis]) Wrapping of numpy.fft.fft
fft.fft2(a[, s, axes]) Wrapping of numpy.fft.fft2
fft.fftn(a[, s, axes]) Wrapping of numpy.fft.fftn
fft.ifft(a[, n, axis]) Wrapping of numpy.fft.ifft
fft.ifft2(a[, s, axes]) Wrapping of numpy.fft.ifft2
fft.ifftn(a[, s, axes]) Wrapping of numpy.fft.ifftn
fft.rfft(a[, n, axis]) Wrapping of numpy.fft.rfft
fft.rfft2(a[, s, axes]) Wrapping of numpy.fft.rfft2
fft.rfftn(a[, s, axes]) Wrapping of numpy.fft.rfftn
fft.irfft(a[, n, axis]) Wrapping of numpy.fft.irfft
fft.irfft2(a[, s, axes]) Wrapping of numpy.fft.irfft2
fft.irfftn(a[, s, axes]) Wrapping of numpy.fft.irfftn
fft.hfft(a[, n, axis]) Wrapping of numpy.fft.hfft
fft.ihfft(a[, n, axis]) Wrapping of numpy.fft.ihfft
fft.fftfreq(n[, d, chunks]) Return the Discrete Fourier Transform sample frequencies.
fft.rfftfreq(n[, d, chunks]) Return the Discrete Fourier Transform sample frequencies (for usage with rfft, irfft).
fft.fftshift(x[, axes]) Shift the zero-frequency component to the center of the spectrum.
fft.ifftshift(x[, axes]) The inverse of fftshift.

Linear Algebra

linalg.cholesky(a[, lower]) Returns the Cholesky decomposition, \(A = L L^*\) or \(A = U^* U\) of a Hermitian positive-definite matrix A.
linalg.inv(a) Compute the inverse of a matrix with LU decomposition and forward / backward substitutions.
linalg.lstsq(a, b) Return the least-squares solution to a linear matrix equation using QR decomposition.
linalg.lu(a) Compute the lu decomposition of a matrix.
linalg.norm(x[, ord, axis, keepdims]) Matrix or vector norm.
linalg.qr(a) Compute the qr factorization of a matrix.
linalg.solve(a, b[, sym_pos]) Solve the equation a x = b for x.
linalg.solve_triangular(a, b[, lower]) Solve the equation a x = b for x, assuming a is a triangular matrix.
linalg.svd(a) Compute the singular value decomposition of a matrix.
linalg.svd_compressed(a, k[, n_power_iter, seed]) Randomly compressed rank-k thin Singular Value Decomposition.
linalg.sfqr(data[, name]) Direct Short-and-Fat QR
linalg.tsqr(data[, compute_svd, …]) Direct Tall-and-Skinny QR algorithm

Masked Arrays

ma.filled(a[, fill_value]) Return input as an array with masked data replaced by a fill value.
ma.fix_invalid(a[, fill_value]) Return input with invalid data masked and replaced by a fill value.
ma.getdata(a) Return the data of a masked array as an ndarray.
ma.getmaskarray(a) Return the mask of a masked array, or full boolean array of False.
ma.masked_array(data[, mask, fill_value]) An array class with possibly masked values.
ma.masked_equal(a, value) Mask an array where equal to a given value.
ma.masked_greater(x, value[, copy]) Mask an array where greater than a given value.
ma.masked_greater_equal(x, value[, copy]) Mask an array where greater than or equal to a given value.
ma.masked_inside(x, v1, v2) Mask an array inside a given interval.
ma.masked_invalid(a) Mask an array where invalid values occur (NaNs or infs).
ma.masked_less(x, value[, copy]) Mask an array where less than a given value.
ma.masked_less_equal(x, value[, copy]) Mask an array where less than or equal to a given value.
ma.masked_not_equal(x, value[, copy]) Mask an array where not equal to a given value.
ma.masked_outside(x, v1, v2) Mask an array outside a given interval.
ma.masked_values(x, value[, rtol, atol, shrink]) Mask using floating point equality.
ma.masked_where(condition, a) Mask an array where a condition is met.
ma.set_fill_value(a, fill_value) Set the filling value of a, if a is a masked array.

Random

random.beta(a, b[, size]) Draw samples from a Beta distribution.
random.binomial(n, p[, size]) Draw samples from a binomial distribution.
random.chisquare(df[, size]) Draw samples from a chi-square distribution.
random.choice(a[, size, replace, p]) Generates a random sample from a given 1-D array
random.exponential([scale, size]) Draw samples from an exponential distribution.
random.f(dfnum, dfden[, size]) Draw samples from an F distribution.
random.gamma(shape[, scale, size]) Draw samples from a Gamma distribution.
random.geometric(p[, size]) Draw samples from the geometric distribution.
random.gumbel([loc, scale, size]) Draw samples from a Gumbel distribution.
random.hypergeometric(ngood, nbad, nsample) Draw samples from a Hypergeometric distribution.
random.laplace([loc, scale, size]) Draw samples from the Laplace or double exponential distribution with specified location (or mean) and scale (decay).
random.logistic([loc, scale, size]) Draw samples from a logistic distribution.
random.lognormal([mean, sigma, size]) Draw samples from a log-normal distribution.
random.logseries(p[, size]) Draw samples from a logarithmic series distribution.
random.negative_binomial(n, p[, size]) Draw samples from a negative binomial distribution.
random.noncentral_chisquare(df, nonc[, size]) Draw samples from a noncentral chi-square distribution.
random.noncentral_f(dfnum, dfden, nonc[, size]) Draw samples from the noncentral F distribution.
random.normal([loc, scale, size]) Draw random samples from a normal (Gaussian) distribution.
random.pareto(a[, size]) Draw samples from a Pareto II or Lomax distribution with specified shape.
random.poisson([lam, size]) Draw samples from a Poisson distribution.
random.power(a[, size]) Draws samples in [0, 1] from a power distribution with positive exponent a - 1.
random.randint(low[, high, size, dtype]) Return random integers from low (inclusive) to high (exclusive).
random.random([size]) Return random floats in the half-open interval [0.0, 1.0).
random.random_sample([size]) Return random floats in the half-open interval [0.0, 1.0).
random.rayleigh([scale, size]) Draw samples from a Rayleigh distribution.
random.standard_cauchy([size]) Draw samples from a standard Cauchy distribution with mode = 0.
random.standard_exponential([size]) Draw samples from the standard exponential distribution.
random.standard_gamma(shape[, size]) Draw samples from a standard Gamma distribution.
random.standard_normal([size]) Draw samples from a standard Normal distribution (mean=0, stdev=1).
random.standard_t(df[, size]) Draw samples from a standard Student’s t distribution with df degrees of freedom.
random.triangular(left, mode, right[, size]) Draw samples from the triangular distribution over the interval [left, right].
random.uniform([low, high, size]) Draw samples from a uniform distribution.
random.vonmises(mu, kappa[, size]) Draw samples from a von Mises distribution.
random.wald(mean, scale[, size]) Draw samples from a Wald, or inverse Gaussian, distribution.
random.weibull(a[, size]) Draw samples from a Weibull distribution.
random.zipf(a[, size]) Standard distributions

Stats

stats.ttest_ind(a, b[, axis, equal_var]) Calculate the T-test for the means of two independent samples of scores.
stats.ttest_1samp(a, popmean[, axis, nan_policy]) Calculate the T-test for the mean of ONE group of scores.
stats.ttest_rel(a, b[, axis, nan_policy]) Calculate the T-test on TWO RELATED samples of scores, a and b.
stats.chisquare(f_obs[, f_exp, ddof, axis]) Calculate a one-way chi square test.
stats.power_divergence(f_obs[, f_exp, ddof, …]) Cressie-Read power divergence statistic and goodness of fit test.
stats.skew(a[, axis, bias, nan_policy]) Compute the skewness of a data set.
stats.skewtest(a[, axis, nan_policy]) Test whether the skew is different from the normal distribution.
stats.kurtosis(a[, axis, fisher, bias, …]) Compute the kurtosis (Fisher or Pearson) of a dataset.
stats.kurtosistest(a[, axis, nan_policy]) Test whether a dataset has normal kurtosis.
stats.normaltest(a[, axis, nan_policy]) Test whether a sample differs from a normal distribution.
stats.f_oneway(*args) Performs a 1-way ANOVA.
stats.moment(a[, moment, axis, nan_policy]) Calculate the nth moment about the mean for a sample.

Image Support

image.imread(filename[, imread, preprocess]) Read a stack of images into a dask array

Slightly Overlapping Computations

overlap.overlap(x, depth, boundary) Share boundaries between neighboring blocks
overlap.map_overlap(x, func, depth[, …]) Map a function over blocks of the array with some overlap
overlap.trim_internal(x, axes[, boundary]) Trim sides from each block
overlap.trim_overlap(x, depth[, boundary]) Trim sides from each block.

Create and Store Arrays

from_array(x[, chunks, name, lock, asarray, …]) Create dask array from something that looks like an array
from_delayed(value, shape, dtype[, name]) Create a dask array from a dask delayed value
from_npy_stack(dirname[, mmap_mode]) Load dask array from stack of npy files
from_zarr(url[, component, storage_options, …]) Load array from the zarr storage format
from_tiledb(uri[, attribute, chunks, …]) Load array from the TileDB storage format
store(sources, targets[, lock, regions, …]) Store dask arrays in array-like objects, overwrite data in target
to_hdf5(filename, *args, **kwargs) Store arrays in HDF5 file
to_zarr(arr, url[, component, …]) Save array to the zarr storage format
to_npy_stack(dirname, x[, axis]) Write dask array to a stack of .npy files
to_tiledb(darray, uri[, compute, …]) Save array to the TileDB storage format

Generalized Ufuncs

apply_gufunc(func, signature, *args, **kwargs) Apply a generalized ufunc or similar python function to arrays.
as_gufunc([signature]) Decorator for dask.array.gufunc.
gufunc(pyfunc, **kwargs) Binds pyfunc into dask.array.apply_gufunc when called.

Internal functions

blockwise(func, out_ind, *args, **kwargs) Tensor operation: Generalized inner and outer products
normalize_chunks(chunks[, shape, limit, …]) Normalize chunks to tuple of tuples

Other functions

dask.array.from_array(x, chunks='auto', name=None, lock=False, asarray=True, fancy=True, getitem=None)

Create dask array from something that looks like an array

Input must have a .shape and support numpy-style slicing.

Parameters:
x : array_like
chunks : int, tuple

How to chunk the array. Must be one of the following forms: - A blocksize like 1000. - A blockshape like (1000, 1000). - Explicit sizes of all blocks along all dimensions like

((1000, 1000, 500), (400, 400)).

  • A size in bytes, like “100 MiB” which will choose a uniform block-like shape
  • The word “auto” which acts like the above, but uses a configuration value array.chunk-size for the chunk size

-1 or None as a blocksize indicate the size of the corresponding dimension.

name : str, optional

The key name to use for the array. Defaults to a hash of x. By default, hash uses python’s standard sha1. This behaviour can be changed by installing cityhash, xxhash or murmurhash. If installed, a large-factor speedup can be obtained in the tokenisation step. Use name=False to generate a random name instead of hashing (fast)

lock : bool or Lock, optional

If x doesn’t support concurrent reads then provide a lock here, or pass in True to have dask.array create one for you.

asarray : bool, optional

If True (default), then chunks will be converted to instances of ndarray. Set to False to pass passed chunks through unchanged.

fancy : bool, optional

If x doesn’t support fancy indexing (e.g. indexing with lists or arrays) then set to False. Default is True.

Examples

>>> x = h5py.File('...')['/data/path']  # doctest: +SKIP
>>> a = da.from_array(x, chunks=(1000, 1000))  # doctest: +SKIP

If your underlying datastore does not support concurrent reads then include the lock=True keyword argument or lock=mylock if you want multiple arrays to coordinate around the same lock.

>>> a = da.from_array(x, chunks=(1000, 1000), lock=True)  # doctest: +SKIP

If your underlying datastore has a .chunks attribute (as h5py and zarr datasets do) then a multiple of that chunk shape will be used if you do not provide a chunk shape.

>>> a = da.from_array(x, chunks='auto')  # doctest: +SKIP
>>> a = da.from_array(x, chunks='100 MiB')  # doctest: +SKIP
>>> a = da.from_array(x)  # doctest: +SKIP
dask.array.from_delayed(value, shape, dtype, name=None)

Create a dask array from a dask delayed value

This routine is useful for constructing dask arrays in an ad-hoc fashion using dask delayed, particularly when combined with stack and concatenate.

The dask array will consist of a single chunk.

Examples

>>> from dask import delayed
>>> value = delayed(np.ones)(5)
>>> array = from_delayed(value, (5,), float)
>>> array
dask.array<from-value, shape=(5,), dtype=float64, chunksize=(5,)>
>>> array.compute()
array([1., 1., 1., 1., 1.])
dask.array.store(sources, targets, lock=True, regions=None, compute=True, return_stored=False, **kwargs)

Store dask arrays in array-like objects, overwrite data in target

This stores dask arrays into object that supports numpy-style setitem indexing. It stores values chunk by chunk so that it does not have to fill up memory. For best performance you can align the block size of the storage target with the block size of your array.

If your data fits in memory then you may prefer calling np.array(myarray) instead.

Parameters:
sources: Array or iterable of Arrays
targets: array-like or Delayed or iterable of array-likes and/or Delayeds

These should support setitem syntax target[10:20] = ...

lock: boolean or threading.Lock, optional

Whether or not to lock the data stores while storing. Pass True (lock each file individually), False (don’t lock) or a particular threading.Lock object to be shared among all writes.

regions: tuple of slices or list of tuples of slices

Each region tuple in regions should be such that target[region].shape = source.shape for the corresponding source and target in sources and targets, respectively. If this is a tuple, the contents will be assumed to be slices, so do not provide a tuple of tuples.

compute: boolean, optional

If true compute immediately, return dask.delayed.Delayed otherwise

return_stored: boolean, optional

Optionally return the stored result (default False).

Examples

>>> x = ...  # doctest: +SKIP
>>> import h5py  # doctest: +SKIP
>>> f = h5py.File('myfile.hdf5')  # doctest: +SKIP
>>> dset = f.create_dataset('/data', shape=x.shape,
...                                  chunks=x.chunks,
...                                  dtype='f8')  # doctest: +SKIP
>>> store(x, dset)  # doctest: +SKIP

Alternatively store many arrays at the same time

>>> store([x, y, z], [dset1, dset2, dset3])  # doctest: +SKIP
dask.array.coarsen(reduction, x, axes, trim_excess=False)

Coarsen array by applying reduction to fixed size neighborhoods

Parameters:
reduction: function

Function like np.sum, np.mean, etc…

x: np.ndarray

Array to be coarsened

axes: dict

Mapping of axis to coarsening factor

Examples

>>> x = np.array([1, 2, 3, 4, 5, 6])
>>> coarsen(np.sum, x, {0: 2})
array([ 3,  7, 11])
>>> coarsen(np.max, x, {0: 3})
array([3, 6])

Provide dictionary of scale per dimension

>>> x = np.arange(24).reshape((4, 6))
>>> x
array([[ 0,  1,  2,  3,  4,  5],
       [ 6,  7,  8,  9, 10, 11],
       [12, 13, 14, 15, 16, 17],
       [18, 19, 20, 21, 22, 23]])
>>> coarsen(np.min, x, {0: 2, 1: 3})
array([[ 0,  3],
       [12, 15]])

You must avoid excess elements explicitly

>>> x = np.array([1, 2, 3, 4, 5, 6, 7, 8])
>>> coarsen(np.min, x, {0: 3}, trim_excess=True)
array([1, 4])
dask.array.stack(seq, axis=0)

Stack arrays along a new axis

Given a sequence of dask arrays, form a new dask array by stacking them along a new dimension (axis=0 by default)

See also

concatenate

Examples

Create slices

>>> import dask.array as da
>>> import numpy as np
>>> data = [from_array(np.ones((4, 4)), chunks=(2, 2))
...          for i in range(3)]
>>> x = da.stack(data, axis=0)
>>> x.shape
(3, 4, 4)
>>> da.stack(data, axis=1).shape
(4, 3, 4)
>>> da.stack(data, axis=-1).shape
(4, 4, 3)

Result is a new dask Array

dask.array.concatenate(seq, axis=0, allow_unknown_chunksizes=False)

Concatenate arrays along an existing axis

Given a sequence of dask Arrays form a new dask Array by stacking them along an existing dimension (axis=0 by default)

Parameters:
seq: list of dask.arrays
axis: int

Dimension along which to align all of the arrays

allow_unknown_chunksizes: bool

Allow unknown chunksizes, such as come from converting from dask dataframes. Dask.array is unable to verify that chunks line up. If data comes from differently aligned sources then this can cause unexpected results.

See also

stack

Examples

Create slices

>>> import dask.array as da
>>> import numpy as np
>>> data = [from_array(np.ones((4, 4)), chunks=(2, 2))
...          for i in range(3)]
>>> x = da.concatenate(data, axis=0)
>>> x.shape
(12, 4)
>>> da.concatenate(data, axis=1).shape
(4, 12)

Result is a new dask Array

dask.array.all(a, axis=None, out=None, keepdims=<no value>)

Test whether all array elements along a given axis evaluate to True.

Parameters:
a : array_like

Input array or object that can be converted to an array.

axis : None or int or tuple of ints, optional

Axis or axes along which a logical AND reduction is performed. The default (axis = None) is to perform a logical AND over all the dimensions of the input array. axis may be negative, in which case it counts from the last to the first axis.

New in version 1.7.0.

If this is a tuple of ints, a reduction is performed on multiple axes, instead of a single axis or all the axes as before.

out : ndarray, optional

Alternate output array in which to place the result. It must have the same shape as the expected output and its type is preserved (e.g., if dtype(out) is float, the result will consist of 0.0’s and 1.0’s). See doc.ufuncs (Section “Output arguments”) for more details.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the input array.

If the default value is passed, then keepdims will not be passed through to the all method of sub-classes of ndarray, however any non-default value will be. If the sub-class’ method does not implement keepdims any exceptions will be raised.

Returns:
all : ndarray, bool

A new boolean or array is returned unless out is specified, in which case a reference to out is returned.

See also

ndarray.all
equivalent method
any
Test whether any element along a given axis evaluates to True.

Notes

Not a Number (NaN), positive infinity and negative infinity evaluate to True because these are not equal to zero.

Examples

>>> np.all([[True,False],[True,True]])
False
>>> np.all([[True,False],[True,True]], axis=0)
array([ True, False])
>>> np.all([-1, 4, 5])
True
>>> np.all([1.0, np.nan])
True
>>> o=np.array([False])
>>> z=np.all([-1, 4, 5], out=o)
>>> id(z), id(o), z                             # doctest: +SKIP
(28293632, 28293632, array([ True]))
dask.array.allclose(arr1, arr2, rtol=1e-05, atol=1e-08, equal_nan=False)

Returns True if two arrays are element-wise equal within a tolerance.

This docstring was copied from numpy.allclose.

Some inconsistencies with the Dask version may exist.

The tolerance values are positive, typically very small numbers. The relative difference (rtol * abs(b)) and the absolute difference atol are added together to compare against the absolute difference between a and b.

If either array contains one or more NaNs, False is returned. Infs are treated as equal if they are in the same place and of the same sign in both arrays.

Parameters:
a, b : array_like

Input arrays to compare.

rtol : float

The relative tolerance parameter (see Notes).

atol : float

The absolute tolerance parameter (see Notes).

equal_nan : bool

Whether to compare NaN’s as equal. If True, NaN’s in a will be considered equal to NaN’s in b in the output array.

New in version 1.10.0.

Returns:
allclose : bool

Returns True if the two arrays are equal within the given tolerance; False otherwise.

See also

isclose, all, any, equal

Notes

If the following equation is element-wise True, then allclose returns True.

absolute(a - b) <= (atol + rtol * absolute(b))

The above equation is not symmetric in a and b, so that allclose(a, b) might be different from allclose(b, a) in some rare cases.

The comparison of a and b uses standard broadcasting, which means that a and b need not have the same shape in order for allclose(a, b) to evaluate to True. The same is true for equal but not array_equal.

Examples

>>> np.allclose([1e10,1e-7], [1.00001e10,1e-8])  # doctest: +SKIP
False
>>> np.allclose([1e10,1e-8], [1.00001e10,1e-9])  # doctest: +SKIP
True
>>> np.allclose([1e10,1e-8], [1.0001e10,1e-9])  # doctest: +SKIP
False
>>> np.allclose([1.0, np.nan], [1.0, np.nan])  # doctest: +SKIP
False
>>> np.allclose([1.0, np.nan], [1.0, np.nan], equal_nan=True)  # doctest: +SKIP
True
dask.array.angle(x, deg=0)

Return the angle of the complex argument.

Parameters:
z : array_like

A complex number or sequence of complex numbers.

deg : bool, optional

Return angle in degrees if True, radians if False (default).

Returns:
angle : ndarray or scalar

The counterclockwise angle from the positive real axis on the complex plane, with dtype as numpy.float64.

..versionchanged:: 1.16.0

This function works on subclasses of ndarray like ma.array.

See also

arctan2, absolute

Examples

>>> np.angle([1.0, 1.0j, 1+1j])               # in radians  # doctest: +SKIP
array([ 0.        ,  1.57079633,  0.78539816])
>>> np.angle(1+1j, deg=True)                  # in degrees  # doctest: +SKIP
45.0
dask.array.any(a, axis=None, out=None, keepdims=<no value>)

Test whether any array element along a given axis evaluates to True.

Returns single boolean unless axis is not None

Parameters:
a : array_like

Input array or object that can be converted to an array.

axis : None or int or tuple of ints, optional

Axis or axes along which a logical OR reduction is performed. The default (axis = None) is to perform a logical OR over all the dimensions of the input array. axis may be negative, in which case it counts from the last to the first axis.

New in version 1.7.0.

If this is a tuple of ints, a reduction is performed on multiple axes, instead of a single axis or all the axes as before.

out : ndarray, optional

Alternate output array in which to place the result. It must have the same shape as the expected output and its type is preserved (e.g., if it is of type float, then it will remain so, returning 1.0 for True and 0.0 for False, regardless of the type of a). See doc.ufuncs (Section “Output arguments”) for details.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the input array.

If the default value is passed, then keepdims will not be passed through to the any method of sub-classes of ndarray, however any non-default value will be. If the sub-class’ method does not implement keepdims any exceptions will be raised.

Returns:
any : bool or ndarray

A new boolean or ndarray is returned unless out is specified, in which case a reference to out is returned.

See also

ndarray.any
equivalent method
all
Test whether all elements along a given axis evaluate to True.

Notes

Not a Number (NaN), positive infinity and negative infinity evaluate to True because these are not equal to zero.

Examples

>>> np.any([[True, False], [True, True]])
True
>>> np.any([[True, False], [False, False]], axis=0)
array([ True, False])
>>> np.any([-1, 0, 5])
True
>>> np.any(np.nan)
True
>>> o=np.array([False])
>>> z=np.any([-1, 4, 5], out=o)
>>> z, o
(array([ True]), array([ True]))
>>> # Check now that z is a reference to o
>>> z is o
True
>>> id(z), id(o) # identity of z and o              # doctest: +SKIP
(191614240, 191614240)
dask.array.apply_along_axis(func1d, axis, arr, *args, **kwargs)

Apply a function to 1-D slices along the given axis.

This docstring was copied from numpy.apply_along_axis.

Some inconsistencies with the Dask version may exist.

Execute func1d(a, *args) where func1d operates on 1-D arrays and a is a 1-D slice of arr along axis.

This is equivalent to (but faster than) the following use of ndindex and s_, which sets each of ii, jj, and kk to a tuple of indices:

Ni, Nk = a.shape[:axis], a.shape[axis+1:]
for ii in ndindex(Ni):
    for kk in ndindex(Nk):
        f = func1d(arr[ii + s_[:,] + kk])
        Nj = f.shape
        for jj in ndindex(Nj):
            out[ii + jj + kk] = f[jj]

Equivalently, eliminating the inner loop, this can be expressed as:

Ni, Nk = a.shape[:axis], a.shape[axis+1:]
for ii in ndindex(Ni):
    for kk in ndindex(Nk):
        out[ii + s_[...,] + kk] = func1d(arr[ii + s_[:,] + kk])
Parameters:
func1d : function (M,) -> (Nj…)

This function should accept 1-D arrays. It is applied to 1-D slices of arr along the specified axis.

axis : integer

Axis along which arr is sliced.

arr : ndarray (Ni…, M, Nk…)

Input array.

args : any

Additional arguments to func1d.

kwargs : any

Additional named arguments to func1d.

New in version 1.9.0.

Returns:
out : ndarray (Ni…, Nj…, Nk…)

The output array. The shape of out is identical to the shape of arr, except along the axis dimension. This axis is removed, and replaced with new dimensions equal to the shape of the return value of func1d. So if func1d returns a scalar out will have one fewer dimensions than arr.

See also

apply_over_axes
Apply a function repeatedly over multiple axes.

Examples

>>> def my_func(a):  # doctest: +SKIP
...     """Average first and last element of a 1-D array"""
...     return (a[0] + a[-1]) * 0.5
>>> b = np.array([[1,2,3], [4,5,6], [7,8,9]])  # doctest: +SKIP
>>> np.apply_along_axis(my_func, 0, b)  # doctest: +SKIP
array([ 4.,  5.,  6.])
>>> np.apply_along_axis(my_func, 1, b)  # doctest: +SKIP
array([ 2.,  5.,  8.])

For a function that returns a 1D array, the number of dimensions in outarr is the same as arr.

>>> b = np.array([[8,1,7], [4,3,9], [5,2,6]])  # doctest: +SKIP
>>> np.apply_along_axis(sorted, 1, b)  # doctest: +SKIP
array([[1, 7, 8],
       [3, 4, 9],
       [2, 5, 6]])

For a function that returns a higher dimensional array, those dimensions are inserted in place of the axis dimension.

>>> b = np.array([[1,2,3], [4,5,6], [7,8,9]])  # doctest: +SKIP
>>> np.apply_along_axis(np.diag, -1, b)  # doctest: +SKIP
array([[[1, 0, 0],
        [0, 2, 0],
        [0, 0, 3]],
       [[4, 0, 0],
        [0, 5, 0],
        [0, 0, 6]],
       [[7, 0, 0],
        [0, 8, 0],
        [0, 0, 9]]])
dask.array.apply_over_axes(func, a, axes)

Apply a function repeatedly over multiple axes.

This docstring was copied from numpy.apply_over_axes.

Some inconsistencies with the Dask version may exist.

func is called as res = func(a, axis), where axis is the first element of axes. The result res of the function call must have either the same dimensions as a or one less dimension. If res has one less dimension than a, a dimension is inserted before axis. The call to func is then repeated for each axis in axes, with res as the first argument.

Parameters:
func : function

This function must take two arguments, func(a, axis).

a : array_like

Input array.

axes : array_like

Axes over which func is applied; the elements must be integers.

Returns:
apply_over_axis : ndarray

The output array. The number of dimensions is the same as a, but the shape can be different. This depends on whether func changes the shape of its output with respect to its input.

See also

apply_along_axis
Apply a function to 1-D slices of an array along the given axis.

Notes

This function is equivalent to tuple axis arguments to reorderable ufuncs with keepdims=True. Tuple axis arguments to ufuncs have been available since version 1.7.0.

Examples

>>> a = np.arange(24).reshape(2,3,4)  # doctest: +SKIP
>>> a  # doctest: +SKIP
array([[[ 0,  1,  2,  3],
        [ 4,  5,  6,  7],
        [ 8,  9, 10, 11]],
       [[12, 13, 14, 15],
        [16, 17, 18, 19],
        [20, 21, 22, 23]]])

Sum over axes 0 and 2. The result has same number of dimensions as the original array:

>>> np.apply_over_axes(np.sum, a, [0,2])  # doctest: +SKIP
array([[[ 60],
        [ 92],
        [124]]])

Tuple axis arguments to ufuncs are equivalent:

>>> np.sum(a, axis=(0,2), keepdims=True)  # doctest: +SKIP
array([[[ 60],
        [ 92],
        [124]]])
dask.array.arange(*args, **kwargs)

Return evenly spaced values from start to stop with step size step.

The values are half-open [start, stop), so including start and excluding stop. This is basically the same as python’s range function but for dask arrays.

When using a non-integer step, such as 0.1, the results will often not be consistent. It is better to use linspace for these cases.

Parameters:
start : int, optional

The starting value of the sequence. The default is 0.

stop : int

The end of the interval, this value is excluded from the interval.

step : int, optional

The spacing between the values. The default is 1 when not specified. The last value of the sequence.

chunks : int

The number of samples on each block. Note that the last block will have fewer samples if len(array) % chunks != 0.

dtype : numpy.dtype

Output dtype. Omit to infer it from start, stop, step

Returns:
samples : dask array
dask.array.arccos(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Trigonometric inverse cosine, element-wise.

The inverse of cos so that, if y = cos(x), then x = arccos(y).

Parameters:
x : array_like

x-coordinate on the unit circle. For real arguments, the domain is [-1, 1].

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
angle : ndarray

The angle of the ray intersecting the unit circle at the given x-coordinate in radians [0, pi]. This is a scalar if x is a scalar.

See also

cos, arctan, arcsin, emath.arccos

Notes

arccos is a multivalued function: for each x there are infinitely many numbers z such that cos(z) = x. The convention is to return the angle z whose real part lies in [0, pi].

For real-valued input data types, arccos always returns real output. For each value that cannot be expressed as a real number or infinity, it yields nan and sets the invalid floating point error flag.

For complex-valued input, arccos is a complex analytic function that has branch cuts [-inf, -1] and [1, inf] and is continuous from above on the former and from below on the latter.

The inverse cos is also known as acos or cos^-1.

References

M. Abramowitz and I.A. Stegun, “Handbook of Mathematical Functions”, 10th printing, 1964, pp. 79. http://www.math.sfu.ca/~cbm/aands/

Examples

We expect the arccos of 1 to be 0, and of -1 to be pi:

>>> np.arccos([1, -1])  # doctest: +SKIP
array([ 0.        ,  3.14159265])

Plot arccos:

>>> import matplotlib.pyplot as plt  # doctest: +SKIP
>>> x = np.linspace(-1, 1, num=100)  # doctest: +SKIP
>>> plt.plot(x, np.arccos(x))  # doctest: +SKIP
>>> plt.axis('tight')  # doctest: +SKIP
>>> plt.show()  # doctest: +SKIP
dask.array.arccosh(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Inverse hyperbolic cosine, element-wise.

Parameters:
x : array_like

Input array.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
arccosh : ndarray

Array of the same shape as x. This is a scalar if x is a scalar.

See also

cosh, arcsinh, sinh, arctanh, tanh

Notes

arccosh is a multivalued function: for each x there are infinitely many numbers z such that cosh(z) = x. The convention is to return the z whose imaginary part lies in [-pi, pi] and the real part in [0, inf].

For real-valued input data types, arccosh always returns real output. For each value that cannot be expressed as a real number or infinity, it yields nan and sets the invalid floating point error flag.

For complex-valued input, arccosh is a complex analytical function that has a branch cut [-inf, 1] and is continuous from above on it.

References

[1]M. Abramowitz and I.A. Stegun, “Handbook of Mathematical Functions”, 10th printing, 1964, pp. 86. http://www.math.sfu.ca/~cbm/aands/
[2]Wikipedia, “Inverse hyperbolic function”, https://en.wikipedia.org/wiki/Arccosh

Examples

>>> np.arccosh([np.e, 10.0])  # doctest: +SKIP
array([ 1.65745445,  2.99322285])
>>> np.arccosh(1)  # doctest: +SKIP
0.0
dask.array.arcsin(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Inverse sine, element-wise.

Parameters:
x : array_like

y-coordinate on the unit circle.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
angle : ndarray

The inverse sine of each element in x, in radians and in the closed interval [-pi/2, pi/2]. This is a scalar if x is a scalar.

See also

sin, cos, arccos, tan, arctan, arctan2, emath.arcsin

Notes

arcsin is a multivalued function: for each x there are infinitely many numbers z such that \(sin(z) = x\). The convention is to return the angle z whose real part lies in [-pi/2, pi/2].

For real-valued input data types, arcsin always returns real output. For each value that cannot be expressed as a real number or infinity, it yields nan and sets the invalid floating point error flag.

For complex-valued input, arcsin is a complex analytic function that has, by convention, the branch cuts [-inf, -1] and [1, inf] and is continuous from above on the former and from below on the latter.

The inverse sine is also known as asin or sin^{-1}.

References

Abramowitz, M. and Stegun, I. A., Handbook of Mathematical Functions, 10th printing, New York: Dover, 1964, pp. 79ff. http://www.math.sfu.ca/~cbm/aands/

Examples

>>> np.arcsin(1)     # pi/2  # doctest: +SKIP
1.5707963267948966
>>> np.arcsin(-1)    # -pi/2  # doctest: +SKIP
-1.5707963267948966
>>> np.arcsin(0)  # doctest: +SKIP
0.0
dask.array.arcsinh(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Inverse hyperbolic sine element-wise.

Parameters:
x : array_like

Input array.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Array of the same shape as x. This is a scalar if x is a scalar.

Notes

arcsinh is a multivalued function: for each x there are infinitely many numbers z such that sinh(z) = x. The convention is to return the z whose imaginary part lies in [-pi/2, pi/2].

For real-valued input data types, arcsinh always returns real output. For each value that cannot be expressed as a real number or infinity, it returns nan and sets the invalid floating point error flag.

For complex-valued input, arccos is a complex analytical function that has branch cuts [1j, infj] and [-1j, -infj] and is continuous from the right on the former and from the left on the latter.

The inverse hyperbolic sine is also known as asinh or sinh^-1.

References

[1]M. Abramowitz and I.A. Stegun, “Handbook of Mathematical Functions”, 10th printing, 1964, pp. 86. http://www.math.sfu.ca/~cbm/aands/
[2]Wikipedia, “Inverse hyperbolic function”, https://en.wikipedia.org/wiki/Arcsinh

Examples

>>> np.arcsinh(np.array([np.e, 10.0]))  # doctest: +SKIP
array([ 1.72538256,  2.99822295])
dask.array.arctan(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Trigonometric inverse tangent, element-wise.

The inverse of tan, so that if y = tan(x) then x = arctan(y).

Parameters:
x : array_like
out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Out has the same shape as x. Its real part is in [-pi/2, pi/2] (arctan(+/-inf) returns +/-pi/2). This is a scalar if x is a scalar.

See also

arctan2
The “four quadrant” arctan of the angle formed by (x, y) and the positive x-axis.
angle
Argument of complex values.

Notes

arctan is a multi-valued function: for each x there are infinitely many numbers z such that tan(z) = x. The convention is to return the angle z whose real part lies in [-pi/2, pi/2].

For real-valued input data types, arctan always returns real output. For each value that cannot be expressed as a real number or infinity, it yields nan and sets the invalid floating point error flag.

For complex-valued input, arctan is a complex analytic function that has [1j, infj] and [-1j, -infj] as branch cuts, and is continuous from the left on the former and from the right on the latter.

The inverse tangent is also known as atan or tan^{-1}.

References

Abramowitz, M. and Stegun, I. A., Handbook of Mathematical Functions, 10th printing, New York: Dover, 1964, pp. 79. http://www.math.sfu.ca/~cbm/aands/

Examples

We expect the arctan of 0 to be 0, and of 1 to be pi/4:

>>> np.arctan([0, 1])  # doctest: +SKIP
array([ 0.        ,  0.78539816])
>>> np.pi/4  # doctest: +SKIP
0.78539816339744828

Plot arctan:

>>> import matplotlib.pyplot as plt  # doctest: +SKIP
>>> x = np.linspace(-10, 10)  # doctest: +SKIP
>>> plt.plot(x, np.arctan(x))  # doctest: +SKIP
>>> plt.axis('tight')  # doctest: +SKIP
>>> plt.show()  # doctest: +SKIP
dask.array.arctan2(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Element-wise arc tangent of x1/x2 choosing the quadrant correctly.

The quadrant (i.e., branch) is chosen so that arctan2(x1, x2) is the signed angle in radians between the ray ending at the origin and passing through the point (1,0), and the ray ending at the origin and passing through the point (x2, x1). (Note the role reversal: the “y-coordinate” is the first function parameter, the “x-coordinate” is the second.) By IEEE convention, this function is defined for x2 = +/-0 and for either or both of x1 and x2 = +/-inf (see Notes for specific values).

This function is not defined for complex-valued arguments; for the so-called argument of complex values, use angle.

Parameters:
x1 : array_like, real-valued

y-coordinates.

x2 : array_like, real-valued

x-coordinates. x2 must be broadcastable to match the shape of x1 or vice versa.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
angle : ndarray

Array of angles in radians, in the range [-pi, pi]. This is a scalar if both x1 and x2 are scalars.

See also

arctan, tan, angle

Notes

arctan2 is identical to the atan2 function of the underlying C library. The following special values are defined in the C standard: [1]

x1 x2 arctan2(x1,x2)
+/- 0 +0 +/- 0
+/- 0 -0 +/- pi
> 0 +/-inf +0 / +pi
< 0 +/-inf -0 / -pi
+/-inf +inf +/- (pi/4)
+/-inf -inf +/- (3*pi/4)

Note that +0 and -0 are distinct floating point numbers, as are +inf and -inf.

References

[1](1, 2) ISO/IEC standard 9899:1999, “Programming language C.”

Examples

Consider four points in different quadrants:

>>> x = np.array([-1, +1, +1, -1])  # doctest: +SKIP
>>> y = np.array([-1, -1, +1, +1])  # doctest: +SKIP
>>> np.arctan2(y, x) * 180 / np.pi  # doctest: +SKIP
array([-135.,  -45.,   45.,  135.])

Note the order of the parameters. arctan2 is defined also when x2 = 0 and at several other special points, obtaining values in the range [-pi, pi]:

>>> np.arctan2([1., -1.], [0., 0.])  # doctest: +SKIP
array([ 1.57079633, -1.57079633])
>>> np.arctan2([0., 0., np.inf], [+0., -0., np.inf])  # doctest: +SKIP
array([ 0.        ,  3.14159265,  0.78539816])
dask.array.arctanh(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Inverse hyperbolic tangent element-wise.

Parameters:
x : array_like

Input array.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Array of the same shape as x. This is a scalar if x is a scalar.

See also

emath.arctanh

Notes

arctanh is a multivalued function: for each x there are infinitely many numbers z such that tanh(z) = x. The convention is to return the z whose imaginary part lies in [-pi/2, pi/2].

For real-valued input data types, arctanh always returns real output. For each value that cannot be expressed as a real number or infinity, it yields nan and sets the invalid floating point error flag.

For complex-valued input, arctanh is a complex analytical function that has branch cuts [-1, -inf] and [1, inf] and is continuous from above on the former and from below on the latter.

The inverse hyperbolic tangent is also known as atanh or tanh^-1.

References

[1]M. Abramowitz and I.A. Stegun, “Handbook of Mathematical Functions”, 10th printing, 1964, pp. 86. http://www.math.sfu.ca/~cbm/aands/
[2]Wikipedia, “Inverse hyperbolic function”, https://en.wikipedia.org/wiki/Arctanh

Examples

>>> np.arctanh([0, -0.5])  # doctest: +SKIP
array([ 0.        , -0.54930614])
dask.array.argmax(a, axis=None, out=None)

Returns the indices of the maximum values along an axis.

Parameters:
a : array_like

Input array.

axis : int, optional

By default, the index is into the flattened array, otherwise along the specified axis.

out : array, optional

If provided, the result will be inserted into this array. It should be of the appropriate shape and dtype.

Returns:
index_array : ndarray of ints

Array of indices into the array. It has the same shape as a.shape with the dimension along axis removed.

See also

ndarray.argmax, argmin

amax
The maximum value along a given axis.
unravel_index
Convert a flat index into an index tuple.

Notes

In case of multiple occurrences of the maximum values, the indices corresponding to the first occurrence are returned.

Examples

>>> a = np.arange(6).reshape(2,3) + 10
>>> a
array([[10, 11, 12],
       [13, 14, 15]])
>>> np.argmax(a)
5
>>> np.argmax(a, axis=0)
array([1, 1, 1])
>>> np.argmax(a, axis=1)
array([2, 2])

Indexes of the maximal elements of a N-dimensional array:

>>> ind = np.unravel_index(np.argmax(a, axis=None), a.shape)
>>> ind
(1, 2)
>>> a[ind]
15
>>> b = np.arange(6)
>>> b[1] = 5
>>> b
array([0, 5, 2, 3, 4, 5])
>>> np.argmax(b)  # Only the first occurrence is returned.
1
dask.array.argmin(a, axis=None, out=None)

Returns the indices of the minimum values along an axis.

Parameters:
a : array_like

Input array.

axis : int, optional

By default, the index is into the flattened array, otherwise along the specified axis.

out : array, optional

If provided, the result will be inserted into this array. It should be of the appropriate shape and dtype.

Returns:
index_array : ndarray of ints

Array of indices into the array. It has the same shape as a.shape with the dimension along axis removed.

See also

ndarray.argmin, argmax

amin
The minimum value along a given axis.
unravel_index
Convert a flat index into an index tuple.

Notes

In case of multiple occurrences of the minimum values, the indices corresponding to the first occurrence are returned.

Examples

>>> a = np.arange(6).reshape(2,3) + 10
>>> a
array([[10, 11, 12],
       [13, 14, 15]])
>>> np.argmin(a)
0
>>> np.argmin(a, axis=0)
array([0, 0, 0])
>>> np.argmin(a, axis=1)
array([0, 0])

Indices of the minimum elements of a N-dimensional array:

>>> ind = np.unravel_index(np.argmin(a, axis=None), a.shape)
>>> ind
(0, 0)
>>> a[ind]
10
>>> b = np.arange(6) + 10
>>> b[4] = 10
>>> b
array([10, 11, 12, 13, 10, 15])
>>> np.argmin(b)  # Only the first occurrence is returned.
0
dask.array.argtopk(a, k, axis=-1, split_every=None)

Extract the indices of the k largest elements from a on the given axis, and return them sorted from largest to smallest. If k is negative, extract the indices of the -k smallest elements instead, and return them sorted from smallest to largest.

This performs best when k is much smaller than the chunk size. All results will be returned in a single chunk along the given axis.

Parameters:
x: Array

Data being sorted

k: int
axis: int, optional
split_every: int >=2, optional

See topk(). The performance considerations for topk also apply here.

Returns:
Selection of np.intp indices of x with size abs(k) along the given axis.

Examples

>>> import dask.array as da
>>> x = np.array([5, 1, 3, 6])
>>> d = da.from_array(x, chunks=2)
>>> d.argtopk(2).compute()
array([3, 0])
>>> d.argtopk(-2).compute()
array([1, 2])
dask.array.argwhere(a)

Find the indices of array elements that are non-zero, grouped by element.

This docstring was copied from numpy.argwhere.

Some inconsistencies with the Dask version may exist.

Parameters:
a : array_like

Input data.

Returns:
index_array : ndarray

Indices of elements that are non-zero. Indices are grouped by element.

See also

where, nonzero

Notes

np.argwhere(a) is the same as np.transpose(np.nonzero(a)).

The output of argwhere is not suitable for indexing arrays. For this purpose use nonzero(a) instead.

Examples

>>> x = np.arange(6).reshape(2,3)  # doctest: +SKIP
>>> x  # doctest: +SKIP
array([[0, 1, 2],
       [3, 4, 5]])
>>> np.argwhere(x>1)  # doctest: +SKIP
array([[0, 2],
       [1, 0],
       [1, 1],
       [1, 2]])
dask.array.around(x, decimals=0)

Evenly round to the given number of decimals.

This docstring was copied from numpy.around.

Some inconsistencies with the Dask version may exist.

Parameters:
a : array_like (Not supported in Dask)

Input data.

decimals : int, optional

Number of decimal places to round to (default: 0). If decimals is negative, it specifies the number of positions to the left of the decimal point.

out : ndarray, optional (Not supported in Dask)

Alternative output array in which to place the result. It must have the same shape as the expected output, but the type of the output values will be cast if necessary. See doc.ufuncs (Section “Output arguments”) for details.

Returns:
rounded_array : ndarray

An array of the same type as a, containing the rounded values. Unless out was specified, a new array is created. A reference to the result is returned.

The real and imaginary parts of complex numbers are rounded separately. The result of rounding a float is a float.

See also

ndarray.round
equivalent method

ceil, fix, floor, rint, trunc

Notes

For values exactly halfway between rounded decimal values, NumPy rounds to the nearest even value. Thus 1.5 and 2.5 round to 2.0, -0.5 and 0.5 round to 0.0, etc. Results may also be surprising due to the inexact representation of decimal fractions in the IEEE floating point standard [1] and errors introduced when scaling by powers of ten.

References

[1](1, 2) “Lecture Notes on the Status of IEEE 754”, William Kahan, https://people.eecs.berkeley.edu/~wkahan/ieee754status/IEEE754.PDF
[2]“How Futile are Mindless Assessments of Roundoff in Floating-Point Computation?”, William Kahan, https://people.eecs.berkeley.edu/~wkahan/Mindless.pdf

Examples

>>> np.around([0.37, 1.64])  # doctest: +SKIP
array([ 0.,  2.])
>>> np.around([0.37, 1.64], decimals=1)  # doctest: +SKIP
array([ 0.4,  1.6])
>>> np.around([.5, 1.5, 2.5, 3.5, 4.5]) # rounds to nearest even value  # doctest: +SKIP
array([ 0.,  2.,  2.,  4.,  4.])
>>> np.around([1,2,3,11], decimals=1) # ndarray of ints is returned  # doctest: +SKIP
array([ 1,  2,  3, 11])
>>> np.around([1,2,3,11], decimals=-1)  # doctest: +SKIP
array([ 0,  0,  0, 10])
dask.array.array(object, dtype=None, copy=True, order='K', subok=False, ndmin=0)

This docstring was copied from numpy.array.

Some inconsistencies with the Dask version may exist.

Create an array.

Parameters:
object : array_like

An array, any object exposing the array interface, an object whose __array__ method returns an array, or any (nested) sequence.

dtype : data-type, optional

The desired data-type for the array. If not given, then the type will be determined as the minimum type required to hold the objects in the sequence. This argument can only be used to ‘upcast’ the array. For downcasting, use the .astype(t) method.

copy : bool, optional

If true (default), then the object is copied. Otherwise, a copy will only be made if __array__ returns a copy, if obj is a nested sequence, or if a copy is needed to satisfy any of the other requirements (dtype, order, etc.).

order : {‘K’, ‘A’, ‘C’, ‘F’}, optional

Specify the memory layout of the array. If object is not an array, the newly created array will be in C order (row major) unless ‘F’ is specified, in which case it will be in Fortran order (column major). If object is an array the following holds.

order no copy copy=True
‘K’ unchanged F & C order preserved, otherwise most similar order
‘A’ unchanged F order if input is F and not C, otherwise C order
‘C’ C order C order
‘F’ F order F order

When copy=False and a copy is made for other reasons, the result is the same as if copy=True, with some exceptions for A, see the Notes section. The default order is ‘K’.

subok : bool, optional

If True, then sub-classes will be passed-through, otherwise the returned array will be forced to be a base-class array (default).

ndmin : int, optional

Specifies the minimum number of dimensions that the resulting array should have. Ones will be pre-pended to the shape as needed to meet this requirement.

Returns:
out : ndarray

An array object satisfying the specified requirements.

See also

empty_like
Return an empty array with shape and type of input.
ones_like
Return an array of ones with shape and type of input.
zeros_like
Return an array of zeros with shape and type of input.
full_like
Return a new array with shape of input filled with value.
empty
Return a new uninitialized array.
ones
Return a new array setting values to one.
zeros
Return a new array setting values to zero.
full
Return a new array of given shape filled with value.

Notes

When order is ‘A’ and object is an array in neither ‘C’ nor ‘F’ order, and a copy is forced by a change in dtype, then the order of the result is not necessarily ‘C’ as expected. This is likely a bug.

Examples

>>> np.array([1, 2, 3])  # doctest: +SKIP
array([1, 2, 3])

Upcasting:

>>> np.array([1, 2, 3.0])  # doctest: +SKIP
array([ 1.,  2.,  3.])

More than one dimension:

>>> np.array([[1, 2], [3, 4]])  # doctest: +SKIP
array([[1, 2],
       [3, 4]])

Minimum dimensions 2:

>>> np.array([1, 2, 3], ndmin=2)  # doctest: +SKIP
array([[1, 2, 3]])

Type provided:

>>> np.array([1, 2, 3], dtype=complex)  # doctest: +SKIP
array([ 1.+0.j,  2.+0.j,  3.+0.j])

Data-type consisting of more than one element:

>>> x = np.array([(1,2),(3,4)],dtype=[('a','<i4'),('b','<i4')])  # doctest: +SKIP
>>> x['a']  # doctest: +SKIP
array([1, 3])

Creating an array from sub-classes:

>>> np.array(np.mat('1 2; 3 4'))  # doctest: +SKIP
array([[1, 2],
       [3, 4]])
>>> np.array(np.mat('1 2; 3 4'), subok=True)  # doctest: +SKIP
matrix([[1, 2],
        [3, 4]])
dask.array.asanyarray(a)

Convert the input to a dask array.

Subclasses of np.ndarray will be passed through as chunks unchanged.

Parameters:
a : array-like

Input data, in any form that can be converted to a dask array.

Returns:
out : dask array

Dask array interpretation of a.

Examples

>>> import dask.array as da
>>> import numpy as np
>>> x = np.arange(3)
>>> da.asanyarray(x)
dask.array<array, shape=(3,), dtype=int64, chunksize=(3,)>
>>> y = [[1, 2, 3], [4, 5, 6]]
>>> da.asanyarray(y)
dask.array<array, shape=(2, 3), dtype=int64, chunksize=(2, 3)>
dask.array.asarray(a, **kwargs)

Convert the input to a dask array.

Parameters:
a : array-like

Input data, in any form that can be converted to a dask array.

Returns:
out : dask array

Dask array interpretation of a.

Examples

>>> import dask.array as da
>>> import numpy as np
>>> x = np.arange(3)
>>> da.asarray(x)
dask.array<array, shape=(3,), dtype=int64, chunksize=(3,)>
>>> y = [[1, 2, 3], [4, 5, 6]]
>>> da.asarray(y)
dask.array<array, shape=(2, 3), dtype=int64, chunksize=(2, 3)>
dask.array.atleast_1d(*arys)

Convert inputs to arrays with at least one dimension.

This docstring was copied from numpy.atleast_1d.

Some inconsistencies with the Dask version may exist.

Scalar inputs are converted to 1-dimensional arrays, whilst higher-dimensional inputs are preserved.

Parameters:
arys1, arys2, … : array_like

One or more input arrays.

Returns:
ret : ndarray

An array, or list of arrays, each with a.ndim >= 1. Copies are made only if necessary.

Examples

>>> np.atleast_1d(1.0)  # doctest: +SKIP
array([ 1.])
>>> x = np.arange(9.0).reshape(3,3)  # doctest: +SKIP
>>> np.atleast_1d(x)  # doctest: +SKIP
array([[ 0.,  1.,  2.],
       [ 3.,  4.,  5.],
       [ 6.,  7.,  8.]])
>>> np.atleast_1d(x) is x  # doctest: +SKIP
True
>>> np.atleast_1d(1, [3, 4])  # doctest: +SKIP
[array([1]), array([3, 4])]
dask.array.atleast_2d(*arys)

View inputs as arrays with at least two dimensions.

This docstring was copied from numpy.atleast_2d.

Some inconsistencies with the Dask version may exist.

Parameters:
arys1, arys2, … : array_like

One or more array-like sequences. Non-array inputs are converted to arrays. Arrays that already have two or more dimensions are preserved.

Returns:
res, res2, … : ndarray

An array, or list of arrays, each with a.ndim >= 2. Copies are avoided where possible, and views with two or more dimensions are returned.

Examples

>>> np.atleast_2d(3.0)  # doctest: +SKIP
array([[ 3.]])
>>> x = np.arange(3.0)  # doctest: +SKIP
>>> np.atleast_2d(x)  # doctest: +SKIP
array([[ 0.,  1.,  2.]])
>>> np.atleast_2d(x).base is x  # doctest: +SKIP
True
>>> np.atleast_2d(1, [1, 2], [[1, 2]])  # doctest: +SKIP
[array([[1]]), array([[1, 2]]), array([[1, 2]])]
dask.array.atleast_3d(*arys)

View inputs as arrays with at least three dimensions.

This docstring was copied from numpy.atleast_3d.

Some inconsistencies with the Dask version may exist.

Parameters:
arys1, arys2, … : array_like

One or more array-like sequences. Non-array inputs are converted to arrays. Arrays that already have three or more dimensions are preserved.

Returns:
res1, res2, … : ndarray

An array, or list of arrays, each with a.ndim >= 3. Copies are avoided where possible, and views with three or more dimensions are returned. For example, a 1-D array of shape (N,) becomes a view of shape (1, N, 1), and a 2-D array of shape (M, N) becomes a view of shape (M, N, 1).

Examples

>>> np.atleast_3d(3.0)  # doctest: +SKIP
array([[[ 3.]]])
>>> x = np.arange(3.0)  # doctest: +SKIP
>>> np.atleast_3d(x).shape  # doctest: +SKIP
(1, 3, 1)
>>> x = np.arange(12.0).reshape(4,3)  # doctest: +SKIP
>>> np.atleast_3d(x).shape  # doctest: +SKIP
(4, 3, 1)
>>> np.atleast_3d(x).base is x.base  # x is a reshape, so not base itself  # doctest: +SKIP
True
>>> for arr in np.atleast_3d([1, 2], [[1, 2]], [[[1, 2]]]):  # doctest: +SKIP
...     print(arr, arr.shape)
...
[[[1]
  [2]]] (1, 2, 1)
[[[1]
  [2]]] (1, 2, 1)
[[[1 2]]] (1, 1, 2)
dask.array.average(a, axis=None, weights=None, returned=False)

Compute the weighted average along the specified axis.

This docstring was copied from numpy.average.

Some inconsistencies with the Dask version may exist.

Parameters:
a : array_like

Array containing data to be averaged. If a is not an array, a conversion is attempted.

axis : None or int or tuple of ints, optional

Axis or axes along which to average a. The default, axis=None, will average over all of the elements of the input array. If axis is negative it counts from the last to the first axis.

New in version 1.7.0.

If axis is a tuple of ints, averaging is performed on all of the axes specified in the tuple instead of a single axis or all the axes as before.

weights : array_like, optional

An array of weights associated with the values in a. Each value in a contributes to the average according to its associated weight. The weights array can either be 1-D (in which case its length must be the size of a along the given axis) or of the same shape as a. If weights=None, then all data in a are assumed to have a weight equal to one.

returned : bool, optional

Default is False. If True, the tuple (average, sum_of_weights) is returned, otherwise only the average is returned. If weights=None, sum_of_weights is equivalent to the number of elements over which the average is taken.

Returns:
retval, [sum_of_weights] : array_type or double

Return the average along the specified axis. When returned is True, return a tuple with the average as the first element and the sum of the weights as the second element. sum_of_weights is of the same type as retval. The result dtype follows a genereal pattern. If weights is None, the result dtype will be that of a , or float64 if a is integral. Otherwise, if weights is not None and a is non- integral, the result type will be the type of lowest precision capable of representing values of both a and weights. If a happens to be integral, the previous rules still applies but the result dtype will at least be float64.

Raises:
ZeroDivisionError

When all weights along axis are zero. See numpy.ma.average for a version robust to this type of error.

TypeError

When the length of 1D weights is not the same as the shape of a along axis.

See also

mean

ma.average
average for masked arrays – useful if your data contains “missing” values
numpy.result_type
Returns the type that results from applying the numpy type promotion rules to the arguments.

Examples

>>> data = range(1,5)  # doctest: +SKIP
>>> data  # doctest: +SKIP
[1, 2, 3, 4]
>>> np.average(data)  # doctest: +SKIP
2.5
>>> np.average(range(1,11), weights=range(10,0,-1))  # doctest: +SKIP
4.0
>>> data = np.arange(6).reshape((3,2))  # doctest: +SKIP
>>> data  # doctest: +SKIP
array([[0, 1],
       [2, 3],
       [4, 5]])
>>> np.average(data, axis=1, weights=[1./4, 3./4])  # doctest: +SKIP
array([ 0.75,  2.75,  4.75])
>>> np.average(data, weights=[1./4, 3./4])  # doctest: +SKIP

Traceback (most recent call last): … TypeError: Axis must be specified when shapes of a and weights differ.

>>> a = np.ones(5, dtype=np.float128)  # doctest: +SKIP
>>> w = np.ones(5, dtype=np.complex64)  # doctest: +SKIP
>>> avg = np.average(a, weights=w)  # doctest: +SKIP
>>> print(avg.dtype)  # doctest: +SKIP
complex256
dask.array.bincount(x, weights=None, minlength=0)

This docstring was copied from numpy.bincount.

Some inconsistencies with the Dask version may exist.

Count number of occurrences of each value in array of non-negative ints.

The number of bins (of size 1) is one larger than the largest value in x. If minlength is specified, there will be at least this number of bins in the output array (though it will be longer if necessary, depending on the contents of x). Each bin gives the number of occurrences of its index value in x. If weights is specified the input array is weighted by it, i.e. if a value n is found at position i, out[n] += weight[i] instead of out[n] += 1.

Parameters:
x : array_like, 1 dimension, nonnegative ints

Input array.

weights : array_like, optional

Weights, array of the same shape as x.

minlength : int, optional

A minimum number of bins for the output array.

New in version 1.6.0.

Returns:
out : ndarray of ints

The result of binning the input array. The length of out is equal to np.amax(x)+1.

Raises:
ValueError

If the input is not 1-dimensional, or contains elements with negative values, or if minlength is negative.

TypeError

If the type of the input is float or complex.

Examples

>>> np.bincount(np.arange(5))  # doctest: +SKIP
array([1, 1, 1, 1, 1])
>>> np.bincount(np.array([0, 1, 1, 3, 2, 1, 7]))  # doctest: +SKIP
array([1, 3, 1, 1, 0, 0, 0, 1])
>>> x = np.array([0, 1, 1, 3, 2, 1, 7, 23])  # doctest: +SKIP
>>> np.bincount(x).size == np.amax(x)+1  # doctest: +SKIP
True

The input array needs to be of integer dtype, otherwise a TypeError is raised:

>>> np.bincount(np.arange(5, dtype=float))  # doctest: +SKIP
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
TypeError: array cannot be safely cast to required type

A possible use of bincount is to perform sums over variable-size chunks of an array, using the weights keyword.

>>> w = np.array([0.3, 0.5, 0.2, 0.7, 1., -0.6]) # weights  # doctest: +SKIP
>>> x = np.array([0, 1, 1, 2, 2, 2])  # doctest: +SKIP
>>> np.bincount(x,  weights=w)  # doctest: +SKIP
array([ 0.3,  0.7,  1.1])
dask.array.bitwise_and(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute the bit-wise AND of two arrays element-wise.

Computes the bit-wise AND of the underlying binary representation of the integers in the input arrays. This ufunc implements the C/Python operator &.

Parameters:
x1, x2 : array_like

Only integer and boolean types are handled.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Result. This is a scalar if both x1 and x2 are scalars.

See also

logical_and, bitwise_or, bitwise_xor

binary_repr
Return the binary representation of the input number as a string.

Examples

The number 13 is represented by 00001101. Likewise, 17 is represented by 00010001. The bit-wise AND of 13 and 17 is therefore 000000001, or 1:

>>> np.bitwise_and(13, 17)  # doctest: +SKIP
1
>>> np.bitwise_and(14, 13)  # doctest: +SKIP
12
>>> np.binary_repr(12)  # doctest: +SKIP
'1100'
>>> np.bitwise_and([14,3], 13)  # doctest: +SKIP
array([12,  1])
>>> np.bitwise_and([11,7], [4,25])  # doctest: +SKIP
array([0, 1])
>>> np.bitwise_and(np.array([2,5,255]), np.array([3,14,16]))  # doctest: +SKIP
array([ 2,  4, 16])
>>> np.bitwise_and([True, True], [False, True])  # doctest: +SKIP
array([False,  True])
dask.array.bitwise_not(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute bit-wise inversion, or bit-wise NOT, element-wise.

Computes the bit-wise NOT of the underlying binary representation of the integers in the input arrays. This ufunc implements the C/Python operator ~.

For signed integer inputs, the two’s complement is returned. In a two’s-complement system negative numbers are represented by the two’s complement of the absolute value. This is the most common method of representing signed integers on computers [1]. A N-bit two’s-complement system can represent every integer in the range \(-2^{N-1}\) to \(+2^{N-1}-1\).

Parameters:
x : array_like

Only integer and boolean types are handled.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Result. This is a scalar if x is a scalar.

See also

bitwise_and, bitwise_or, bitwise_xor, logical_not

binary_repr
Return the binary representation of the input number as a string.

Notes

bitwise_not is an alias for invert:

>>> np.bitwise_not is np.invert  # doctest: +SKIP
True

References

[1](1, 2) Wikipedia, “Two’s complement”, https://en.wikipedia.org/wiki/Two’s_complement

Examples

We’ve seen that 13 is represented by 00001101. The invert or bit-wise NOT of 13 is then:

>>> np.invert(np.array([13], dtype=uint8))  # doctest: +SKIP
array([242], dtype=uint8)
>>> np.binary_repr(x, width=8)  # doctest: +SKIP
'00001101'
>>> np.binary_repr(242, width=8)  # doctest: +SKIP
'11110010'

The result depends on the bit-width:

>>> np.invert(np.array([13], dtype=uint16))  # doctest: +SKIP
array([65522], dtype=uint16)
>>> np.binary_repr(x, width=16)  # doctest: +SKIP
'0000000000001101'
>>> np.binary_repr(65522, width=16)  # doctest: +SKIP
'1111111111110010'

When using signed integer types the result is the two’s complement of the result for the unsigned type:

>>> np.invert(np.array([13], dtype=int8))  # doctest: +SKIP
array([-14], dtype=int8)
>>> np.binary_repr(-14, width=8)  # doctest: +SKIP
'11110010'

Booleans are accepted as well:

>>> np.invert(array([True, False]))  # doctest: +SKIP
array([False,  True])
dask.array.bitwise_or(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute the bit-wise OR of two arrays element-wise.

Computes the bit-wise OR of the underlying binary representation of the integers in the input arrays. This ufunc implements the C/Python operator |.

Parameters:
x1, x2 : array_like

Only integer and boolean types are handled.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Result. This is a scalar if both x1 and x2 are scalars.

See also

logical_or, bitwise_and, bitwise_xor

binary_repr
Return the binary representation of the input number as a string.

Examples

The number 13 has the binaray representation 00001101. Likewise, 16 is represented by 00010000. The bit-wise OR of 13 and 16 is then 000111011, or 29:

>>> np.bitwise_or(13, 16)  # doctest: +SKIP
29
>>> np.binary_repr(29)  # doctest: +SKIP
'11101'
>>> np.bitwise_or(32, 2)  # doctest: +SKIP
34
>>> np.bitwise_or([33, 4], 1)  # doctest: +SKIP
array([33,  5])
>>> np.bitwise_or([33, 4], [1, 2])  # doctest: +SKIP
array([33,  6])
>>> np.bitwise_or(np.array([2, 5, 255]), np.array([4, 4, 4]))  # doctest: +SKIP
array([  6,   5, 255])
>>> np.array([2, 5, 255]) | np.array([4, 4, 4])  # doctest: +SKIP
array([  6,   5, 255])
>>> np.bitwise_or(np.array([2, 5, 255, 2147483647L], dtype=np.int32),  # doctest: +SKIP
...               np.array([4, 4, 4, 2147483647L], dtype=np.int32))
array([         6,          5,        255, 2147483647])
>>> np.bitwise_or([True, True], [False, True])  # doctest: +SKIP
array([ True,  True])
dask.array.bitwise_xor(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute the bit-wise XOR of two arrays element-wise.

Computes the bit-wise XOR of the underlying binary representation of the integers in the input arrays. This ufunc implements the C/Python operator ^.

Parameters:
x1, x2 : array_like

Only integer and boolean types are handled.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Result. This is a scalar if both x1 and x2 are scalars.

See also

logical_xor, bitwise_and, bitwise_or

binary_repr
Return the binary representation of the input number as a string.

Examples

The number 13 is represented by 00001101. Likewise, 17 is represented by 00010001. The bit-wise XOR of 13 and 17 is therefore 00011100, or 28:

>>> np.bitwise_xor(13, 17)  # doctest: +SKIP
28
>>> np.binary_repr(28)  # doctest: +SKIP
'11100'
>>> np.bitwise_xor(31, 5)  # doctest: +SKIP
26
>>> np.bitwise_xor([31,3], 5)  # doctest: +SKIP
array([26,  6])
>>> np.bitwise_xor([31,3], [5,6])  # doctest: +SKIP
array([26,  5])
>>> np.bitwise_xor([True, True], [False, True])  # doctest: +SKIP
array([ True, False])
dask.array.block(arrays, allow_unknown_chunksizes=False)

Assemble an nd-array from nested lists of blocks.

Blocks in the innermost lists are concatenated along the last dimension (-1), then these are concatenated along the second-last dimension (-2), and so on until the outermost list is reached

Blocks can be of any dimension, but will not be broadcasted using the normal rules. Instead, leading axes of size 1 are inserted, to make block.ndim the same for all blocks. This is primarily useful for working with scalars, and means that code like block([v, 1]) is valid, where v.ndim == 1.

When the nested list is two levels deep, this allows block matrices to be constructed from their components.

Parameters:
arrays : nested list of array_like or scalars (but not tuples)

If passed a single ndarray or scalar (a nested list of depth 0), this is returned unmodified (and not copied).

Elements shapes must match along the appropriate axes (without broadcasting), but leading 1s will be prepended to the shape as necessary to make the dimensions match.

allow_unknown_chunksizes: bool

Allow unknown chunksizes, such as come from converting from dask dataframes. Dask.array is unable to verify that chunks line up. If data comes from differently aligned sources then this can cause unexpected results.

Returns:
block_array : ndarray

The array assembled from the given blocks.

The dimensionality of the output is equal to the greatest of: * the dimensionality of all the inputs * the depth to which the input list is nested

Raises:
ValueError
  • If list depths are mismatched - for instance, [[a, b], c] is illegal, and should be spelt [[a, b], [c]]
  • If lists are empty - for instance, [[a, b], []]

See also

concatenate
Join a sequence of arrays together.
stack
Stack arrays in sequence along a new dimension.
hstack
Stack arrays in sequence horizontally (column wise).
vstack
Stack arrays in sequence vertically (row wise).
dstack
Stack arrays in sequence depth wise (along third dimension).
vsplit
Split array into a list of multiple sub-arrays vertically.

Notes

When called with only scalars, block is equivalent to an ndarray call. So block([[1, 2], [3, 4]]) is equivalent to array([[1, 2], [3, 4]]).

This function does not enforce that the blocks lie on a fixed grid. block([[a, b], [c, d]]) is not restricted to arrays of the form:

AAAbb
AAAbb
cccDD

But is also allowed to produce, for some a, b, c, d:

AAAbb
AAAbb
cDDDD

Since concatenation happens along the last axis first, block is _not_ capable of producing the following directly:

AAAbb
cccbb
cccDD

Matlab’s “square bracket stacking”, [A, B, ...; p, q, ...], is equivalent to block([[A, B, ...], [p, q, ...]]).

dask.array.broadcast_arrays(*args, **kwargs)

Broadcast any number of arrays against each other.

This docstring was copied from numpy.broadcast_arrays.

Some inconsistencies with the Dask version may exist.

Parameters:
`*args` : array_likes

The arrays to broadcast.

subok : bool, optional

If True, then sub-classes will be passed-through, otherwise the returned arrays will be forced to be a base-class array (default).

Returns:
broadcasted : list of arrays

These arrays are views on the original arrays. They are typically not contiguous. Furthermore, more than one element of a broadcasted array may refer to a single memory location. If you need to write to the arrays, make copies first.

Examples

>>> x = np.array([[1,2,3]])  # doctest: +SKIP
>>> y = np.array([[4],[5]])  # doctest: +SKIP
>>> np.broadcast_arrays(x, y)  # doctest: +SKIP
[array([[1, 2, 3],
       [1, 2, 3]]), array([[4, 4, 4],
       [5, 5, 5]])]

Here is a useful idiom for getting contiguous copies instead of non-contiguous views.

>>> [np.array(a) for a in np.broadcast_arrays(x, y)]  # doctest: +SKIP
[array([[1, 2, 3],
       [1, 2, 3]]), array([[4, 4, 4],
       [5, 5, 5]])]
dask.array.broadcast_to(x, shape, chunks=None)

Broadcast an array to a new shape.

Parameters:
x : array_like

The array to broadcast.

shape : tuple

The shape of the desired array.

chunks : tuple, optional

If provided, then the result will use these chunks instead of the same chunks as the source array. Setting chunks explicitly as part of broadcast_to is more efficient than rechunking afterwards. Chunks are only allowed to differ from the original shape along dimensions that are new on the result or have size 1 the input array.

Returns:
broadcast : dask array
dask.array.coarsen(reduction, x, axes, trim_excess=False)

Coarsen array by applying reduction to fixed size neighborhoods

Parameters:
reduction: function

Function like np.sum, np.mean, etc…

x: np.ndarray

Array to be coarsened

axes: dict

Mapping of axis to coarsening factor

Examples

>>> x = np.array([1, 2, 3, 4, 5, 6])
>>> coarsen(np.sum, x, {0: 2})
array([ 3,  7, 11])
>>> coarsen(np.max, x, {0: 3})
array([3, 6])

Provide dictionary of scale per dimension

>>> x = np.arange(24).reshape((4, 6))
>>> x
array([[ 0,  1,  2,  3,  4,  5],
       [ 6,  7,  8,  9, 10, 11],
       [12, 13, 14, 15, 16, 17],
       [18, 19, 20, 21, 22, 23]])
>>> coarsen(np.min, x, {0: 2, 1: 3})
array([[ 0,  3],
       [12, 15]])

You must avoid excess elements explicitly

>>> x = np.array([1, 2, 3, 4, 5, 6, 7, 8])
>>> coarsen(np.min, x, {0: 3}, trim_excess=True)
array([1, 4])
dask.array.ceil(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Return the ceiling of the input, element-wise.

The ceil of the scalar x is the smallest integer i, such that i >= x. It is often denoted as \(\lceil x \rceil\).

Parameters:
x : array_like

Input data.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or scalar

The ceiling of each element in x, with float dtype. This is a scalar if x is a scalar.

See also

floor, trunc, rint

Examples

>>> a = np.array([-1.7, -1.5, -0.2, 0.2, 1.5, 1.7, 2.0])  # doctest: +SKIP
>>> np.ceil(a)  # doctest: +SKIP
array([-1., -1., -0.,  1.,  2.,  2.,  2.])
dask.array.choose(a, choices)

Construct an array from an index array and a set of arrays to choose from.

This docstring was copied from numpy.choose.

Some inconsistencies with the Dask version may exist.

First of all, if confused or uncertain, definitely look at the Examples - in its full generality, this function is less simple than it might seem from the following code description (below ndi = numpy.lib.index_tricks):

np.choose(a,c) == np.array([c[a[I]][I] for I in ndi.ndindex(a.shape)]).

But this omits some subtleties. Here is a fully general summary:

Given an “index” array (a) of integers and a sequence of n arrays (choices), a and each choice array are first broadcast, as necessary, to arrays of a common shape; calling these Ba and Bchoices[i], i = 0,…,n-1 we have that, necessarily, Ba.shape == Bchoices[i].shape for each i. Then, a new array with shape Ba.shape is created as follows:

  • if mode=raise (the default), then, first of all, each element of a (and thus Ba) must be in the range [0, n-1]; now, suppose that i (in that range) is the value at the (j0, j1, …, jm) position in Ba - then the value at the same position in the new array is the value in Bchoices[i] at that same position;
  • if mode=wrap, values in a (and thus Ba) may be any (signed) integer; modular arithmetic is used to map integers outside the range [0, n-1] back into that range; and then the new array is constructed as above;
  • if mode=clip, values in a (and thus Ba) may be any (signed) integer; negative integers are mapped to 0; values greater than n-1 are mapped to n-1; and then the new array is constructed as above.
Parameters:
a : int array

This array must contain integers in [0, n-1], where n is the number of choices, unless mode=wrap or mode=clip, in which cases any integers are permissible.

choices : sequence of arrays

Choice arrays. a and all of the choices must be broadcastable to the same shape. If choices is itself an array (not recommended), then its outermost dimension (i.e., the one corresponding to choices.shape[0]) is taken as defining the “sequence”.

out : array, optional (Not supported in Dask)

If provided, the result will be inserted into this array. It should be of the appropriate shape and dtype.

mode : {‘raise’ (default), ‘wrap’, ‘clip’}, optional (Not supported in Dask)

Specifies how indices outside [0, n-1] will be treated:

  • ‘raise’ : an exception is raised
  • ‘wrap’ : value becomes value mod n
  • ‘clip’ : values < 0 are mapped to 0, values > n-1 are mapped to n-1
Returns:
merged_array : array

The merged result.

Raises:
ValueError: shape mismatch

If a and each choice array are not all broadcastable to the same shape.

See also

ndarray.choose
equivalent method

Notes

To reduce the chance of misinterpretation, even though the following “abuse” is nominally supported, choices should neither be, nor be thought of as, a single array, i.e., the outermost sequence-like container should be either a list or a tuple.

Examples

>>> choices = [[0, 1, 2, 3], [10, 11, 12, 13],  # doctest: +SKIP
...   [20, 21, 22, 23], [30, 31, 32, 33]]
>>> np.choose([2, 3, 1, 0], choices  # doctest: +SKIP
... # the first element of the result will be the first element of the
... # third (2+1) "array" in choices, namely, 20; the second element
... # will be the second element of the fourth (3+1) choice array, i.e.,
... # 31, etc.
... )
array([20, 31, 12,  3])
>>> np.choose([2, 4, 1, 0], choices, mode='clip') # 4 goes to 3 (4-1)  # doctest: +SKIP
array([20, 31, 12,  3])
>>> # because there are 4 choice arrays
>>> np.choose([2, 4, 1, 0], choices, mode='wrap') # 4 goes to (4 mod 4)  # doctest: +SKIP
array([20,  1, 12,  3])
>>> # i.e., 0

A couple examples illustrating how choose broadcasts:

>>> a = [[1, 0, 1], [0, 1, 0], [1, 0, 1]]  # doctest: +SKIP
>>> choices = [-10, 10]  # doctest: +SKIP
>>> np.choose(a, choices)  # doctest: +SKIP
array([[ 10, -10,  10],
       [-10,  10, -10],
       [ 10, -10,  10]])
>>> # With thanks to Anne Archibald
>>> a = np.array([0, 1]).reshape((2,1,1))  # doctest: +SKIP
>>> c1 = np.array([1, 2, 3]).reshape((1,3,1))  # doctest: +SKIP
>>> c2 = np.array([-1, -2, -3, -4, -5]).reshape((1,1,5))  # doctest: +SKIP
>>> np.choose(a, (c1, c2)) # result is 2x3x5, res[0,:,:]=c1, res[1,:,:]=c2  # doctest: +SKIP
array([[[ 1,  1,  1,  1,  1],
        [ 2,  2,  2,  2,  2],
        [ 3,  3,  3,  3,  3]],
       [[-1, -2, -3, -4, -5],
        [-1, -2, -3, -4, -5],
        [-1, -2, -3, -4, -5]]])
dask.array.clip(*args, **kwargs)

Clip (limit) the values in an array.

Given an interval, values outside the interval are clipped to the interval edges. For example, if an interval of [0, 1] is specified, values smaller than 0 become 0, and values larger than 1 become 1.

Parameters:
a : array_like

Array containing elements to clip.

a_min : scalar or array_like or None

Minimum value. If None, clipping is not performed on lower interval edge. Not more than one of a_min and a_max may be None.

a_max : scalar or array_like or None

Maximum value. If None, clipping is not performed on upper interval edge. Not more than one of a_min and a_max may be None. If a_min or a_max are array_like, then the three arrays will be broadcasted to match their shapes.

out : ndarray, optional

The results will be placed in this array. It may be the input array for in-place clipping. out must be of the right shape to hold the output. Its type is preserved.

Returns:
clipped_array : ndarray

An array with the elements of a, but where values < a_min are replaced with a_min, and those > a_max with a_max.

See also

numpy.doc.ufuncs
Section “Output arguments”

Examples

>>> a = np.arange(10)  # doctest: +SKIP
>>> np.clip(a, 1, 8)  # doctest: +SKIP
array([1, 1, 2, 3, 4, 5, 6, 7, 8, 8])
>>> a  # doctest: +SKIP
array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
>>> np.clip(a, 3, 6, out=a)  # doctest: +SKIP
array([3, 3, 3, 3, 4, 5, 6, 6, 6, 6])
>>> a = np.arange(10)  # doctest: +SKIP
>>> a  # doctest: +SKIP
array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
>>> np.clip(a, [3, 4, 1, 1, 1, 4, 4, 4, 4, 4], 8)  # doctest: +SKIP
array([3, 4, 2, 3, 4, 5, 6, 7, 8, 8])
dask.array.compress(condition, a, axis=None)

Return selected slices of an array along given axis.

This docstring was copied from numpy.compress.

Some inconsistencies with the Dask version may exist.

When working along a given axis, a slice along that axis is returned in output for each index where condition evaluates to True. When working on a 1-D array, compress is equivalent to extract.

Parameters:
condition : 1-D array of bools

Array that selects which entries to return. If len(condition) is less than the size of a along the given axis, then output is truncated to the length of the condition array.

a : array_like

Array from which to extract a part.

axis : int, optional

Axis along which to take slices. If None (default), work on the flattened array.

out : ndarray, optional (Not supported in Dask)

Output array. Its type is preserved and it must be of the right shape to hold the output.

Returns:
compressed_array : ndarray

A copy of a without the slices along axis for which condition is false.

See also

take, choose, diag, diagonal, select

ndarray.compress
Equivalent method in ndarray
np.extract
Equivalent method when working on 1-D arrays
numpy.doc.ufuncs
Section “Output arguments”

Examples

>>> a = np.array([[1, 2], [3, 4], [5, 6]])  # doctest: +SKIP
>>> a  # doctest: +SKIP
array([[1, 2],
       [3, 4],
       [5, 6]])
>>> np.compress([0, 1], a, axis=0)  # doctest: +SKIP
array([[3, 4]])
>>> np.compress([False, True, True], a, axis=0)  # doctest: +SKIP
array([[3, 4],
       [5, 6]])
>>> np.compress([False, True], a, axis=1)  # doctest: +SKIP
array([[2],
       [4],
       [6]])

Working on the flattened array does not return slices along an axis but selects elements.

>>> np.compress([False, True], a)  # doctest: +SKIP
array([2])
dask.array.concatenate(seq, axis=0, allow_unknown_chunksizes=False)

Concatenate arrays along an existing axis

Given a sequence of dask Arrays form a new dask Array by stacking them along an existing dimension (axis=0 by default)

Parameters:
seq: list of dask.arrays
axis: int

Dimension along which to align all of the arrays

allow_unknown_chunksizes: bool

Allow unknown chunksizes, such as come from converting from dask dataframes. Dask.array is unable to verify that chunks line up. If data comes from differently aligned sources then this can cause unexpected results.

See also

stack

Examples

Create slices

>>> import dask.array as da
>>> import numpy as np
>>> data = [from_array(np.ones((4, 4)), chunks=(2, 2))
...          for i in range(3)]
>>> x = da.concatenate(data, axis=0)
>>> x.shape
(12, 4)
>>> da.concatenate(data, axis=1).shape
(4, 12)

Result is a new dask Array

dask.array.conj(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Return the complex conjugate, element-wise.

The complex conjugate of a complex number is obtained by changing the sign of its imaginary part.

Parameters:
x : array_like

Input value.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray

The complex conjugate of x, with same dtype as y. This is a scalar if x is a scalar.

Examples

>>> np.conjugate(1+2j)  # doctest: +SKIP
(1-2j)
>>> x = np.eye(2) + 1j * np.eye(2)  # doctest: +SKIP
>>> np.conjugate(x)  # doctest: +SKIP
array([[ 1.-1.j,  0.-0.j],
       [ 0.-0.j,  1.-1.j]])
dask.array.copysign(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Change the sign of x1 to that of x2, element-wise.

If both arguments are arrays or sequences, they have to be of the same length. If x2 is a scalar, its sign will be copied to all elements of x1.

Parameters:
x1 : array_like

Values to change the sign of.

x2 : array_like

The sign of x2 is copied to x1.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

The values of x1 with the sign of x2. This is a scalar if both x1 and x2 are scalars.

Examples

>>> np.copysign(1.3, -1)  # doctest: +SKIP
-1.3
>>> 1/np.copysign(0, 1)  # doctest: +SKIP
inf
>>> 1/np.copysign(0, -1)  # doctest: +SKIP
-inf
>>> np.copysign([-1, 0, 1], -1.1)  # doctest: +SKIP
array([-1., -0., -1.])
>>> np.copysign([-1, 0, 1], np.arange(3)-1)  # doctest: +SKIP
array([-1.,  0.,  1.])
dask.array.corrcoef(x, y=None, rowvar=1)

Return Pearson product-moment correlation coefficients.

This docstring was copied from numpy.corrcoef.

Some inconsistencies with the Dask version may exist.

Please refer to the documentation for cov for more detail. The relationship between the correlation coefficient matrix, R, and the covariance matrix, C, is

\[R_{ij} = \frac{ C_{ij} } { \sqrt{ C_{ii} * C_{jj} } }\]

The values of R are between -1 and 1, inclusive.

Parameters:
x : array_like

A 1-D or 2-D array containing multiple variables and observations. Each row of x represents a variable, and each column a single observation of all those variables. Also see rowvar below.

y : array_like, optional

An additional set of variables and observations. y has the same shape as x.

rowvar : bool, optional

If rowvar is True (default), then each row represents a variable, with observations in the columns. Otherwise, the relationship is transposed: each column represents a variable, while the rows contain observations.

bias : _NoValue, optional (Not supported in Dask)

Has no effect, do not use.

Deprecated since version 1.10.0.

ddof : _NoValue, optional (Not supported in Dask)

Has no effect, do not use.

Deprecated since version 1.10.0.

Returns:
R : ndarray

The correlation coefficient matrix of the variables.

See also

cov
Covariance matrix

Notes

Due to floating point rounding the resulting array may not be Hermitian, the diagonal elements may not be 1, and the elements may not satisfy the inequality abs(a) <= 1. The real and imaginary parts are clipped to the interval [-1, 1] in an attempt to improve on that situation but is not much help in the complex case.

This function accepts but discards arguments bias and ddof. This is for backwards compatibility with previous versions of this function. These arguments had no effect on the return values of the function and can be safely ignored in this and previous versions of numpy.

dask.array.cos(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Cosine element-wise.

Parameters:
x : array_like

Input array in radians.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray

The corresponding cosine values. This is a scalar if x is a scalar.

Notes

If out is provided, the function writes the result into it, and returns a reference to out. (See Examples)

References

M. Abramowitz and I. A. Stegun, Handbook of Mathematical Functions. New York, NY: Dover, 1972.

Examples

>>> np.cos(np.array([0, np.pi/2, np.pi]))  # doctest: +SKIP
array([  1.00000000e+00,   6.12303177e-17,  -1.00000000e+00])
>>>
>>> # Example of providing the optional output parameter
>>> out2 = np.cos([0.1], out1)  # doctest: +SKIP
>>> out2 is out1  # doctest: +SKIP
True
>>>
>>> # Example of ValueError due to provision of shape mis-matched `out`
>>> np.cos(np.zeros((3,3)),np.zeros((2,2)))  # doctest: +SKIP
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ValueError: invalid return array shape
dask.array.cosh(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Hyperbolic cosine, element-wise.

Equivalent to 1/2 * (np.exp(x) + np.exp(-x)) and np.cos(1j*x).

Parameters:
x : array_like

Input array.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Output array of same shape as x. This is a scalar if x is a scalar.

Examples

>>> np.cosh(0)  # doctest: +SKIP
1.0

The hyperbolic cosine describes the shape of a hanging cable:

>>> import matplotlib.pyplot as plt  # doctest: +SKIP
>>> x = np.linspace(-4, 4, 1000)  # doctest: +SKIP
>>> plt.plot(x, np.cosh(x))  # doctest: +SKIP
>>> plt.show()  # doctest: +SKIP
dask.array.count_nonzero(a, axis=None)

Counts the number of non-zero values in the array a.

This docstring was copied from numpy.count_nonzero.

Some inconsistencies with the Dask version may exist.

The word “non-zero” is in reference to the Python 2.x built-in method __nonzero__() (renamed __bool__() in Python 3.x) of Python objects that tests an object’s “truthfulness”. For example, any number is considered truthful if it is nonzero, whereas any string is considered truthful if it is not the empty string. Thus, this function (recursively) counts how many elements in a (and in sub-arrays thereof) have their __nonzero__() or __bool__() method evaluated to True.

Parameters:
a : array_like

The array for which to count non-zeros.

axis : int or tuple, optional

Axis or tuple of axes along which to count non-zeros. Default is None, meaning that non-zeros will be counted along a flattened version of a.

New in version 1.12.0.

Returns:
count : int or array of int

Number of non-zero values in the array along a given axis. Otherwise, the total number of non-zero values in the array is returned.

See also

nonzero
Return the coordinates of all the non-zero values.

Examples

>>> np.count_nonzero(np.eye(4))  # doctest: +SKIP
4
>>> np.count_nonzero([[0,1,7,0,0],[3,0,0,2,19]])  # doctest: +SKIP
5
>>> np.count_nonzero([[0,1,7,0,0],[3,0,0,2,19]], axis=0)  # doctest: +SKIP
array([1, 1, 1, 1, 1])
>>> np.count_nonzero([[0,1,7,0,0],[3,0,0,2,19]], axis=1)  # doctest: +SKIP
array([2, 3])
dask.array.cov(m, y=None, rowvar=1, bias=0, ddof=None)

Estimate a covariance matrix, given data and weights.

This docstring was copied from numpy.cov.

Some inconsistencies with the Dask version may exist.

Covariance indicates the level to which two variables vary together. If we examine N-dimensional samples, \(X = [x_1, x_2, ... x_N]^T\), then the covariance matrix element \(C_{ij}\) is the covariance of \(x_i\) and \(x_j\). The element \(C_{ii}\) is the variance of \(x_i\).

See the notes for an outline of the algorithm.

Parameters:
m : array_like

A 1-D or 2-D array containing multiple variables and observations. Each row of m represents a variable, and each column a single observation of all those variables. Also see rowvar below.

y : array_like, optional

An additional set of variables and observations. y has the same form as that of m.

rowvar : bool, optional

If rowvar is True (default), then each row represents a variable, with observations in the columns. Otherwise, the relationship is transposed: each column represents a variable, while the rows contain observations.

bias : bool, optional

Default normalization (False) is by (N - 1), where N is the number of observations given (unbiased estimate). If bias is True, then normalization is by N. These values can be overridden by using the keyword ddof in numpy versions >= 1.5.

ddof : int, optional

If not None the default value implied by bias is overridden. Note that ddof=1 will return the unbiased estimate, even if both fweights and aweights are specified, and ddof=0 will return the simple average. See the notes for the details. The default value is None.

New in version 1.5.

fweights : array_like, int, optional (Not supported in Dask)

1-D array of integer frequency weights; the number of times each observation vector should be repeated.

New in version 1.10.

aweights : array_like, optional (Not supported in Dask)

1-D array of observation vector weights. These relative weights are typically large for observations considered “important” and smaller for observations considered less “important”. If ddof=0 the array of weights can be used to assign probabilities to observation vectors.

New in version 1.10.

Returns:
out : ndarray

The covariance matrix of the variables.

See also

corrcoef
Normalized covariance matrix

Notes

Assume that the observations are in the columns of the observation array m and let f = fweights and a = aweights for brevity. The steps to compute the weighted covariance are as follows:

>>> w = f * a  
>>> v1 = np.sum(w)  
>>> v2 = np.sum(w * a)  
>>> m -= np.sum(m * w, axis=1, keepdims=True) / v1  
>>> cov = np.dot(m * w, m.T) * v1 / (v1**2 - ddof * v2)  

Note that when a == 1, the normalization factor v1 / (v1**2 - ddof * v2) goes over to 1 / (np.sum(f) - ddof) as it should.

Examples

Consider two variables, \(x_0\) and \(x_1\), which correlate perfectly, but in opposite directions:

>>> x = np.array([[0, 2], [1, 1], [2, 0]]).T  # doctest: +SKIP
>>> x  # doctest: +SKIP
array([[0, 1, 2],
       [2, 1, 0]])

Note how \(x_0\) increases while \(x_1\) decreases. The covariance matrix shows this clearly:

>>> np.cov(x)  # doctest: +SKIP
array([[ 1., -1.],
       [-1.,  1.]])

Note that element \(C_{0,1}\), which shows the correlation between \(x_0\) and \(x_1\), is negative.

Further, note how x and y are combined:

>>> x = [-2.1, -1,  4.3]  # doctest: +SKIP
>>> y = [3,  1.1,  0.12]  # doctest: +SKIP
>>> X = np.stack((x, y), axis=0)  # doctest: +SKIP
>>> print(np.cov(X))  # doctest: +SKIP
[[ 11.71        -4.286     ]
 [ -4.286        2.14413333]]
>>> print(np.cov(x, y))  # doctest: +SKIP
[[ 11.71        -4.286     ]
 [ -4.286        2.14413333]]
>>> print(np.cov(x))  # doctest: +SKIP
11.71
dask.array.cumprod(a, axis=None, dtype=None, out=None)

Return the cumulative product of elements along a given axis.

Parameters:
a : array_like

Input array.

axis : int, optional

Axis along which the cumulative product is computed. By default the input is flattened.

dtype : dtype, optional

Type of the returned array, as well as of the accumulator in which the elements are multiplied. If dtype is not specified, it defaults to the dtype of a, unless a has an integer dtype with a precision less than that of the default platform integer. In that case, the default platform integer is used instead.

out : ndarray, optional

Alternative output array in which to place the result. It must have the same shape and buffer length as the expected output but the type of the resulting values will be cast if necessary.

Returns:
cumprod : ndarray

A new array holding the result is returned unless out is specified, in which case a reference to out is returned.

See also

numpy.doc.ufuncs
Section “Output arguments”

Notes

Arithmetic is modular when using integer types, and no error is raised on overflow.

Examples

>>> a = np.array([1,2,3])
>>> np.cumprod(a) # intermediate results 1, 1*2
...               # total product 1*2*3 = 6
array([1, 2, 6])
>>> a = np.array([[1, 2, 3], [4, 5, 6]])
>>> np.cumprod(a, dtype=float) # specify type of output
array([   1.,    2.,    6.,   24.,  120.,  720.])

The cumulative product for each column (i.e., over the rows) of a:

>>> np.cumprod(a, axis=0)
array([[ 1,  2,  3],
       [ 4, 10, 18]])

The cumulative product for each row (i.e. over the columns) of a:

>>> np.cumprod(a,axis=1)
array([[  1,   2,   6],
       [  4,  20, 120]])
dask.array.cumsum(a, axis=None, dtype=None, out=None)

Return the cumulative sum of the elements along a given axis.

Parameters:
a : array_like

Input array.

axis : int, optional

Axis along which the cumulative sum is computed. The default (None) is to compute the cumsum over the flattened array.

dtype : dtype, optional

Type of the returned array and of the accumulator in which the elements are summed. If dtype is not specified, it defaults to the dtype of a, unless a has an integer dtype with a precision less than that of the default platform integer. In that case, the default platform integer is used.

out : ndarray, optional

Alternative output array in which to place the result. It must have the same shape and buffer length as the expected output but the type will be cast if necessary. See doc.ufuncs (Section “Output arguments”) for more details.

Returns:
cumsum_along_axis : ndarray.

A new array holding the result is returned unless out is specified, in which case a reference to out is returned. The result has the same size as a, and the same shape as a if axis is not None or a is a 1-d array.

See also

sum
Sum array elements.
trapz
Integration of array values using the composite trapezoidal rule.
diff
Calculate the n-th discrete difference along given axis.

Notes

Arithmetic is modular when using integer types, and no error is raised on overflow.

Examples

>>> a = np.array([[1,2,3], [4,5,6]])
>>> a
array([[1, 2, 3],
       [4, 5, 6]])
>>> np.cumsum(a)
array([ 1,  3,  6, 10, 15, 21])
>>> np.cumsum(a, dtype=float)     # specifies type of output value(s)
array([  1.,   3.,   6.,  10.,  15.,  21.])
>>> np.cumsum(a,axis=0)      # sum over rows for each of the 3 columns
array([[1, 2, 3],
       [5, 7, 9]])
>>> np.cumsum(a,axis=1)      # sum over columns for each of the 2 rows
array([[ 1,  3,  6],
       [ 4,  9, 15]])
dask.array.deg2rad(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Convert angles from degrees to radians.

Parameters:
x : array_like

Angles in degrees.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray

The corresponding angle in radians. This is a scalar if x is a scalar.

See also

rad2deg
Convert angles from radians to degrees.
unwrap
Remove large jumps in angle by wrapping.

Notes

New in version 1.3.0.

deg2rad(x) is x * pi / 180.

Examples

>>> np.deg2rad(180)  # doctest: +SKIP
3.1415926535897931
dask.array.degrees(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Convert angles from radians to degrees.

Parameters:
x : array_like

Input array in radians.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray of floats

The corresponding degree values; if out was supplied this is a reference to it. This is a scalar if x is a scalar.

See also

rad2deg
equivalent function

Examples

Convert a radian array to degrees

>>> rad = np.arange(12.)*np.pi/6  # doctest: +SKIP
>>> np.degrees(rad)  # doctest: +SKIP
array([   0.,   30.,   60.,   90.,  120.,  150.,  180.,  210.,  240.,
        270.,  300.,  330.])
>>> out = np.zeros((rad.shape))  # doctest: +SKIP
>>> r = degrees(rad, out)  # doctest: +SKIP
>>> np.all(r == out)  # doctest: +SKIP
True
dask.array.diag(v)

Extract a diagonal or construct a diagonal array.

This docstring was copied from numpy.diag.

Some inconsistencies with the Dask version may exist.

See the more detailed documentation for numpy.diagonal if you use this function to extract a diagonal and wish to write to the resulting array; whether it returns a copy or a view depends on what version of numpy you are using.

Parameters:
v : array_like

If v is a 2-D array, return a copy of its k-th diagonal. If v is a 1-D array, return a 2-D array with v on the k-th diagonal.

k : int, optional (Not supported in Dask)

Diagonal in question. The default is 0. Use k>0 for diagonals above the main diagonal, and k<0 for diagonals below the main diagonal.

Returns:
out : ndarray

The extracted diagonal or constructed diagonal array.

See also

diagonal
Return specified diagonals.
diagflat
Create a 2-D array with the flattened input as a diagonal.
trace
Sum along diagonals.
triu
Upper triangle of an array.
tril
Lower triangle of an array.

Examples

>>> x = np.arange(9).reshape((3,3))  # doctest: +SKIP
>>> x  # doctest: +SKIP
array([[0, 1, 2],
       [3, 4, 5],
       [6, 7, 8]])
>>> np.diag(x)  # doctest: +SKIP
array([0, 4, 8])
>>> np.diag(x, k=1)  # doctest: +SKIP
array([1, 5])
>>> np.diag(x, k=-1)  # doctest: +SKIP
array([3, 7])
>>> np.diag(np.diag(x))  # doctest: +SKIP
array([[0, 0, 0],
       [0, 4, 0],
       [0, 0, 8]])
dask.array.diagonal(a, offset=0, axis1=0, axis2=1)

Return specified diagonals.

This docstring was copied from numpy.diagonal.

Some inconsistencies with the Dask version may exist.

If a is 2-D, returns the diagonal of a with the given offset, i.e., the collection of elements of the form a[i, i+offset]. If a has more than two dimensions, then the axes specified by axis1 and axis2 are used to determine the 2-D sub-array whose diagonal is returned. The shape of the resulting array can be determined by removing axis1 and axis2 and appending an index to the right equal to the size of the resulting diagonals.

In versions of NumPy prior to 1.7, this function always returned a new, independent array containing a copy of the values in the diagonal.

In NumPy 1.7 and 1.8, it continues to return a copy of the diagonal, but depending on this fact is deprecated. Writing to the resulting array continues to work as it used to, but a FutureWarning is issued.

Starting in NumPy 1.9 it returns a read-only view on the original array. Attempting to write to the resulting array will produce an error.

In some future release, it will return a read/write view and writing to the returned array will alter your original array. The returned array will have the same type as the input array.

If you don’t write to the array returned by this function, then you can just ignore all of the above.

If you depend on the current behavior, then we suggest copying the returned array explicitly, i.e., use np.diagonal(a).copy() instead of just np.diagonal(a). This will work with both past and future versions of NumPy.

Parameters:
a : array_like

Array from which the diagonals are taken.

offset : int, optional

Offset of the diagonal from the main diagonal. Can be positive or negative. Defaults to main diagonal (0).

axis1 : int, optional

Axis to be used as the first axis of the 2-D sub-arrays from which the diagonals should be taken. Defaults to first axis (0).

axis2 : int, optional

Axis to be used as the second axis of the 2-D sub-arrays from which the diagonals should be taken. Defaults to second axis (1).

Returns:
array_of_diagonals : ndarray

If a is 2-D, then a 1-D array containing the diagonal and of the same type as a is returned unless a is a matrix, in which case a 1-D array rather than a (2-D) matrix is returned in order to maintain backward compatibility.

If a.ndim > 2, then the dimensions specified by axis1 and axis2 are removed, and a new axis inserted at the end corresponding to the diagonal.

Raises:
ValueError

If the dimension of a is less than 2.

See also

diag
MATLAB work-a-like for 1-D and 2-D arrays.
diagflat
Create diagonal arrays.
trace
Sum along diagonals.

Examples

>>> a = np.arange(4).reshape(2,2)  # doctest: +SKIP
>>> a  # doctest: +SKIP
array([[0, 1],
       [2, 3]])
>>> a.diagonal()  # doctest: +SKIP
array([0, 3])
>>> a.diagonal(1)  # doctest: +SKIP
array([1])

A 3-D example:

>>> a = np.arange(8).reshape(2,2,2); a  # doctest: +SKIP
array([[[0, 1],
        [2, 3]],
       [[4, 5],
        [6, 7]]])
>>> a.diagonal(0, # Main diagonals of two arrays created by skipping  # doctest: +SKIP
...            0, # across the outer(left)-most axis last and
...            1) # the "middle" (row) axis first.
array([[0, 6],
       [1, 7]])

The sub-arrays whose main diagonals we just obtained; note that each corresponds to fixing the right-most (column) axis, and that the diagonals are “packed” in rows.

>>> a[:,:,0] # main diagonal is [0 6]  # doctest: +SKIP
array([[0, 2],
       [4, 6]])
>>> a[:,:,1] # main diagonal is [1 7]  # doctest: +SKIP
array([[1, 3],
       [5, 7]])
dask.array.diff(a, n=1, axis=-1)

Calculate the n-th discrete difference along the given axis.

This docstring was copied from numpy.diff.

Some inconsistencies with the Dask version may exist.

The first difference is given by out[n] = a[n+1] - a[n] along the given axis, higher differences are calculated by using diff recursively.

Parameters:
a : array_like

Input array

n : int, optional

The number of times values are differenced. If zero, the input is returned as-is.

axis : int, optional

The axis along which the difference is taken, default is the last axis.

prepend, append : array_like, optional

Values to prepend or append to “a” along axis prior to performing the difference. Scalar values are expanded to arrays with length 1 in the direction of axis and the shape of the input array in along all other axes. Otherwise the dimension and shape must match “a” except along axis.

Returns:
diff : ndarray

The n-th differences. The shape of the output is the same as a except along axis where the dimension is smaller by n. The type of the output is the same as the type of the difference between any two elements of a. This is the same as the type of a in most cases. A notable exception is datetime64, which results in a timedelta64 output array.

See also

gradient, ediff1d, cumsum

Notes

Type is preserved for boolean arrays, so the result will contain False when consecutive elements are the same and True when they differ.

For unsigned integer arrays, the results will also be unsigned. This should not be surprising, as the result is consistent with calculating the difference directly:

>>> u8_arr = np.array([1, 0], dtype=np.uint8)  # doctest: +SKIP
>>> np.diff(u8_arr)  # doctest: +SKIP
array([255], dtype=uint8)
>>> u8_arr[1,...] - u8_arr[0,...]  # doctest: +SKIP
array(255, np.uint8)

If this is not desirable, then the array should be cast to a larger integer type first:

>>> i16_arr = u8_arr.astype(np.int16)  # doctest: +SKIP
>>> np.diff(i16_arr)  # doctest: +SKIP
array([-1], dtype=int16)

Examples

>>> x = np.array([1, 2, 4, 7, 0])  # doctest: +SKIP
>>> np.diff(x)  # doctest: +SKIP
array([ 1,  2,  3, -7])
>>> np.diff(x, n=2)  # doctest: +SKIP
array([  1,   1, -10])
>>> x = np.array([[1, 3, 6, 10], [0, 5, 6, 8]])  # doctest: +SKIP
>>> np.diff(x)  # doctest: +SKIP
array([[2, 3, 4],
       [5, 1, 2]])
>>> np.diff(x, axis=0)  # doctest: +SKIP
array([[-1,  2,  0, -2]])
>>> x = np.arange('1066-10-13', '1066-10-16', dtype=np.datetime64)  # doctest: +SKIP
>>> np.diff(x)  # doctest: +SKIP
array([1, 1], dtype='timedelta64[D]')
dask.array.digitize(a, bins, right=False)

Return the indices of the bins to which each value in input array belongs.

This docstring was copied from numpy.digitize.

Some inconsistencies with the Dask version may exist.

right order of bins returned index i satisfies
False increasing bins[i-1] <= x < bins[i]
True increasing bins[i-1] < x <= bins[i]
False decreasing bins[i-1] > x >= bins[i]
True decreasing bins[i-1] >= x > bins[i]

If values in x are beyond the bounds of bins, 0 or len(bins) is returned as appropriate.

Parameters:
x : array_like (Not supported in Dask)

Input array to be binned. Prior to NumPy 1.10.0, this array had to be 1-dimensional, but can now have any shape.

bins : array_like

Array of bins. It has to be 1-dimensional and monotonic.

right : bool, optional

Indicating whether the intervals include the right or the left bin edge. Default behavior is (right==False) indicating that the interval does not include the right edge. The left bin end is open in this case, i.e., bins[i-1] <= x < bins[i] is the default behavior for monotonically increasing bins.

Returns:
indices : ndarray of ints

Output array of indices, of same shape as x.

Raises:
ValueError

If bins is not monotonic.

TypeError

If the type of the input is complex.

See also

bincount, histogram, unique, searchsorted

Notes

If values in x are such that they fall outside the bin range, attempting to index bins with the indices that digitize returns will result in an IndexError.

New in version 1.10.0.

np.digitize is implemented in terms of np.searchsorted. This means that a binary search is used to bin the values, which scales much better for larger number of bins than the previous linear search. It also removes the requirement for the input array to be 1-dimensional.

For monotonically _increasing_ bins, the following are equivalent:

np.digitize(x, bins, right=True)
np.searchsorted(bins, x, side='left')

Note that as the order of the arguments are reversed, the side must be too. The searchsorted call is marginally faster, as it does not do any monotonicity checks. Perhaps more importantly, it supports all dtypes.

Examples

>>> x = np.array([0.2, 6.4, 3.0, 1.6])  # doctest: +SKIP
>>> bins = np.array([0.0, 1.0, 2.5, 4.0, 10.0])  # doctest: +SKIP
>>> inds = np.digitize(x, bins)  # doctest: +SKIP
>>> inds  # doctest: +SKIP
array([1, 4, 3, 2])
>>> for n in range(x.size):  # doctest: +SKIP
...   print(bins[inds[n]-1], "<=", x[n], "<", bins[inds[n]])
...
0.0 <= 0.2 < 1.0
4.0 <= 6.4 < 10.0
2.5 <= 3.0 < 4.0
1.0 <= 1.6 < 2.5
>>> x = np.array([1.2, 10.0, 12.4, 15.5, 20.])  # doctest: +SKIP
>>> bins = np.array([0, 5, 10, 15, 20])  # doctest: +SKIP
>>> np.digitize(x,bins,right=True)  # doctest: +SKIP
array([1, 2, 3, 4, 4])
>>> np.digitize(x,bins,right=False)  # doctest: +SKIP
array([1, 3, 3, 4, 5])
dask.array.dot(a, b, out=None)

This docstring was copied from numpy.dot.

Some inconsistencies with the Dask version may exist.

Dot product of two arrays. Specifically,

  • If both a and b are 1-D arrays, it is inner product of vectors (without complex conjugation).

  • If both a and b are 2-D arrays, it is matrix multiplication, but using matmul() or a @ b is preferred.

  • If either a or b is 0-D (scalar), it is equivalent to multiply() and using numpy.multiply(a, b) or a * b is preferred.

  • If a is an N-D array and b is a 1-D array, it is a sum product over the last axis of a and b.

  • If a is an N-D array and b is an M-D array (where M>=2), it is a sum product over the last axis of a and the second-to-last axis of b:

    dot(a, b)[i,j,k,m] = sum(a[i,j,:] * b[k,:,m])
    
Parameters:
a : array_like

First argument.

b : array_like

Second argument.

out : ndarray, optional

Output argument. This must have the exact kind that would be returned if it was not used. In particular, it must have the right type, must be C-contiguous, and its dtype must be the dtype that would be returned for dot(a,b). This is a performance feature. Therefore, if these conditions are not met, an exception is raised, instead of attempting to be flexible.

Returns:
output : ndarray

Returns the dot product of a and b. If a and b are both scalars or both 1-D arrays then a scalar is returned; otherwise an array is returned. If out is given, then it is returned.

Raises:
ValueError

If the last dimension of a is not the same size as the second-to-last dimension of b.

See also

vdot
Complex-conjugating dot product.
tensordot
Sum products over arbitrary axes.
einsum
Einstein summation convention.
matmul
‘@’ operator as method with out parameter.

Examples

>>> np.dot(3, 4)  # doctest: +SKIP
12

Neither argument is complex-conjugated:

>>> np.dot([2j, 3j], [2j, 3j])  # doctest: +SKIP
(-13+0j)

For 2-D arrays it is the matrix product:

>>> a = [[1, 0], [0, 1]]  # doctest: +SKIP
>>> b = [[4, 1], [2, 2]]  # doctest: +SKIP
>>> np.dot(a, b)  # doctest: +SKIP
array([[4, 1],
       [2, 2]])
>>> a = np.arange(3*4*5*6).reshape((3,4,5,6))  # doctest: +SKIP
>>> b = np.arange(3*4*5*6)[::-1].reshape((5,4,6,3))  # doctest: +SKIP
>>> np.dot(a, b)[2,3,2,1,2,2]  # doctest: +SKIP
499128
>>> sum(a[2,3,2,:] * b[1,2,:,2])  # doctest: +SKIP
499128
dask.array.dstack(tup, allow_unknown_chunksizes=False)

Stack arrays in sequence depth wise (along third axis).

This docstring was copied from numpy.dstack.

Some inconsistencies with the Dask version may exist.

This is equivalent to concatenation along the third axis after 2-D arrays of shape (M,N) have been reshaped to (M,N,1) and 1-D arrays of shape (N,) have been reshaped to (1,N,1). Rebuilds arrays divided by dsplit.

This function makes most sense for arrays with up to 3 dimensions. For instance, for pixel-data with a height (first axis), width (second axis), and r/g/b channels (third axis). The functions concatenate, stack and block provide more general stacking and concatenation operations.

Parameters:
tup : sequence of arrays

The arrays must have the same shape along all but the third axis. 1-D or 2-D arrays must have the same shape.

Returns:
stacked : ndarray

The array formed by stacking the given arrays, will be at least 3-D.

See also

stack
Join a sequence of arrays along a new axis.
vstack
Stack along first axis.
hstack
Stack along second axis.
concatenate
Join a sequence of arrays along an existing axis.
dsplit
Split array along third axis.

Examples

>>> a = np.array((1,2,3))  # doctest: +SKIP
>>> b = np.array((2,3,4))  # doctest: +SKIP
>>> np.dstack((a,b))  # doctest: +SKIP
array([[[1, 2],
        [2, 3],
        [3, 4]]])
>>> a = np.array([[1],[2],[3]])  # doctest: +SKIP
>>> b = np.array([[2],[3],[4]])  # doctest: +SKIP
>>> np.dstack((a,b))  # doctest: +SKIP
array([[[1, 2]],
       [[2, 3]],
       [[3, 4]]])
dask.array.ediff1d(ary, to_end=None, to_begin=None)

The differences between consecutive elements of an array.

This docstring was copied from numpy.ediff1d.

Some inconsistencies with the Dask version may exist.

Parameters:
ary : array_like

If necessary, will be flattened before the differences are taken.

to_end : array_like, optional

Number(s) to append at the end of the returned differences.

to_begin : array_like, optional

Number(s) to prepend at the beginning of the returned differences.

Returns:
ediff1d : ndarray

The differences. Loosely, this is ary.flat[1:] - ary.flat[:-1].

See also

diff, gradient

Notes

When applied to masked arrays, this function drops the mask information if the to_begin and/or to_end parameters are used.

Examples

>>> x = np.array([1, 2, 4, 7, 0])  # doctest: +SKIP
>>> np.ediff1d(x)  # doctest: +SKIP
array([ 1,  2,  3, -7])
>>> np.ediff1d(x, to_begin=-99, to_end=np.array([88, 99]))  # doctest: +SKIP
array([-99,   1,   2,   3,  -7,  88,  99])

The returned array is always 1D.

>>> y = [[1, 2, 4], [1, 6, 24]]  # doctest: +SKIP
>>> np.ediff1d(y)  # doctest: +SKIP
array([ 1,  2, -3,  5, 18])
dask.array.empty(*args, **kwargs)

Blocked variant of empty

Follows the signature of empty exactly except that it also requires a keyword argument chunks=(…)

Original signature follows below. empty(shape, dtype=float, order=’C’)

Return a new array of given shape and type, without initializing entries.

Parameters:
shape : int or tuple of int

Shape of the empty array, e.g., (2, 3) or 2.

dtype : data-type, optional

Desired output data-type for the array, e.g, numpy.int8. Default is numpy.float64.

order : {‘C’, ‘F’}, optional, default: ‘C’

Whether to store multi-dimensional data in row-major (C-style) or column-major (Fortran-style) order in memory.

Returns:
out : ndarray

Array of uninitialized (arbitrary) data of the given shape, dtype, and order. Object arrays will be initialized to None.

See also

empty_like
Return an empty array with shape and type of input.
ones
Return a new array setting values to one.
zeros
Return a new array setting values to zero.
full
Return a new array of given shape filled with value.

Notes

empty, unlike zeros, does not set the array values to zero, and may therefore be marginally faster. On the other hand, it requires the user to manually set all the values in the array, and should be used with caution.

Examples

>>> np.empty([2, 2])
array([[ -9.74499359e+001,   6.69583040e-309],
       [  2.13182611e-314,   3.06959433e-309]])         #random
>>> np.empty([2, 2], dtype=int)
array([[-1073741821, -1067949133],
       [  496041986,    19249760]])                     #random
dask.array.empty_like(a, dtype=None, chunks=None)

Return a new array with the same shape and type as a given array.

Parameters:
a : array_like

The shape and data-type of a define these same attributes of the returned array.

dtype : data-type, optional

Overrides the data type of the result.

chunks : sequence of ints

The number of samples on each block. Note that the last block will have fewer samples if len(array) % chunks != 0.

Returns:
out : ndarray

Array of uninitialized (arbitrary) data with the same shape and type as a.

See also

ones_like
Return an array of ones with shape and type of input.
zeros_like
Return an array of zeros with shape and type of input.
empty
Return a new uninitialized array.
ones
Return a new array setting values to one.
zeros
Return a new array setting values to zero.

Notes

This function does not initialize the returned array; to do that use zeros_like or ones_like instead. It may be marginally faster than the functions that do set the array values.

dask.array.einsum(subscripts, *operands, out=None, dtype=None, order='K', casting='safe', optimize=False)

This docstring was copied from numpy.einsum.

Some inconsistencies with the Dask version may exist.

Evaluates the Einstein summation convention on the operands.

Using the Einstein summation convention, many common multi-dimensional, linear algebraic array operations can be represented in a simple fashion. In implicit mode einsum computes these values.

In explicit mode, einsum provides further flexibility to compute other array operations that might not be considered classical Einstein summation operations, by disabling, or forcing summation over specified subscript labels.

See the notes and examples for clarification.

Parameters:
subscripts : str

Specifies the subscripts for summation as comma separated list of subscript labels. An implicit (classical Einstein summation) calculation is performed unless the explicit indicator ‘->’ is included as well as subscript labels of the precise output form.

operands : list of array_like

These are the arrays for the operation.

out : ndarray, optional

If provided, the calculation is done into this array.

dtype : {data-type, None}, optional

If provided, forces the calculation to use the data type specified. Note that you may have to also give a more liberal casting parameter to allow the conversions. Default is None.

order : {‘C’, ‘F’, ‘A’, ‘K’}, optional

Controls the memory layout of the output. ‘C’ means it should be C contiguous. ‘F’ means it should be Fortran contiguous, ‘A’ means it should be ‘F’ if the inputs are all ‘F’, ‘C’ otherwise. ‘K’ means it should be as close to the layout as the inputs as is possible, including arbitrarily permuted axes. Default is ‘K’.

casting : {‘no’, ‘equiv’, ‘safe’, ‘same_kind’, ‘unsafe’}, optional

Controls what kind of data casting may occur. Setting this to ‘unsafe’ is not recommended, as it can adversely affect accumulations.

  • ‘no’ means the data types should not be cast at all.
  • ‘equiv’ means only byte-order changes are allowed.
  • ‘safe’ means only casts which can preserve values are allowed.
  • ‘same_kind’ means only safe casts or casts within a kind, like float64 to float32, are allowed.
  • ‘unsafe’ means any data conversions may be done.

Default is ‘safe’.

optimize : {False, True, ‘greedy’, ‘optimal’}, optional

Controls if intermediate optimization should occur. No optimization will occur if False and True will default to the ‘greedy’ algorithm. Also accepts an explicit contraction list from the np.einsum_path function. See np.einsum_path for more details. Defaults to False.

Returns:
output : ndarray

The calculation based on the Einstein summation convention.

See also

einsum_path, dot, inner, outer, tensordot, linalg.multi_dot

Notes

New in version 1.6.0.

The Einstein summation convention can be used to compute many multi-dimensional, linear algebraic array operations. einsum provides a succinct way of representing these.

A non-exhaustive list of these operations, which can be computed by einsum, is shown below along with examples:

The subscripts string is a comma-separated list of subscript labels, where each label refers to a dimension of the corresponding operand. Whenever a label is repeated it is summed, so np.einsum('i,i', a, b) is equivalent to np.inner(a,b). If a label appears only once, it is not summed, so np.einsum('i', a) produces a view of a with no changes. A further example np.einsum('ij,jk', a, b) describes traditional matrix multiplication and is equivalent to np.matmul(a,b). Repeated subscript labels in one operand take the diagonal. For example, np.einsum('ii', a) is equivalent to np.trace(a).

In implicit mode, the chosen subscripts are important since the axes of the output are reordered alphabetically. This means that np.einsum('ij', a) doesn’t affect a 2D array, while np.einsum('ji', a) takes its transpose. Additionally, np.einsum('ij,jk', a, b) returns a matrix multiplication, while, np.einsum('ij,jh', a, b) returns the transpose of the multiplication since subscript ‘h’ precedes subscript ‘i’.

In explicit mode the output can be directly controlled by specifying output subscript labels. This requires the identifier ‘->’ as well as the list of output subscript labels. This feature increases the flexibility of the function since summing can be disabled or forced when required. The call np.einsum('i->', a) is like np.sum(a, axis=-1), and np.einsum('ii->i', a) is like np.diag(a). The difference is that einsum does not allow broadcasting by default. Additionally np.einsum('ij,jh->ih', a, b) directly specifies the order of the output subscript labels and therefore returns matrix multiplication, unlike the example above in implicit mode.

To enable and control broadcasting, use an ellipsis. Default NumPy-style broadcasting is done by adding an ellipsis to the left of each term, like np.einsum('...ii->...i', a). To take the trace along the first and last axes, you can do np.einsum('i...i', a), or to do a matrix-matrix product with the left-most indices instead of rightmost, one can do np.einsum('ij...,jk...->ik...', a, b).

When there is only one operand, no axes are summed, and no output parameter is provided, a view into the operand is returned instead of a new array. Thus, taking the diagonal as np.einsum('ii->i', a) produces a view (changed in version 1.10.0).

einsum also provides an alternative way to provide the subscripts and operands as einsum(op0, sublist0, op1, sublist1, ..., [sublistout]). If the output shape is not provided in this format einsum will be calculated in implicit mode, otherwise it will be performed explicitly. The examples below have corresponding einsum calls with the two parameter methods.

New in version 1.10.0.

Views returned from einsum are now writeable whenever the input array is writeable. For example, np.einsum('ijk...->kji...', a) will now have the same effect as np.swapaxes(a, 0, 2) and np.einsum('ii->i', a) will return a writeable view of the diagonal of a 2D array.

New in version 1.12.0.

Added the optimize argument which will optimize the contraction order of an einsum expression. For a contraction with three or more operands this can greatly increase the computational efficiency at the cost of a larger memory footprint during computation.

Typically a ‘greedy’ algorithm is applied which empirical tests have shown returns the optimal path in the majority of cases. In some cases ‘optimal’ will return the superlative path through a more expensive, exhaustive search. For iterative calculations it may be advisable to calculate the optimal path once and reuse that path by supplying it as an argument. An example is given below.

See numpy.einsum_path() for more details.

Examples

>>> a = np.arange(25).reshape(5,5)  # doctest: +SKIP
>>> b = np.arange(5)  # doctest: +SKIP
>>> c = np.arange(6).reshape(2,3)  # doctest: +SKIP

Trace of a matrix:

>>> np.einsum('ii', a)  # doctest: +SKIP
60
>>> np.einsum(a, [0,0])  # doctest: +SKIP
60
>>> np.trace(a)  # doctest: +SKIP
60

Extract the diagonal (requires explicit form):

>>> np.einsum('ii->i', a)  # doctest: +SKIP
array([ 0,  6, 12, 18, 24])
>>> np.einsum(a, [0,0], [0])  # doctest: +SKIP
array([ 0,  6, 12, 18, 24])
>>> np.diag(a)  # doctest: +SKIP
array([ 0,  6, 12, 18, 24])

Sum over an axis (requires explicit form):

>>> np.einsum('ij->i', a)  # doctest: +SKIP
array([ 10,  35,  60,  85, 110])
>>> np.einsum(a, [0,1], [0])  # doctest: +SKIP
array([ 10,  35,  60,  85, 110])
>>> np.sum(a, axis=1)  # doctest: +SKIP
array([ 10,  35,  60,  85, 110])

For higher dimensional arrays summing a single axis can be done with ellipsis:

>>> np.einsum('...j->...', a)  # doctest: +SKIP
array([ 10,  35,  60,  85, 110])
>>> np.einsum(a, [Ellipsis,1], [Ellipsis])  # doctest: +SKIP
array([ 10,  35,  60,  85, 110])

Compute a matrix transpose, or reorder any number of axes:

>>> np.einsum('ji', c)  # doctest: +SKIP
array([[0, 3],
       [1, 4],
       [2, 5]])
>>> np.einsum('ij->ji', c)  # doctest: +SKIP
array([[0, 3],
       [1, 4],
       [2, 5]])
>>> np.einsum(c, [1,0])  # doctest: +SKIP
array([[0, 3],
       [1, 4],
       [2, 5]])
>>> np.transpose(c)  # doctest: +SKIP
array([[0, 3],
       [1, 4],
       [2, 5]])

Vector inner products:

>>> np.einsum('i,i', b, b)  # doctest: +SKIP
30
>>> np.einsum(b, [0], b, [0])  # doctest: +SKIP
30
>>> np.inner(b,b)  # doctest: +SKIP
30

Matrix vector multiplication:

>>> np.einsum('ij,j', a, b)  # doctest: +SKIP
array([ 30,  80, 130, 180, 230])
>>> np.einsum(a, [0,1], b, [1])  # doctest: +SKIP
array([ 30,  80, 130, 180, 230])
>>> np.dot(a, b)  # doctest: +SKIP
array([ 30,  80, 130, 180, 230])
>>> np.einsum('...j,j', a, b)  # doctest: +SKIP
array([ 30,  80, 130, 180, 230])

Broadcasting and scalar multiplication:

>>> np.einsum('..., ...', 3, c)  # doctest: +SKIP
array([[ 0,  3,  6],
       [ 9, 12, 15]])
>>> np.einsum(',ij', 3, c)  # doctest: +SKIP
array([[ 0,  3,  6],
       [ 9, 12, 15]])
>>> np.einsum(3, [Ellipsis], c, [Ellipsis])  # doctest: +SKIP
array([[ 0,  3,  6],
       [ 9, 12, 15]])
>>> np.multiply(3, c)  # doctest: +SKIP
array([[ 0,  3,  6],
       [ 9, 12, 15]])

Vector outer product:

>>> np.einsum('i,j', np.arange(2)+1, b)  # doctest: +SKIP
array([[0, 1, 2, 3, 4],
       [0, 2, 4, 6, 8]])
>>> np.einsum(np.arange(2)+1, [0], b, [1])  # doctest: +SKIP
array([[0, 1, 2, 3, 4],
       [0, 2, 4, 6, 8]])
>>> np.outer(np.arange(2)+1, b)  # doctest: +SKIP
array([[0, 1, 2, 3, 4],
       [0, 2, 4, 6, 8]])

Tensor contraction:

>>> a = np.arange(60.).reshape(3,4,5)  # doctest: +SKIP
>>> b = np.arange(24.).reshape(4,3,2)  # doctest: +SKIP
>>> np.einsum('ijk,jil->kl', a, b)  # doctest: +SKIP
array([[ 4400.,  4730.],
       [ 4532.,  4874.],
       [ 4664.,  5018.],
       [ 4796.,  5162.],
       [ 4928.,  5306.]])
>>> np.einsum(a, [0,1,2], b, [1,0,3], [2,3])  # doctest: +SKIP
array([[ 4400.,  4730.],
       [ 4532.,  4874.],
       [ 4664.,  5018.],
       [ 4796.,  5162.],
       [ 4928.,  5306.]])
>>> np.tensordot(a,b, axes=([1,0],[0,1]))  # doctest: +SKIP
array([[ 4400.,  4730.],
       [ 4532.,  4874.],
       [ 4664.,  5018.],
       [ 4796.,  5162.],
       [ 4928.,  5306.]])

Writeable returned arrays (since version 1.10.0):

>>> a = np.zeros((3, 3))  # doctest: +SKIP
>>> np.einsum('ii->i', a)[:] = 1  # doctest: +SKIP
>>> a  # doctest: +SKIP
array([[ 1.,  0.,  0.],
       [ 0.,  1.,  0.],
       [ 0.,  0.,  1.]])

Example of ellipsis use:

>>> a = np.arange(6).reshape((3,2))  # doctest: +SKIP
>>> b = np.arange(12).reshape((4,3))  # doctest: +SKIP
>>> np.einsum('ki,jk->ij', a, b)  # doctest: +SKIP
array([[10, 28, 46, 64],
       [13, 40, 67, 94]])
>>> np.einsum('ki,...k->i...', a, b)  # doctest: +SKIP
array([[10, 28, 46, 64],
       [13, 40, 67, 94]])
>>> np.einsum('k...,jk', a, b)  # doctest: +SKIP
array([[10, 28, 46, 64],
       [13, 40, 67, 94]])

Chained array operations. For more complicated contractions, speed ups might be achieved by repeatedly computing a ‘greedy’ path or pre-computing the ‘optimal’ path and repeatedly applying it, using an einsum_path insertion (since version 1.12.0). Performance improvements can be particularly significant with larger arrays:

>>> a = np.ones(64).reshape(2,4,8)  # doctest: +SKIP
# Basic `einsum`: ~1520ms  (benchmarked on 3.1GHz Intel i5.)
>>> for iteration in range(500):  # doctest: +SKIP
...     np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a)
# Sub-optimal `einsum` (due to repeated path calculation time): ~330ms
>>> for iteration in range(500):  # doctest: +SKIP
...     np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize='optimal')
# Greedy `einsum` (faster optimal path approximation): ~160ms
>>> for iteration in range(500):  # doctest: +SKIP
...     np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize='greedy')
# Optimal `einsum` (best usage pattern in some use cases): ~110ms
>>> path = np.einsum_path('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize='optimal')[0]  # doctest: +SKIP
>>> for iteration in range(500):  # doctest: +SKIP
...     np.einsum('ijk,ilm,njm,nlk,abc->',a,a,a,a,a, optimize=path)
dask.array.exp(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Calculate the exponential of all elements in the input array.

Parameters:
x : array_like

Input values.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Output array, element-wise exponential of x. This is a scalar if x is a scalar.

See also

expm1
Calculate exp(x) - 1 for all elements in the array.
exp2
Calculate 2**x for all elements in the array.

Notes

The irrational number e is also known as Euler’s number. It is approximately 2.718281, and is the base of the natural logarithm, ln (this means that, if \(x = \ln y = \log_e y\), then \(e^x = y\). For real input, exp(x) is always positive.

For complex arguments, x = a + ib, we can write \(e^x = e^a e^{ib}\). The first term, \(e^a\), is already known (it is the real argument, described above). The second term, \(e^{ib}\), is \(\cos b + i \sin b\), a function with magnitude 1 and a periodic phase.

References

[1]Wikipedia, “Exponential function”, https://en.wikipedia.org/wiki/Exponential_function
[2]M. Abramovitz and I. A. Stegun, “Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables,” Dover, 1964, p. 69, http://www.math.sfu.ca/~cbm/aands/page_69.htm

Examples

Plot the magnitude and phase of exp(x) in the complex plane:

>>> import matplotlib.pyplot as plt  # doctest: +SKIP
>>> x = np.linspace(-2*np.pi, 2*np.pi, 100)  # doctest: +SKIP
>>> xx = x + 1j * x[:, np.newaxis] # a + ib over complex plane  # doctest: +SKIP
>>> out = np.exp(xx)  # doctest: +SKIP
>>> plt.subplot(121)  # doctest: +SKIP
>>> plt.imshow(np.abs(out),  # doctest: +SKIP
...            extent=[-2*np.pi, 2*np.pi, -2*np.pi, 2*np.pi], cmap='gray')
>>> plt.title('Magnitude of exp(x)')  # doctest: +SKIP
>>> plt.subplot(122)  # doctest: +SKIP
>>> plt.imshow(np.angle(out),  # doctest: +SKIP
...            extent=[-2*np.pi, 2*np.pi, -2*np.pi, 2*np.pi], cmap='hsv')
>>> plt.title('Phase (angle) of exp(x)')  # doctest: +SKIP
>>> plt.show()  # doctest: +SKIP
dask.array.expm1(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Calculate exp(x) - 1 for all elements in the array.

Parameters:
x : array_like

Input values.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Element-wise exponential minus one: out = exp(x) - 1. This is a scalar if x is a scalar.

See also

log1p
log(1 + x), the inverse of expm1.

Notes

This function provides greater precision than exp(x) - 1 for small values of x.

Examples

The true value of exp(1e-10) - 1 is 1.00000000005e-10 to about 32 significant digits. This example shows the superiority of expm1 in this case.

>>> np.expm1(1e-10)  # doctest: +SKIP
1.00000000005e-10
>>> np.exp(1e-10) - 1  # doctest: +SKIP
1.000000082740371e-10
dask.array.eye(N, chunks='auto', M=None, k=0, dtype=<class 'float'>)

Return a 2-D Array with ones on the diagonal and zeros elsewhere.

Parameters:
N : int

Number of rows in the output.

chunks : int, str

How to chunk the array. Must be one of the following forms: - A blocksize like 1000. - A size in bytes, like “100 MiB” which will choose a uniform

block-like shape

  • The word “auto” which acts like the above, but uses a configuration value array.chunk-size for the chunk size
M : int, optional

Number of columns in the output. If None, defaults to N.

k : int, optional

Index of the diagonal: 0 (the default) refers to the main diagonal, a positive value refers to an upper diagonal, and a negative value to a lower diagonal.

dtype : data-type, optional

Data-type of the returned array.

Returns:
I : Array of shape (N,M)

An array where all elements are equal to zero, except for the k-th diagonal, whose values are equal to one.

dask.array.fabs(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute the absolute values element-wise.

This function returns the absolute values (positive magnitude) of the data in x. Complex values are not handled, use absolute to find the absolute values of complex data.

Parameters:
x : array_like

The array of numbers for which the absolute values are required. If x is a scalar, the result y will also be a scalar.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or scalar

The absolute values of x, the returned values are always floats. This is a scalar if x is a scalar.

See also

absolute
Absolute values including complex types.

Examples

>>> np.fabs(-1)  # doctest: +SKIP
1.0
>>> np.fabs([-1.2, 1.2])  # doctest: +SKIP
array([ 1.2,  1.2])
dask.array.fix(*args, **kwargs)

Round to nearest integer towards zero.

Round an array of floats element-wise to nearest integer towards zero. The rounded values are returned as floats.

Parameters:
x : array_like

An array of floats to be rounded

y : ndarray, optional

Output array

Returns:
out : ndarray of floats

The array of rounded numbers

See also

trunc, floor, ceil

around
Round to given number of decimals

Examples

>>> np.fix(3.14)  # doctest: +SKIP
3.0
>>> np.fix(3)  # doctest: +SKIP
3.0
>>> np.fix([2.1, 2.9, -2.1, -2.9])  # doctest: +SKIP
array([ 2.,  2., -2., -2.])
dask.array.flatnonzero(a)

Return indices that are non-zero in the flattened version of a.

This docstring was copied from numpy.flatnonzero.

Some inconsistencies with the Dask version may exist.

This is equivalent to np.nonzero(np.ravel(a))[0].

Parameters:
a : array_like

Input data.

Returns:
res : ndarray

Output array, containing the indices of the elements of a.ravel() that are non-zero.

See also

nonzero
Return the indices of the non-zero elements of the input array.
ravel
Return a 1-D array containing the elements of the input array.

Examples

>>> x = np.arange(-2, 3)  # doctest: +SKIP
>>> x  # doctest: +SKIP
array([-2, -1,  0,  1,  2])
>>> np.flatnonzero(x)  # doctest: +SKIP
array([0, 1, 3, 4])

Use the indices of the non-zero elements as an index array to extract these elements:

>>> x.ravel()[np.flatnonzero(x)]  # doctest: +SKIP
array([-2, -1,  1,  2])
dask.array.flip(m, axis)

Reverse element order along axis.

Parameters:
axis : int

Axis to reverse element order of.

Returns:
reversed array : ndarray
dask.array.flipud(m)

Flip array in the up/down direction.

This docstring was copied from numpy.flipud.

Some inconsistencies with the Dask version may exist.

Flip the entries in each column in the up/down direction. Rows are preserved, but appear in a different order than before.

Parameters:
m : array_like

Input array.

Returns:
out : array_like

A view of m with the rows reversed. Since a view is returned, this operation is \(\mathcal O(1)\).

See also

fliplr
Flip array in the left/right direction.
rot90
Rotate array counterclockwise.

Notes

Equivalent to m[::-1,...]. Does not require the array to be two-dimensional.

Examples

>>> A = np.diag([1.0, 2, 3])  # doctest: +SKIP
>>> A  # doctest: +SKIP
array([[ 1.,  0.,  0.],
       [ 0.,  2.,  0.],
       [ 0.,  0.,  3.]])
>>> np.flipud(A)  # doctest: +SKIP
array([[ 0.,  0.,  3.],
       [ 0.,  2.,  0.],
       [ 1.,  0.,  0.]])
>>> A = np.random.randn(2,3,5)  # doctest: +SKIP
>>> np.all(np.flipud(A) == A[::-1,...])  # doctest: +SKIP
True
>>> np.flipud([1,2])  # doctest: +SKIP
array([2, 1])
dask.array.fliplr(m)

Flip array in the left/right direction.

This docstring was copied from numpy.fliplr.

Some inconsistencies with the Dask version may exist.

Flip the entries in each row in the left/right direction. Columns are preserved, but appear in a different order than before.

Parameters:
m : array_like

Input array, must be at least 2-D.

Returns:
f : ndarray

A view of m with the columns reversed. Since a view is returned, this operation is \(\mathcal O(1)\).

See also

flipud
Flip array in the up/down direction.
rot90
Rotate array counterclockwise.

Notes

Equivalent to m[:,::-1]. Requires the array to be at least 2-D.

Examples

>>> A = np.diag([1.,2.,3.])  # doctest: +SKIP
>>> A  # doctest: +SKIP
array([[ 1.,  0.,  0.],
       [ 0.,  2.,  0.],
       [ 0.,  0.,  3.]])
>>> np.fliplr(A)  # doctest: +SKIP
array([[ 0.,  0.,  1.],
       [ 0.,  2.,  0.],
       [ 3.,  0.,  0.]])
>>> A = np.random.randn(2,3,5)  # doctest: +SKIP
>>> np.all(np.fliplr(A) == A[:,::-1,...])  # doctest: +SKIP
True
dask.array.floor(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Return the floor of the input, element-wise.

The floor of the scalar x is the largest integer i, such that i <= x. It is often denoted as \(\lfloor x \rfloor\).

Parameters:
x : array_like

Input data.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or scalar

The floor of each element in x. This is a scalar if x is a scalar.

See also

ceil, trunc, rint

Notes

Some spreadsheet programs calculate the “floor-towards-zero”, in other words floor(-2.5) == -2. NumPy instead uses the definition of floor where floor(-2.5) == -3.

Examples

>>> a = np.array([-1.7, -1.5, -0.2, 0.2, 1.5, 1.7, 2.0])  # doctest: +SKIP
>>> np.floor(a)  # doctest: +SKIP
array([-2., -2., -1.,  0.,  1.,  1.,  2.])
dask.array.fmax(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Element-wise maximum of array elements.

Compare two arrays and returns a new array containing the element-wise maxima. If one of the elements being compared is a NaN, then the non-nan element is returned. If both elements are NaNs then the first is returned. The latter distinction is important for complex NaNs, which are defined as at least one of the real or imaginary parts being a NaN. The net effect is that NaNs are ignored when possible.

Parameters:
x1, x2 : array_like

The arrays holding the elements to be compared. They must have the same shape.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or scalar

The maximum of x1 and x2, element-wise. This is a scalar if both x1 and x2 are scalars.

See also

fmin
Element-wise minimum of two arrays, ignores NaNs.
maximum
Element-wise maximum of two arrays, propagates NaNs.
amax
The maximum value of an array along a given axis, propagates NaNs.
nanmax
The maximum value of an array along a given axis, ignores NaNs.

minimum, amin, nanmin

Notes

New in version 1.3.0.

The fmax is equivalent to np.where(x1 >= x2, x1, x2) when neither x1 nor x2 are NaNs, but it is faster and does proper broadcasting.

Examples

>>> np.fmax([2, 3, 4], [1, 5, 2])  # doctest: +SKIP
array([ 2.,  5.,  4.])
>>> np.fmax(np.eye(2), [0.5, 2])  # doctest: +SKIP
array([[ 1. ,  2. ],
       [ 0.5,  2. ]])
>>> np.fmax([np.nan, 0, np.nan],[0, np.nan, np.nan])  # doctest: +SKIP
array([  0.,   0.,  NaN])
dask.array.fmin(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Element-wise minimum of array elements.

Compare two arrays and returns a new array containing the element-wise minima. If one of the elements being compared is a NaN, then the non-nan element is returned. If both elements are NaNs then the first is returned. The latter distinction is important for complex NaNs, which are defined as at least one of the real or imaginary parts being a NaN. The net effect is that NaNs are ignored when possible.

Parameters:
x1, x2 : array_like

The arrays holding the elements to be compared. They must have the same shape.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or scalar

The minimum of x1 and x2, element-wise. This is a scalar if both x1 and x2 are scalars.

See also

fmax
Element-wise maximum of two arrays, ignores NaNs.
minimum
Element-wise minimum of two arrays, propagates NaNs.
amin
The minimum value of an array along a given axis, propagates NaNs.
nanmin
The minimum value of an array along a given axis, ignores NaNs.

maximum, amax, nanmax

Notes

New in version 1.3.0.

The fmin is equivalent to np.where(x1 <= x2, x1, x2) when neither x1 nor x2 are NaNs, but it is faster and does proper broadcasting.

Examples

>>> np.fmin([2, 3, 4], [1, 5, 2])  # doctest: +SKIP
array([1, 3, 2])
>>> np.fmin(np.eye(2), [0.5, 2])  # doctest: +SKIP
array([[ 0.5,  0. ],
       [ 0. ,  1. ]])
>>> np.fmin([np.nan, 0, np.nan],[0, np.nan, np.nan])  # doctest: +SKIP
array([  0.,   0.,  NaN])
dask.array.fmod(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Return the element-wise remainder of division.

This is the NumPy implementation of the C library function fmod, the remainder has the same sign as the dividend x1. It is equivalent to the Matlab(TM) rem function and should not be confused with the Python modulus operator x1 % x2.

Parameters:
x1 : array_like

Dividend.

x2 : array_like

Divisor.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : array_like

The remainder of the division of x1 by x2. This is a scalar if both x1 and x2 are scalars.

See also

remainder
Equivalent to the Python % operator.

divide

Notes

The result of the modulo operation for negative dividend and divisors is bound by conventions. For fmod, the sign of result is the sign of the dividend, while for remainder the sign of the result is the sign of the divisor. The fmod function is equivalent to the Matlab(TM) rem function.

Examples

>>> np.fmod([-3, -2, -1, 1, 2, 3], 2)  # doctest: +SKIP
array([-1,  0, -1,  1,  0,  1])
>>> np.remainder([-3, -2, -1, 1, 2, 3], 2)  # doctest: +SKIP
array([1, 0, 1, 1, 0, 1])
>>> np.fmod([5, 3], [2, 2.])  # doctest: +SKIP
array([ 1.,  1.])
>>> a = np.arange(-3, 3).reshape(3, 2)  # doctest: +SKIP
>>> a  # doctest: +SKIP
array([[-3, -2],
       [-1,  0],
       [ 1,  2]])
>>> np.fmod(a, [2,2])  # doctest: +SKIP
array([[-1,  0],
       [-1,  0],
       [ 1,  0]])
dask.array.frexp(x, [out1, out2, ]/, [out=(None, None), ]*, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Decompose the elements of x into mantissa and twos exponent.

Returns (mantissa, exponent), where x = mantissa * 2**exponent`. The mantissa is lies in the open interval(-1, 1), while the twos exponent is a signed integer.

Parameters:
x : array_like

Array of numbers to be decomposed.

out1 : ndarray, optional

Output array for the mantissa. Must have the same shape as x.

out2 : ndarray, optional

Output array for the exponent. Must have the same shape as x.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
mantissa : ndarray

Floating values between -1 and 1. This is a scalar if x is a scalar.

exponent : ndarray

Integer exponents of 2. This is a scalar if x is a scalar.

See also

ldexp
Compute y = x1 * 2**x2, the inverse of frexp.

Notes

Complex dtypes are not supported, they will raise a TypeError.

Examples

>>> x = np.arange(9)  # doctest: +SKIP
>>> y1, y2 = np.frexp(x)  # doctest: +SKIP
>>> y1  # doctest: +SKIP
array([ 0.   ,  0.5  ,  0.5  ,  0.75 ,  0.5  ,  0.625,  0.75 ,  0.875,
        0.5  ])
>>> y2  # doctest: +SKIP
array([0, 1, 2, 2, 3, 3, 3, 3, 4])
>>> y1 * 2**y2  # doctest: +SKIP
array([ 0.,  1.,  2.,  3.,  4.,  5.,  6.,  7.,  8.])
dask.array.fromfunction(func, chunks='auto', shape=None, dtype=None, **kwargs)

Construct an array by executing a function over each coordinate.

This docstring was copied from numpy.fromfunction.

Some inconsistencies with the Dask version may exist.

The resulting array therefore has a value fn(x, y, z) at coordinate (x, y, z).

Parameters:
function : callable (Not supported in Dask)

The function is called with N parameters, where N is the rank of shape. Each parameter represents the coordinates of the array varying along a specific axis. For example, if shape were (2, 2), then the parameters would be array([[0, 0], [1, 1]]) and array([[0, 1], [0, 1]])

shape : (N,) tuple of ints

Shape of the output array, which also determines the shape of the coordinate arrays passed to function.

dtype : data-type, optional

Data-type of the coordinate arrays passed to function. By default, dtype is float.

Returns:
fromfunction : any

The result of the call to function is passed back directly. Therefore the shape of fromfunction is completely determined by function. If function returns a scalar value, the shape of fromfunction would not match the shape parameter.

See also

indices, meshgrid

Notes

Keywords other than dtype are passed to function.

Examples

>>> np.fromfunction(lambda i, j: i == j, (3, 3), dtype=int)  # doctest: +SKIP
array([[ True, False, False],
       [False,  True, False],
       [False, False,  True]])
>>> np.fromfunction(lambda i, j: i + j, (3, 3), dtype=int)  # doctest: +SKIP
array([[0, 1, 2],
       [1, 2, 3],
       [2, 3, 4]])
dask.array.frompyfunc(func, nin, nout)

This docstring was copied from numpy.frompyfunc.

Some inconsistencies with the Dask version may exist.

Takes an arbitrary Python function and returns a NumPy ufunc.

Can be used, for example, to add broadcasting to a built-in Python function (see Examples section).

Parameters:
func : Python function object

An arbitrary Python function.

nin : int

The number of input arguments.

nout : int

The number of objects returned by func.

Returns:
out : ufunc

Returns a NumPy universal function (ufunc) object.

See also

vectorize
evaluates pyfunc over input arrays using broadcasting rules of numpy

Notes

The returned ufunc always returns PyObject arrays.

Examples

Use frompyfunc to add broadcasting to the Python function oct:

>>> oct_array = np.frompyfunc(oct, 1, 1)  # doctest: +SKIP
>>> oct_array(np.array((10, 30, 100)))  # doctest: +SKIP
array([012, 036, 0144], dtype=object)
>>> np.array((oct(10), oct(30), oct(100))) # for comparison  # doctest: +SKIP
array(['012', '036', '0144'],
      dtype='|S4')
dask.array.full(*args, **kwargs)

Blocked variant of full

Follows the signature of full exactly except that it also requires a keyword argument chunks=(…)

Original signature follows below.

Return a new array of given shape and type, filled with fill_value.

Parameters:
shape : int or sequence of ints

Shape of the new array, e.g., (2, 3) or 2.

fill_value : scalar

Fill value.

dtype : data-type, optional
The desired data-type for the array The default, None, means

np.array(fill_value).dtype.

order : {‘C’, ‘F’}, optional

Whether to store multidimensional data in C- or Fortran-contiguous (row- or column-wise) order in memory.

Returns:
out : ndarray

Array of fill_value with the given shape, dtype, and order.

See also

full_like
Return a new array with shape of input filled with value.
empty
Return a new uninitialized array.
ones
Return a new array setting values to one.
zeros
Return a new array setting values to zero.

Examples

>>> np.full((2, 2), np.inf)
array([[ inf,  inf],
       [ inf,  inf]])
>>> np.full((2, 2), 10)
array([[10, 10],
       [10, 10]])
dask.array.full_like(a, fill_value, dtype=None, chunks=None)

Return a full array with the same shape and type as a given array.

Parameters:
a : array_like

The shape and data-type of a define these same attributes of the returned array.

fill_value : scalar

Fill value.

dtype : data-type, optional

Overrides the data type of the result.

chunks : sequence of ints

The number of samples on each block. Note that the last block will have fewer samples if len(array) % chunks != 0.

Returns:
out : ndarray

Array of fill_value with the same shape and type as a.

See also

zeros_like
Return an array of zeros with shape and type of input.
ones_like
Return an array of ones with shape and type of input.
empty_like
Return an empty array with shape and type of input.
zeros
Return a new array setting values to zero.
ones
Return a new array setting values to one.
empty
Return a new uninitialized array.
full
Fill a new array.
dask.array.gradient(f, *varargs, **kwargs)

Return the gradient of an N-dimensional array.

This docstring was copied from numpy.gradient.

Some inconsistencies with the Dask version may exist.

The gradient is computed using second order accurate central differences in the interior points and either first or second order accurate one-sides (forward or backwards) differences at the boundaries. The returned gradient hence has the same shape as the input array.

Parameters:
f : array_like

An N-dimensional array containing samples of a scalar function.

varargs : list of scalar or array, optional

Spacing between f values. Default unitary spacing for all dimensions. Spacing can be specified using:

  1. single scalar to specify a sample distance for all dimensions.
  2. N scalars to specify a constant sample distance for each dimension. i.e. dx, dy, dz, …
  3. N arrays to specify the coordinates of the values along each dimension of F. The length of the array must match the size of the corresponding dimension
  4. Any combination of N scalars/arrays with the meaning of 2. and 3.

If axis is given, the number of varargs must equal the number of axes. Default: 1.

edge_order : {1, 2}, optional

Gradient is calculated using N-th order accurate differences at the boundaries. Default: 1.

New in version 1.9.1.

axis : None or int or tuple of ints, optional

Gradient is calculated only along the given axis or axes The default (axis = None) is to calculate the gradient for all the axes of the input array. axis may be negative, in which case it counts from the last to the first axis.

New in version 1.11.0.

Returns:
gradient : ndarray or list of ndarray

A set of ndarrays (or a single ndarray if there is only one dimension) corresponding to the derivatives of f with respect to each dimension. Each derivative has the same shape as f.

Notes

Assuming that \(f\in C^{3}\) (i.e., \(f\) has at least 3 continuous derivatives) and let \(h_{*}\) be a non-homogeneous stepsize, we minimize the “consistency error” \(\eta_{i}\) between the true gradient and its estimate from a linear combination of the neighboring grid-points:

\[\eta_{i} = f_{i}^{\left(1\right)} - \left[ \alpha f\left(x_{i}\right) + \beta f\left(x_{i} + h_{d}\right) + \gamma f\left(x_{i}-h_{s}\right) \right]\]

By substituting \(f(x_{i} + h_{d})\) and \(f(x_{i} - h_{s})\) with their Taylor series expansion, this translates into solving the following the linear system:

\[\begin{split}\left\{ \begin{array}{r} \alpha+\beta+\gamma=0 \\ \beta h_{d}-\gamma h_{s}=1 \\ \beta h_{d}^{2}+\gamma h_{s}^{2}=0 \end{array} \right.\end{split}\]

The resulting approximation of \(f_{i}^{(1)}\) is the following:

\[\hat f_{i}^{(1)} = \frac{ h_{s}^{2}f\left(x_{i} + h_{d}\right) + \left(h_{d}^{2} - h_{s}^{2}\right)f\left(x_{i}\right) - h_{d}^{2}f\left(x_{i}-h_{s}\right)} { h_{s}h_{d}\left(h_{d} + h_{s}\right)} + \mathcal{O}\left(\frac{h_{d}h_{s}^{2} + h_{s}h_{d}^{2}}{h_{d} + h_{s}}\right)\]

It is worth noting that if \(h_{s}=h_{d}\) (i.e., data are evenly spaced) we find the standard second order approximation:

\[\hat f_{i}^{(1)}= \frac{f\left(x_{i+1}\right) - f\left(x_{i-1}\right)}{2h} + \mathcal{O}\left(h^{2}\right)\]

With a similar procedure the forward/backward approximations used for boundaries can be derived.

References

[1]Quarteroni A., Sacco R., Saleri F. (2007) Numerical Mathematics (Texts in Applied Mathematics). New York: Springer.
[2]Durran D. R. (1999) Numerical Methods for Wave Equations in Geophysical Fluid Dynamics. New York: Springer.
[3]Fornberg B. (1988) Generation of Finite Difference Formulas on Arbitrarily Spaced Grids, Mathematics of Computation 51, no. 184 : 699-706. PDF.

Examples

>>> f = np.array([1, 2, 4, 7, 11, 16], dtype=float)  # doctest: +SKIP
>>> np.gradient(f)  # doctest: +SKIP
array([ 1. ,  1.5,  2.5,  3.5,  4.5,  5. ])
>>> np.gradient(f, 2)  # doctest: +SKIP
array([ 0.5 ,  0.75,  1.25,  1.75,  2.25,  2.5 ])

Spacing can be also specified with an array that represents the coordinates of the values F along the dimensions. For instance a uniform spacing:

>>> x = np.arange(f.size)  # doctest: +SKIP
>>> np.gradient(f, x)  # doctest: +SKIP
array([ 1. ,  1.5,  2.5,  3.5,  4.5,  5. ])

Or a non uniform one:

>>> x = np.array([0., 1., 1.5, 3.5, 4., 6.], dtype=float)  # doctest: +SKIP
>>> np.gradient(f, x)  # doctest: +SKIP
array([ 1. ,  3. ,  3.5,  6.7,  6.9,  2.5])

For two dimensional arrays, the return will be two arrays ordered by axis. In this example the first array stands for the gradient in rows and the second one in columns direction:

>>> np.gradient(np.array([[1, 2, 6], [3, 4, 5]], dtype=float))  # doctest: +SKIP
[array([[ 2.,  2., -1.],
        [ 2.,  2., -1.]]), array([[ 1. ,  2.5,  4. ],
        [ 1. ,  1. ,  1. ]])]

In this example the spacing is also specified: uniform for axis=0 and non uniform for axis=1

>>> dx = 2.  # doctest: +SKIP
>>> y = [1., 1.5, 3.5]  # doctest: +SKIP
>>> np.gradient(np.array([[1, 2, 6], [3, 4, 5]], dtype=float), dx, y)  # doctest: +SKIP
[array([[ 1. ,  1. , -0.5],
        [ 1. ,  1. , -0.5]]), array([[ 2. ,  2. ,  2. ],
        [ 2. ,  1.7,  0.5]])]

It is possible to specify how boundaries are treated using edge_order

>>> x = np.array([0, 1, 2, 3, 4])  # doctest: +SKIP
>>> f = x**2  # doctest: +SKIP
>>> np.gradient(f, edge_order=1)  # doctest: +SKIP
array([ 1.,  2.,  4.,  6.,  7.])
>>> np.gradient(f, edge_order=2)  # doctest: +SKIP
array([-0.,  2.,  4.,  6.,  8.])

The axis keyword can be used to specify a subset of axes of which the gradient is calculated

>>> np.gradient(np.array([[1, 2, 6], [3, 4, 5]], dtype=float), axis=0)  # doctest: +SKIP
array([[ 2.,  2., -1.],
       [ 2.,  2., -1.]])
dask.array.histogram(a, bins=None, range=None, normed=False, weights=None, density=None)

Blocked variant of numpy.histogram().

Follows the signature of numpy.histogram() exactly with the following exceptions:

  • Either an iterable specifying the bins or the number of bins and a range argument is required as computing min and max over blocked arrays is an expensive operation that must be performed explicitly.
  • weights must be a dask.array.Array with the same block structure as a.

Examples

Using number of bins and range:

>>> import dask.array as da
>>> import numpy as np
>>> x = da.from_array(np.arange(10000), chunks=10)
>>> h, bins = da.histogram(x, bins=10, range=[0, 10000])
>>> bins
array([    0.,  1000.,  2000.,  3000.,  4000.,  5000.,  6000.,  7000.,
        8000.,  9000., 10000.])
>>> h.compute()
array([1000, 1000, 1000, 1000, 1000, 1000, 1000, 1000, 1000, 1000])

Explicitly specifying the bins:

>>> h, bins = da.histogram(x, bins=np.array([0, 5000, 10000]))
>>> bins
array([    0,  5000, 10000])
>>> h.compute()
array([5000, 5000])
dask.array.hstack(tup, allow_unknown_chunksizes=False)

Stack arrays in sequence horizontally (column wise).

This docstring was copied from numpy.hstack.

Some inconsistencies with the Dask version may exist.

This is equivalent to concatenation along the second axis, except for 1-D arrays where it concatenates along the first axis. Rebuilds arrays divided by hsplit.

This function makes most sense for arrays with up to 3 dimensions. For instance, for pixel-data with a height (first axis), width (second axis), and r/g/b channels (third axis). The functions concatenate, stack and block provide more general stacking and concatenation operations.

Parameters:
tup : sequence of ndarrays

The arrays must have the same shape along all but the second axis, except 1-D arrays which can be any length.

Returns:
stacked : ndarray

The array formed by stacking the given arrays.

See also

stack
Join a sequence of arrays along a new axis.
vstack
Stack arrays in sequence vertically (row wise).
dstack
Stack arrays in sequence depth wise (along third axis).
concatenate
Join a sequence of arrays along an existing axis.
hsplit
Split array along second axis.
block
Assemble arrays from blocks.

Examples

>>> a = np.array((1,2,3))  # doctest: +SKIP
>>> b = np.array((2,3,4))  # doctest: +SKIP
>>> np.hstack((a,b))  # doctest: +SKIP
array([1, 2, 3, 2, 3, 4])
>>> a = np.array([[1],[2],[3]])  # doctest: +SKIP
>>> b = np.array([[2],[3],[4]])  # doctest: +SKIP
>>> np.hstack((a,b))  # doctest: +SKIP
array([[1, 2],
       [2, 3],
       [3, 4]])
dask.array.hypot(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Given the “legs” of a right triangle, return its hypotenuse.

Equivalent to sqrt(x1**2 + x2**2), element-wise. If x1 or x2 is scalar_like (i.e., unambiguously cast-able to a scalar type), it is broadcast for use with each element of the other argument. (See Examples)

Parameters:
x1, x2 : array_like

Leg of the triangle(s).

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
z : ndarray

The hypotenuse of the triangle(s). This is a scalar if both x1 and x2 are scalars.

Examples

>>> np.hypot(3*np.ones((3, 3)), 4*np.ones((3, 3)))  # doctest: +SKIP
array([[ 5.,  5.,  5.],
       [ 5.,  5.,  5.],
       [ 5.,  5.,  5.]])

Example showing broadcast of scalar_like argument:

>>> np.hypot(3*np.ones((3, 3)), [4])  # doctest: +SKIP
array([[ 5.,  5.,  5.],
       [ 5.,  5.,  5.],
       [ 5.,  5.,  5.]])
dask.array.imag(*args, **kwargs)

Return the imaginary part of the complex argument.

Parameters:
val : array_like

Input array.

Returns:
out : ndarray or scalar

The imaginary component of the complex argument. If val is real, the type of val is used for the output. If val has complex elements, the returned type is float.

See also

real, angle, real_if_close

Examples

>>> a = np.array([1+2j, 3+4j, 5+6j])  # doctest: +SKIP
>>> a.imag  # doctest: +SKIP
array([ 2.,  4.,  6.])
>>> a.imag = np.array([8, 10, 12])  # doctest: +SKIP
>>> a  # doctest: +SKIP
array([ 1. +8.j,  3.+10.j,  5.+12.j])
>>> np.imag(1 + 1j)  # doctest: +SKIP
1.0
dask.array.indices(dimensions, dtype=<class 'int'>, chunks='auto')

Implements NumPy’s indices for Dask Arrays.

Generates a grid of indices covering the dimensions provided.

The final array has the shape (len(dimensions), *dimensions). The chunks are used to specify the chunking for axis 1 up to len(dimensions). The 0th axis always has chunks of length 1.

Parameters:
dimensions : sequence of ints

The shape of the index grid.

dtype : dtype, optional

Type to use for the array. Default is int.

chunks : sequence of ints

The number of samples on each block. Note that the last block will have fewer samples if len(array) % chunks != 0.

Returns:
grid : dask array
dask.array.insert(arr, obj, values, axis)

Insert values along the given axis before the given indices.

This docstring was copied from numpy.insert.

Some inconsistencies with the Dask version may exist.

Parameters:
arr : array_like

Input array.

obj : int, slice or sequence of ints

Object that defines the index or indices before which values is inserted.

New in version 1.8.0.

Support for multiple insertions when obj is a single scalar or a sequence with one element (similar to calling insert multiple times).

values : array_like

Values to insert into arr. If the type of values is different from that of arr, values is converted to the type of arr. values should be shaped so that arr[...,obj,...] = values is legal.

axis : int, optional

Axis along which to insert values. If axis is None then arr is flattened first.

Returns:
out : ndarray

A copy of arr with values inserted. Note that insert does not occur in-place: a new array is returned. If axis is None, out is a flattened array.

See also

append
Append elements at the end of an array.
concatenate
Join a sequence of arrays along an existing axis.
delete
Delete elements from an array.

Notes

Note that for higher dimensional inserts obj=0 behaves very different from obj=[0] just like arr[:,0,:] = values is different from arr[:,[0],:] = values.

Examples

>>> a = np.array([[1, 1], [2, 2], [3, 3]])  # doctest: +SKIP
>>> a  # doctest: +SKIP
array([[1, 1],
       [2, 2],
       [3, 3]])
>>> np.insert(a, 1, 5)  # doctest: +SKIP
array([1, 5, 1, 2, 2, 3, 3])
>>> np.insert(a, 1, 5, axis=1)  # doctest: +SKIP
array([[1, 5, 1],
       [2, 5, 2],
       [3, 5, 3]])

Difference between sequence and scalars:

>>> np.insert(a, [1], [[1],[2],[3]], axis=1)  # doctest: +SKIP
array([[1, 1, 1],
       [2, 2, 2],
       [3, 3, 3]])
>>> np.array_equal(np.insert(a, 1, [1, 2, 3], axis=1),  # doctest: +SKIP
...                np.insert(a, [1], [[1],[2],[3]], axis=1))
True
>>> b = a.flatten()  # doctest: +SKIP
>>> b  # doctest: +SKIP
array([1, 1, 2, 2, 3, 3])
>>> np.insert(b, [2, 2], [5, 6])  # doctest: +SKIP
array([1, 1, 5, 6, 2, 2, 3, 3])
>>> np.insert(b, slice(2, 4), [5, 6])  # doctest: +SKIP
array([1, 1, 5, 2, 6, 2, 3, 3])
>>> np.insert(b, [2, 2], [7.13, False]) # type casting  # doctest: +SKIP
array([1, 1, 7, 0, 2, 2, 3, 3])
>>> x = np.arange(8).reshape(2, 4)  # doctest: +SKIP
>>> idx = (1, 3)  # doctest: +SKIP
>>> np.insert(x, idx, 999, axis=1)  # doctest: +SKIP
array([[  0, 999,   1,   2, 999,   3],
       [  4, 999,   5,   6, 999,   7]])
dask.array.invert(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute bit-wise inversion, or bit-wise NOT, element-wise.

Computes the bit-wise NOT of the underlying binary representation of the integers in the input arrays. This ufunc implements the C/Python operator ~.

For signed integer inputs, the two’s complement is returned. In a two’s-complement system negative numbers are represented by the two’s complement of the absolute value. This is the most common method of representing signed integers on computers [1]. A N-bit two’s-complement system can represent every integer in the range \(-2^{N-1}\) to \(+2^{N-1}-1\).

Parameters:
x : array_like

Only integer and boolean types are handled.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Result. This is a scalar if x is a scalar.

See also

bitwise_and, bitwise_or, bitwise_xor, logical_not

binary_repr
Return the binary representation of the input number as a string.

Notes

bitwise_not is an alias for invert:

>>> np.bitwise_not is np.invert  # doctest: +SKIP
True

References

[1](1, 2) Wikipedia, “Two’s complement”, https://en.wikipedia.org/wiki/Two’s_complement

Examples

We’ve seen that 13 is represented by 00001101. The invert or bit-wise NOT of 13 is then:

>>> np.invert(np.array([13], dtype=uint8))  # doctest: +SKIP
array([242], dtype=uint8)
>>> np.binary_repr(x, width=8)  # doctest: +SKIP
'00001101'
>>> np.binary_repr(242, width=8)  # doctest: +SKIP
'11110010'

The result depends on the bit-width:

>>> np.invert(np.array([13], dtype=uint16))  # doctest: +SKIP
array([65522], dtype=uint16)
>>> np.binary_repr(x, width=16)  # doctest: +SKIP
'0000000000001101'
>>> np.binary_repr(65522, width=16)  # doctest: +SKIP
'1111111111110010'

When using signed integer types the result is the two’s complement of the result for the unsigned type:

>>> np.invert(np.array([13], dtype=int8))  # doctest: +SKIP
array([-14], dtype=int8)
>>> np.binary_repr(-14, width=8)  # doctest: +SKIP
'11110010'

Booleans are accepted as well:

>>> np.invert(array([True, False]))  # doctest: +SKIP
array([False,  True])
dask.array.isclose(arr1, arr2, rtol=1e-05, atol=1e-08, equal_nan=False)

Returns a boolean array where two arrays are element-wise equal within a tolerance.

This docstring was copied from numpy.isclose.

Some inconsistencies with the Dask version may exist.

The tolerance values are positive, typically very small numbers. The relative difference (rtol * abs(b)) and the absolute difference atol are added together to compare against the absolute difference between a and b.

Warning

The default atol is not appropriate for comparing numbers that are much smaller than one (see Notes).

Parameters:
a, b : array_like

Input arrays to compare.

rtol : float

The relative tolerance parameter (see Notes).

atol : float

The absolute tolerance parameter (see Notes).

equal_nan : bool

Whether to compare NaN’s as equal. If True, NaN’s in a will be considered equal to NaN’s in b in the output array.

Returns:
y : array_like

Returns a boolean array of where a and b are equal within the given tolerance. If both a and b are scalars, returns a single boolean value.

See also

allclose

Notes

New in version 1.7.0.

For finite values, isclose uses the following equation to test whether two floating point values are equivalent.

absolute(a - b) <= (atol + rtol * absolute(b))

Unlike the built-in math.isclose, the above equation is not symmetric in a and b – it assumes b is the reference value – so that isclose(a, b) might be different from isclose(b, a). Furthermore, the default value of atol is not zero, and is used to determine what small values should be considered close to zero. The default value is appropriate for expected values of order unity: if the expected values are significantly smaller than one, it can result in false positives. atol should be carefully selected for the use case at hand. A zero value for atol will result in False if either a or b is zero.

Examples

>>> np.isclose([1e10,1e-7], [1.00001e10,1e-8])  # doctest: +SKIP
array([True, False])
>>> np.isclose([1e10,1e-8], [1.00001e10,1e-9])  # doctest: +SKIP
array([True, True])
>>> np.isclose([1e10,1e-8], [1.0001e10,1e-9])  # doctest: +SKIP
array([False, True])
>>> np.isclose([1.0, np.nan], [1.0, np.nan])  # doctest: +SKIP
array([True, False])
>>> np.isclose([1.0, np.nan], [1.0, np.nan], equal_nan=True)  # doctest: +SKIP
array([True, True])
>>> np.isclose([1e-8, 1e-7], [0.0, 0.0])  # doctest: +SKIP
array([ True, False], dtype=bool)
>>> np.isclose([1e-100, 1e-7], [0.0, 0.0], atol=0.0)  # doctest: +SKIP
array([False, False], dtype=bool)
>>> np.isclose([1e-10, 1e-10], [1e-20, 0.0])  # doctest: +SKIP
array([ True,  True], dtype=bool)
>>> np.isclose([1e-10, 1e-10], [1e-20, 0.999999e-10], atol=0.0)  # doctest: +SKIP
array([False,  True], dtype=bool)
dask.array.iscomplex(*args, **kwargs)

Returns a bool array, where True if input element is complex.

What is tested is whether the input has a non-zero imaginary part, not if the input type is complex.

Parameters:
x : array_like

Input array.

Returns:
out : ndarray of bools

Output array.

See also

isreal

iscomplexobj
Return True if x is a complex type or an array of complex numbers.

Examples

>>> np.iscomplex([1+1j, 1+0j, 4.5, 3, 2, 2j])  # doctest: +SKIP
array([ True, False, False, False, False,  True])
dask.array.isfinite(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Test element-wise for finiteness (not infinity or not Not a Number).

The result is returned as a boolean array.

Parameters:
x : array_like

Input values.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray, bool

True where x is not positive infinity, negative infinity, or NaN; false otherwise. This is a scalar if x is a scalar.

Notes

Not a Number, positive infinity and negative infinity are considered to be non-finite.

NumPy uses the IEEE Standard for Binary Floating-Point for Arithmetic (IEEE 754). This means that Not a Number is not equivalent to infinity. Also that positive infinity is not equivalent to negative infinity. But infinity is equivalent to positive infinity. Errors result if the second argument is also supplied when x is a scalar input, or if first and second arguments have different shapes.

Examples

>>> np.isfinite(1)  # doctest: +SKIP
True
>>> np.isfinite(0)  # doctest: +SKIP
True
>>> np.isfinite(np.nan)  # doctest: +SKIP
False
>>> np.isfinite(np.inf)  # doctest: +SKIP
False
>>> np.isfinite(np.NINF)  # doctest: +SKIP
False
>>> np.isfinite([np.log(-1.),1.,np.log(0)])  # doctest: +SKIP
array([False,  True, False])
>>> x = np.array([-np.inf, 0., np.inf])  # doctest: +SKIP
>>> y = np.array([2, 2, 2])  # doctest: +SKIP
>>> np.isfinite(x, y)  # doctest: +SKIP
array([0, 1, 0])
>>> y  # doctest: +SKIP
array([0, 1, 0])
dask.array.isin(element, test_elements, assume_unique=False, invert=False)

Calculates element in test_elements, broadcasting over element only. Returns a boolean array of the same shape as element that is True where an element of element is in test_elements and False otherwise.

Parameters:
element : array_like

Input array.

test_elements : array_like

The values against which to test each value of element. This argument is flattened if it is an array or array_like. See notes for behavior with non-array-like parameters.

assume_unique : bool, optional

If True, the input arrays are both assumed to be unique, which can speed up the calculation. Default is False.

invert : bool, optional

If True, the values in the returned array are inverted, as if calculating element not in test_elements. Default is False. np.isin(a, b, invert=True) is equivalent to (but faster than) np.invert(np.isin(a, b)).

Returns:
isin : ndarray, bool

Has the same shape as element. The values element[isin] are in test_elements.

See also

in1d
Flattened version of this function.
numpy.lib.arraysetops
Module with a number of other functions for performing set operations on arrays.

Notes

isin is an element-wise function version of the python keyword in. isin(a, b) is roughly equivalent to np.array([item in b for item in a]) if a and b are 1-D sequences.

element and test_elements are converted to arrays if they are not already. If test_elements is a set (or other non-sequence collection) it will be converted to an object array with one element, rather than an array of the values contained in test_elements. This is a consequence of the array constructor’s way of handling non-sequence collections. Converting the set to a list usually gives the desired behavior.

New in version 1.13.0.

Examples

>>> element = 2*np.arange(4).reshape((2, 2))
>>> element
array([[0, 2],
       [4, 6]])
>>> test_elements = [1, 2, 4, 8]
>>> mask = np.isin(element, test_elements)
>>> mask
array([[ False,  True],
       [ True,  False]])
>>> element[mask]
array([2, 4])

The indices of the matched values can be obtained with nonzero:

>>> np.nonzero(mask)
(array([0, 1]), array([1, 0]))

The test can also be inverted:

>>> mask = np.isin(element, test_elements, invert=True)
>>> mask
array([[ True, False],
       [ False, True]])
>>> element[mask]
array([0, 6])

Because of how array handles sets, the following does not work as expected:

>>> test_set = {1, 2, 4, 8}
>>> np.isin(element, test_set)
array([[ False, False],
       [ False, False]])

Casting the set to a list gives the expected result:

>>> np.isin(element, list(test_set))
array([[ False,  True],
       [ True,  False]])
dask.array.isinf(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Test element-wise for positive or negative infinity.

Returns a boolean array of the same shape as x, True where x == +/-inf, otherwise False.

Parameters:
x : array_like

Input values

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : bool (scalar) or boolean ndarray

True where x is positive or negative infinity, false otherwise. This is a scalar if x is a scalar.

Notes

NumPy uses the IEEE Standard for Binary Floating-Point for Arithmetic (IEEE 754).

Errors result if the second argument is supplied when the first argument is a scalar, or if the first and second arguments have different shapes.

Examples

>>> np.isinf(np.inf)  # doctest: +SKIP
True
>>> np.isinf(np.nan)  # doctest: +SKIP
False
>>> np.isinf(np.NINF)  # doctest: +SKIP
True
>>> np.isinf([np.inf, -np.inf, 1.0, np.nan])  # doctest: +SKIP
array([ True,  True, False, False])
>>> x = np.array([-np.inf, 0., np.inf])  # doctest: +SKIP
>>> y = np.array([2, 2, 2])  # doctest: +SKIP
>>> np.isinf(x, y)  # doctest: +SKIP
array([1, 0, 1])
>>> y  # doctest: +SKIP
array([1, 0, 1])
dask.array.isneginf(*args, **kwargs)

Return (x1 == x2) element-wise.

Parameters:
x1, x2 : array_like

Input arrays of the same shape.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Output array, element-wise comparison of x1 and x2. Typically of type bool, unless dtype=object is passed. This is a scalar if both x1 and x2 are scalars.

See also

not_equal, greater_equal, less_equal, greater, less

Examples

>>> np.equal([0, 1, 3], np.arange(3))  # doctest: +SKIP
array([ True,  True, False])

What is compared are values, not types. So an int (1) and an array of length one can evaluate as True:

>>> np.equal(1, np.ones(1))  # doctest: +SKIP
array([ True])
dask.array.isnan(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Test element-wise for NaN and return result as a boolean array.

Parameters:
x : array_like

Input array.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or bool

True where x is NaN, false otherwise. This is a scalar if x is a scalar.

See also

isinf, isneginf, isposinf, isfinite, isnat

Notes

NumPy uses the IEEE Standard for Binary Floating-Point for Arithmetic (IEEE 754). This means that Not a Number is not equivalent to infinity.

Examples

>>> np.isnan(np.nan)  # doctest: +SKIP
True
>>> np.isnan(np.inf)  # doctest: +SKIP
False
>>> np.isnan([np.log(-1.),1.,np.log(0)])  # doctest: +SKIP
array([ True, False, False])
dask.array.isnull(values)

pandas.isnull for dask arrays

dask.array.isposinf(*args, **kwargs)

Return (x1 == x2) element-wise.

Parameters:
x1, x2 : array_like

Input arrays of the same shape.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

Output array, element-wise comparison of x1 and x2. Typically of type bool, unless dtype=object is passed. This is a scalar if both x1 and x2 are scalars.

See also

not_equal, greater_equal, less_equal, greater, less

Examples

>>> np.equal([0, 1, 3], np.arange(3))  # doctest: +SKIP
array([ True,  True, False])

What is compared are values, not types. So an int (1) and an array of length one can evaluate as True:

>>> np.equal(1, np.ones(1))  # doctest: +SKIP
array([ True])
dask.array.isreal(*args, **kwargs)

Returns a bool array, where True if input element is real.

If element has complex type with zero complex part, the return value for that element is True.

Parameters:
x : array_like

Input array.

Returns:
out : ndarray, bool

Boolean array of same shape as x.

See also

iscomplex

isrealobj
Return True if x is not a complex type.

Examples

>>> np.isreal([1+1j, 1+0j, 4.5, 3, 2, 2j])  # doctest: +SKIP
array([False,  True,  True,  True,  True, False])
dask.array.ldexp(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Returns x1 * 2**x2, element-wise.

The mantissas x1 and twos exponents x2 are used to construct floating point numbers x1 * 2**x2.

Parameters:
x1 : array_like

Array of multipliers.

x2 : array_like, int

Array of twos exponents.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or scalar

The result of x1 * 2**x2. This is a scalar if both x1 and x2 are scalars.

See also

frexp
Return (y1, y2) from x = y1 * 2**y2, inverse to ldexp.

Notes

Complex dtypes are not supported, they will raise a TypeError.

ldexp is useful as the inverse of frexp, if used by itself it is more clear to simply use the expression x1 * 2**x2.

Examples

>>> np.ldexp(5, np.arange(4))  # doctest: +SKIP
array([  5.,  10.,  20.,  40.], dtype=float32)
>>> x = np.arange(6)  # doctest: +SKIP
>>> np.ldexp(*np.frexp(x))  # doctest: +SKIP
array([ 0.,  1.,  2.,  3.,  4.,  5.])
dask.array.linspace(start, stop, num=50, endpoint=True, retstep=False, chunks='auto', dtype=None)

Return num evenly spaced values over the closed interval [start, stop].

Parameters:
start : scalar

The starting value of the sequence.

stop : scalar

The last value of the sequence.

num : int, optional

Number of samples to include in the returned dask array, including the endpoints. Default is 50.

endpoint : bool, optional

If True, stop is the last sample. Otherwise, it is not included. Default is True.

retstep : bool, optional

If True, return (samples, step), where step is the spacing between samples. Default is False.

chunks : int

The number of samples on each block. Note that the last block will have fewer samples if num % blocksize != 0

dtype : dtype, optional

The type of the output array.

Returns:
samples : dask array
step : float, optional

Only returned if retstep is True. Size of spacing between samples.

dask.array.log(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Natural logarithm, element-wise.

The natural logarithm log is the inverse of the exponential function, so that log(exp(x)) = x. The natural logarithm is logarithm in base e.

Parameters:
x : array_like

Input value.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray

The natural logarithm of x, element-wise. This is a scalar if x is a scalar.

See also

log10, log2, log1p, emath.log

Notes

Logarithm is a multivalued function: for each x there is an infinite number of z such that exp(z) = x. The convention is to return the z whose imaginary part lies in [-pi, pi].

For real-valued input data types, log always returns real output. For each value that cannot be expressed as a real number or infinity, it yields nan and sets the invalid floating point error flag.

For complex-valued input, log is a complex analytical function that has a branch cut [-inf, 0] and is continuous from above on it. log handles the floating-point negative zero as an infinitesimal negative number, conforming to the C99 standard.

References

[1]M. Abramowitz and I.A. Stegun, “Handbook of Mathematical Functions”, 10th printing, 1964, pp. 67. http://www.math.sfu.ca/~cbm/aands/
[2]Wikipedia, “Logarithm”. https://en.wikipedia.org/wiki/Logarithm

Examples

>>> np.log([1, np.e, np.e**2, 0])  # doctest: +SKIP
array([  0.,   1.,   2., -Inf])
dask.array.log10(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Return the base 10 logarithm of the input array, element-wise.

Parameters:
x : array_like

Input values.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray

The logarithm to the base 10 of x, element-wise. NaNs are returned where x is negative. This is a scalar if x is a scalar.

See also

emath.log10

Notes

Logarithm is a multivalued function: for each x there is an infinite number of z such that 10**z = x. The convention is to return the z whose imaginary part lies in [-pi, pi].

For real-valued input data types, log10 always returns real output. For each value that cannot be expressed as a real number or infinity, it yields nan and sets the invalid floating point error flag.

For complex-valued input, log10 is a complex analytical function that has a branch cut [-inf, 0] and is continuous from above on it. log10 handles the floating-point negative zero as an infinitesimal negative number, conforming to the C99 standard.

References

[1]M. Abramowitz and I.A. Stegun, “Handbook of Mathematical Functions”, 10th printing, 1964, pp. 67. http://www.math.sfu.ca/~cbm/aands/
[2]Wikipedia, “Logarithm”. https://en.wikipedia.org/wiki/Logarithm

Examples

>>> np.log10([1e-15, -3.])  # doctest: +SKIP
array([-15.,  NaN])
dask.array.log1p(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Return the natural logarithm of one plus the input array, element-wise.

Calculates log(1 + x).

Parameters:
x : array_like

Input values.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray

Natural logarithm of 1 + x, element-wise. This is a scalar if x is a scalar.

See also

expm1
exp(x) - 1, the inverse of log1p.

Notes

For real-valued input, log1p is accurate also for x so small that 1 + x == 1 in floating-point accuracy.

Logarithm is a multivalued function: for each x there is an infinite number of z such that exp(z) = 1 + x. The convention is to return the z whose imaginary part lies in [-pi, pi].

For real-valued input data types, log1p always returns real output. For each value that cannot be expressed as a real number or infinity, it yields nan and sets the invalid floating point error flag.

For complex-valued input, log1p is a complex analytical function that has a branch cut [-inf, -1] and is continuous from above on it. log1p handles the floating-point negative zero as an infinitesimal negative number, conforming to the C99 standard.

References

[1]M. Abramowitz and I.A. Stegun, “Handbook of Mathematical Functions”, 10th printing, 1964, pp. 67. http://www.math.sfu.ca/~cbm/aands/
[2]Wikipedia, “Logarithm”. https://en.wikipedia.org/wiki/Logarithm

Examples

>>> np.log1p(1e-99)  # doctest: +SKIP
1e-99
>>> np.log(1 + 1e-99)  # doctest: +SKIP
0.0
dask.array.log2(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Base-2 logarithm of x.

Parameters:
x : array_like

Input values.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray

Base-2 logarithm of x. This is a scalar if x is a scalar.

See also

log, log10, log1p, emath.log2

Notes

New in version 1.3.0.

Logarithm is a multivalued function: for each x there is an infinite number of z such that 2**z = x. The convention is to return the z whose imaginary part lies in [-pi, pi].

For real-valued input data types, log2 always returns real output. For each value that cannot be expressed as a real number or infinity, it yields nan and sets the invalid floating point error flag.

For complex-valued input, log2 is a complex analytical function that has a branch cut [-inf, 0] and is continuous from above on it. log2 handles the floating-point negative zero as an infinitesimal negative number, conforming to the C99 standard.

Examples

>>> x = np.array([0, 1, 2, 2**4])  # doctest: +SKIP
>>> np.log2(x)  # doctest: +SKIP
array([-Inf,   0.,   1.,   4.])
>>> xi = np.array([0+1.j, 1, 2+0.j, 4.j])  # doctest: +SKIP
>>> np.log2(xi)  # doctest: +SKIP
array([ 0.+2.26618007j,  0.+0.j        ,  1.+0.j        ,  2.+2.26618007j])
dask.array.logaddexp(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Logarithm of the sum of exponentiations of the inputs.

Calculates log(exp(x1) + exp(x2)). This function is useful in statistics where the calculated probabilities of events may be so small as to exceed the range of normal floating point numbers. In such cases the logarithm of the calculated probability is stored. This function allows adding probabilities stored in such a fashion.

Parameters:
x1, x2 : array_like

Input values.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
result : ndarray

Logarithm of exp(x1) + exp(x2). This is a scalar if both x1 and x2 are scalars.

See also

logaddexp2
Logarithm of the sum of exponentiations of inputs in base 2.

Notes

New in version 1.3.0.

Examples

>>> prob1 = np.log(1e-50)  # doctest: +SKIP
>>> prob2 = np.log(2.5e-50)  # doctest: +SKIP
>>> prob12 = np.logaddexp(prob1, prob2)  # doctest: +SKIP
>>> prob12  # doctest: +SKIP
-113.87649168120691
>>> np.exp(prob12)  # doctest: +SKIP
3.5000000000000057e-50
dask.array.logaddexp2(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Logarithm of the sum of exponentiations of the inputs in base-2.

Calculates log2(2**x1 + 2**x2). This function is useful in machine learning when the calculated probabilities of events may be so small as to exceed the range of normal floating point numbers. In such cases the base-2 logarithm of the calculated probability can be used instead. This function allows adding probabilities stored in such a fashion.

Parameters:
x1, x2 : array_like

Input values.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
result : ndarray

Base-2 logarithm of 2**x1 + 2**x2. This is a scalar if both x1 and x2 are scalars.

See also

logaddexp
Logarithm of the sum of exponentiations of the inputs.

Notes

New in version 1.3.0.

Examples

>>> prob1 = np.log2(1e-50)  # doctest: +SKIP
>>> prob2 = np.log2(2.5e-50)  # doctest: +SKIP
>>> prob12 = np.logaddexp2(prob1, prob2)  # doctest: +SKIP
>>> prob1, prob2, prob12  # doctest: +SKIP
(-166.09640474436813, -164.77447664948076, -164.28904982231052)
>>> 2**prob12  # doctest: +SKIP
3.4999999999999914e-50
dask.array.logical_and(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute the truth value of x1 AND x2 element-wise.

Parameters:
x1, x2 : array_like

Input arrays. x1 and x2 must be of the same shape.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or bool

Boolean result with the same shape as x1 and x2 of the logical AND operation on corresponding elements of x1 and x2. This is a scalar if both x1 and x2 are scalars.

Examples

>>> np.logical_and(True, False)  # doctest: +SKIP
False
>>> np.logical_and([True, False], [False, False])  # doctest: +SKIP
array([False, False])
>>> x = np.arange(5)  # doctest: +SKIP
>>> np.logical_and(x>1, x<4)  # doctest: +SKIP
array([False, False,  True,  True, False])
dask.array.logical_not(x, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute the truth value of NOT x element-wise.

Parameters:
x : array_like

Logical NOT is applied to the elements of x.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : bool or ndarray of bool

Boolean result with the same shape as x of the NOT operation on elements of x. This is a scalar if x is a scalar.

Examples

>>> np.logical_not(3)  # doctest: +SKIP
False
>>> np.logical_not([True, False, 0, 1])  # doctest: +SKIP
array([False,  True,  True, False])
>>> x = np.arange(5)  # doctest: +SKIP
>>> np.logical_not(x<3)  # doctest: +SKIP
array([False, False, False,  True,  True])
dask.array.logical_or(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute the truth value of x1 OR x2 element-wise.

Parameters:
x1, x2 : array_like

Logical OR is applied to the elements of x1 and x2. They have to be of the same shape.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or bool

Boolean result with the same shape as x1 and x2 of the logical OR operation on elements of x1 and x2. This is a scalar if both x1 and x2 are scalars.

Examples

>>> np.logical_or(True, False)  # doctest: +SKIP
True
>>> np.logical_or([True, False], [False, False])  # doctest: +SKIP
array([ True, False])
>>> x = np.arange(5)  # doctest: +SKIP
>>> np.logical_or(x < 1, x > 3)  # doctest: +SKIP
array([ True, False, False, False,  True])
dask.array.logical_xor(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Compute the truth value of x1 XOR x2, element-wise.

Parameters:
x1, x2 : array_like

Logical XOR is applied to the elements of x1 and x2. They must be broadcastable to the same shape.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : bool or ndarray of bool

Boolean result of the logical XOR operation applied to the elements of x1 and x2; the shape is determined by whether or not broadcasting of one or both arrays was required. This is a scalar if both x1 and x2 are scalars.

Examples

>>> np.logical_xor(True, False)  # doctest: +SKIP
True
>>> np.logical_xor([True, True, False, False], [True, False, True, False])  # doctest: +SKIP
array([False,  True,  True, False])
>>> x = np.arange(5)  # doctest: +SKIP
>>> np.logical_xor(x < 1, x > 3)  # doctest: +SKIP
array([ True, False, False, False,  True])

Simple example showing support of broadcasting

>>> np.logical_xor(0, np.eye(2))  # doctest: +SKIP
array([[ True, False],
       [False,  True]])
dask.array.matmul(x1, x2, /, out=None, *, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

This docstring was copied from numpy.matmul.

Some inconsistencies with the Dask version may exist.

Matrix product of two arrays.

Parameters:
x1, x2 : array_like

Input arrays, scalars not allowed.

out : ndarray, optional

A location into which the result is stored. If provided, it must have a shape that matches the signature (n,k),(k,m)->(n,m). If not provided or None, a freshly-allocated array is returned.

**kwargs

For other keyword-only arguments, see the ufunc docs.

..versionadded:: 1.16

Now handles ufunc kwargs

Returns:
y : ndarray

The matrix product of the inputs. This is a scalar only when both x1, x2 are 1-d vectors.

Raises:
ValueError

If the last dimension of a is not the same size as the second-to-last dimension of b.

If a scalar value is passed in.

See also

vdot
Complex-conjugating dot product.
tensordot
Sum products over arbitrary axes.
einsum
Einstein summation convention.
dot
alternative matrix product with different broadcasting rules.

Notes

The behavior depends on the arguments in the following way.

  • If both arguments are 2-D they are multiplied like conventional matrices.
  • If either argument is N-D, N > 2, it is treated as a stack of matrices residing in the last two indexes and broadcast accordingly.
  • If the first argument is 1-D, it is promoted to a matrix by prepending a 1 to its dimensions. After matrix multiplication the prepended 1 is removed.
  • If the second argument is 1-D, it is promoted to a matrix by appending a 1 to its dimensions. After matrix multiplication the appended 1 is removed.

matmul differs from dot in two important ways:

  • Multiplication by scalars is not allowed, use * instead.

  • Stacks of matrices are broadcast together as if the matrices were elements, respecting the signature (n,k),(k,m)->(n,m):

    >>> a = a = np.full([9,5,7,3], True, dtype=bool)  # doctest: +SKIP
    >>> c = np.full([9, 5, 4,3], True, dtype=bool)  # doctest: +SKIP
    >>> np.dot(a, c).shape  # doctest: +SKIP
    (9, 5, 7, 9, 5, 4)
    >>> np.matmul(a, c).shape # n is 5, k is 3, m is 4  # doctest: +SKIP
    (9, 5, 7, 4)
    

The matmul function implements the semantics of the @ operator introduced in Python 3.5 following PEP465.

Examples

For 2-D arrays it is the matrix product:

>>> a = np.array([[1, 0],  # doctest: +SKIP
...               [0, 1]])
>>> b = np.array([[4, 1],   # doctest: +SKIP
...               [2, 2]]
>>> np.matmul(a, b)  # doctest: +SKIP
array([[4, 1],
       [2, 2]])

For 2-D mixed with 1-D, the result is the usual.

>>> a = np.array([[1, 0],  # doctest: +SKIP
...               [0, 1]]
>>> b = np.array([1, 2])  # doctest: +SKIP
>>> np.matmul(a, b)  # doctest: +SKIP
array([1, 2])
>>> np.matmul(b, a)  # doctest: +SKIP
array([1, 2])

Broadcasting is conventional for stacks of arrays

>>> a = np.arange(2 * 2 * 4).reshape((2, 2, 4))  # doctest: +SKIP
>>> b = np.arange(2 * 2 * 4).reshape((2, 4, 2))  # doctest: +SKIP
>>> np.matmul(a,b).shape  # doctest: +SKIP
(2, 2, 2)
>>> np.matmul(a, b)[0, 1, 1]  # doctest: +SKIP
98
>>> sum(a[0, 1, :] * b[0 , :, 1])  # doctest: +SKIP
98

Vector, vector returns the scalar inner product, but neither argument is complex-conjugated:

>>> np.matmul([2j, 3j], [2j, 3j])  # doctest: +SKIP
(-13+0j)

Scalar multiplication raises an error.

>>> np.matmul([1,2], 3)  # doctest: +SKIP
Traceback (most recent call last):
...
ValueError: matmul: Input operand 1 does not have enough dimensions ...

New in version 1.10.0.

dask.array.max(a, axis=None, out=None, keepdims=<no value>, initial=<no value>)

Return the maximum of an array or maximum along an axis.

Parameters:
a : array_like

Input data.

axis : None or int or tuple of ints, optional

Axis or axes along which to operate. By default, flattened input is used.

New in version 1.7.0.

If this is a tuple of ints, the maximum is selected over multiple axes, instead of a single axis or all the axes as before.

out : ndarray, optional

Alternative output array in which to place the result. Must be of the same shape and buffer length as the expected output. See doc.ufuncs (Section “Output arguments”) for more details.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the input array.

If the default value is passed, then keepdims will not be passed through to the amax method of sub-classes of ndarray, however any non-default value will be. If the sub-class’ method does not implement keepdims any exceptions will be raised.

initial : scalar, optional

The minimum value of an output element. Must be present to allow computation on empty slice. See ~numpy.ufunc.reduce for details.

New in version 1.15.0.

Returns:
amax : ndarray or scalar

Maximum of a. If axis is None, the result is a scalar value. If axis is given, the result is an array of dimension a.ndim - 1.

See also

amin
The minimum value of an array along a given axis, propagating any NaNs.
nanmax
The maximum value of an array along a given axis, ignoring any NaNs.
maximum
Element-wise maximum of two arrays, propagating any NaNs.
fmax
Element-wise maximum of two arrays, ignoring any NaNs.
argmax
Return the indices of the maximum values.

nanmin, minimum, fmin

Notes

NaN values are propagated, that is if at least one item is NaN, the corresponding max value will be NaN as well. To ignore NaN values (MATLAB behavior), please use nanmax.

Don’t use amax for element-wise comparison of 2 arrays; when a.shape[0] is 2, maximum(a[0], a[1]) is faster than amax(a, axis=0).

Examples

>>> a = np.arange(4).reshape((2,2))
>>> a
array([[0, 1],
       [2, 3]])
>>> np.amax(a)           # Maximum of the flattened array
3
>>> np.amax(a, axis=0)   # Maxima along the first axis
array([2, 3])
>>> np.amax(a, axis=1)   # Maxima along the second axis
array([1, 3])
>>> b = np.arange(5, dtype=float)
>>> b[2] = np.NaN
>>> np.amax(b)
nan
>>> np.nanmax(b)
4.0

You can use an initial value to compute the maximum of an empty slice, or to initialize it to a different value:

>>> np.max([[-50], [10]], axis=-1, initial=0)
array([ 0, 10])

Notice that the initial value is used as one of the elements for which the maximum is determined, unlike for the default argument Python’s max function, which is only used for empty iterables.

>>> np.max([5], initial=6)
6
>>> max([5], default=6)
5
dask.array.maximum(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Element-wise maximum of array elements.

Compare two arrays and returns a new array containing the element-wise maxima. If one of the elements being compared is a NaN, then that element is returned. If both elements are NaNs then the first is returned. The latter distinction is important for complex NaNs, which are defined as at least one of the real or imaginary parts being a NaN. The net effect is that NaNs are propagated.

Parameters:
x1, x2 : array_like

The arrays holding the elements to be compared. They must have the same shape, or shapes that can be broadcast to a single shape.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or scalar

The maximum of x1 and x2, element-wise. This is a scalar if both x1 and x2 are scalars.

See also

minimum
Element-wise minimum of two arrays, propagates NaNs.
fmax
Element-wise maximum of two arrays, ignores NaNs.
amax
The maximum value of an array along a given axis, propagates NaNs.
nanmax
The maximum value of an array along a given axis, ignores NaNs.

fmin, amin, nanmin

Notes

The maximum is equivalent to np.where(x1 >= x2, x1, x2) when neither x1 nor x2 are nans, but it is faster and does proper broadcasting.

Examples

>>> np.maximum([2, 3, 4], [1, 5, 2])  # doctest: +SKIP
array([2, 5, 4])
>>> np.maximum(np.eye(2), [0.5, 2]) # broadcasting  # doctest: +SKIP
array([[ 1. ,  2. ],
       [ 0.5,  2. ]])
>>> np.maximum([np.nan, 0, np.nan], [0, np.nan, np.nan])  # doctest: +SKIP
array([ NaN,  NaN,  NaN])
>>> np.maximum(np.Inf, 1)  # doctest: +SKIP
inf
dask.array.mean(a, axis=None, dtype=None, out=None, keepdims=<no value>)

Compute the arithmetic mean along the specified axis.

Returns the average of the array elements. The average is taken over the flattened array by default, otherwise over the specified axis. float64 intermediate and return values are used for integer inputs.

Parameters:
a : array_like

Array containing numbers whose mean is desired. If a is not an array, a conversion is attempted.

axis : None or int or tuple of ints, optional

Axis or axes along which the means are computed. The default is to compute the mean of the flattened array.

New in version 1.7.0.

If this is a tuple of ints, a mean is performed over multiple axes, instead of a single axis or all the axes as before.

dtype : data-type, optional

Type to use in computing the mean. For integer inputs, the default is float64; for floating point inputs, it is the same as the input dtype.

out : ndarray, optional

Alternate output array in which to place the result. The default is None; if provided, it must have the same shape as the expected output, but the type will be cast if necessary. See doc.ufuncs for details.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the input array.

If the default value is passed, then keepdims will not be passed through to the mean method of sub-classes of ndarray, however any non-default value will be. If the sub-class’ method does not implement keepdims any exceptions will be raised.

Returns:
m : ndarray, see dtype parameter above

If out=None, returns a new array containing the mean values, otherwise a reference to the output array is returned.

See also

average
Weighted average

std, var, nanmean, nanstd, nanvar

Notes

The arithmetic mean is the sum of the elements along the axis divided by the number of elements.

Note that for floating-point input, the mean is computed using the same precision the input has. Depending on the input data, this can cause the results to be inaccurate, especially for float32 (see example below). Specifying a higher-precision accumulator using the dtype keyword can alleviate this issue.

By default, float16 results are computed using float32 intermediates for extra precision.

Examples

>>> a = np.array([[1, 2], [3, 4]])
>>> np.mean(a)
2.5
>>> np.mean(a, axis=0)
array([ 2.,  3.])
>>> np.mean(a, axis=1)
array([ 1.5,  3.5])

In single precision, mean can be inaccurate:

>>> a = np.zeros((2, 512*512), dtype=np.float32)
>>> a[0, :] = 1.0
>>> a[1, :] = 0.1
>>> np.mean(a)
0.54999924

Computing the mean in float64 is more accurate:

>>> np.mean(a, dtype=np.float64)
0.55000000074505806
dask.array.meshgrid(*xi, **kwargs)

Return coordinate matrices from coordinate vectors.

This docstring was copied from numpy.meshgrid.

Some inconsistencies with the Dask version may exist.

Make N-D coordinate arrays for vectorized evaluations of N-D scalar/vector fields over N-D grids, given one-dimensional coordinate arrays x1, x2,…, xn.

Changed in version 1.9: 1-D and 0-D cases are allowed.

Parameters:
x1, x2,…, xn : array_like

1-D arrays representing the coordinates of a grid.

indexing : {‘xy’, ‘ij’}, optional

Cartesian (‘xy’, default) or matrix (‘ij’) indexing of output. See Notes for more details.

New in version 1.7.0.

sparse : bool, optional

If True a sparse grid is returned in order to conserve memory. Default is False.

New in version 1.7.0.

copy : bool, optional

If False, a view into the original arrays are returned in order to conserve memory. Default is True. Please note that sparse=False, copy=False will likely return non-contiguous arrays. Furthermore, more than one element of a broadcast array may refer to a single memory location. If you need to write to the arrays, make copies first.

New in version 1.7.0.

Returns:
X1, X2,…, XN : ndarray

For vectors x1, x2,…, ‘xn’ with lengths Ni=len(xi) , return (N1, N2, N3,...Nn) shaped arrays if indexing=’ij’ or (N2, N1, N3,...Nn) shaped arrays if indexing=’xy’ with the elements of xi repeated to fill the matrix along the first dimension for x1, the second for x2 and so on.

See also

index_tricks.mgrid
Construct a multi-dimensional “meshgrid” using indexing notation.
index_tricks.ogrid
Construct an open multi-dimensional “meshgrid” using indexing notation.

Notes

This function supports both indexing conventions through the indexing keyword argument. Giving the string ‘ij’ returns a meshgrid with matrix indexing, while ‘xy’ returns a meshgrid with Cartesian indexing. In the 2-D case with inputs of length M and N, the outputs are of shape (N, M) for ‘xy’ indexing and (M, N) for ‘ij’ indexing. In the 3-D case with inputs of length M, N and P, outputs are of shape (N, M, P) for ‘xy’ indexing and (M, N, P) for ‘ij’ indexing. The difference is illustrated by the following code snippet:

xv, yv = np.meshgrid(x, y, sparse=False, indexing='ij')
for i in range(nx):
    for j in range(ny):
        # treat xv[i,j], yv[i,j]

xv, yv = np.meshgrid(x, y, sparse=False, indexing='xy')
for i in range(nx):
    for j in range(ny):
        # treat xv[j,i], yv[j,i]

In the 1-D and 0-D case, the indexing and sparse keywords have no effect.

Examples

>>> nx, ny = (3, 2)  # doctest: +SKIP
>>> x = np.linspace(0, 1, nx)  # doctest: +SKIP
>>> y = np.linspace(0, 1, ny)  # doctest: +SKIP
>>> xv, yv = np.meshgrid(x, y)  # doctest: +SKIP
>>> xv  # doctest: +SKIP
array([[ 0. ,  0.5,  1. ],
       [ 0. ,  0.5,  1. ]])
>>> yv  # doctest: +SKIP
array([[ 0.,  0.,  0.],
       [ 1.,  1.,  1.]])
>>> xv, yv = np.meshgrid(x, y, sparse=True)  # make sparse output arrays  # doctest: +SKIP
>>> xv  # doctest: +SKIP
array([[ 0. ,  0.5,  1. ]])
>>> yv  # doctest: +SKIP
array([[ 0.],
       [ 1.]])

meshgrid is very useful to evaluate functions on a grid.

>>> import matplotlib.pyplot as plt  # doctest: +SKIP
>>> x = np.arange(-5, 5, 0.1)  # doctest: +SKIP
>>> y = np.arange(-5, 5, 0.1)  # doctest: +SKIP
>>> xx, yy = np.meshgrid(x, y, sparse=True)  # doctest: +SKIP
>>> z = np.sin(xx**2 + yy**2) / (xx**2 + yy**2)  # doctest: +SKIP
>>> h = plt.contourf(x,y,z)  # doctest: +SKIP
>>> plt.show()  # doctest: +SKIP
dask.array.min(a, axis=None, out=None, keepdims=<no value>, initial=<no value>)

Return the minimum of an array or minimum along an axis.

Parameters:
a : array_like

Input data.

axis : None or int or tuple of ints, optional

Axis or axes along which to operate. By default, flattened input is used.

New in version 1.7.0.

If this is a tuple of ints, the minimum is selected over multiple axes, instead of a single axis or all the axes as before.

out : ndarray, optional

Alternative output array in which to place the result. Must be of the same shape and buffer length as the expected output. See doc.ufuncs (Section “Output arguments”) for more details.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the input array.

If the default value is passed, then keepdims will not be passed through to the amin method of sub-classes of ndarray, however any non-default value will be. If the sub-class’ method does not implement keepdims any exceptions will be raised.

initial : scalar, optional

The maximum value of an output element. Must be present to allow computation on empty slice. See ~numpy.ufunc.reduce for details.

New in version 1.15.0.

Returns:
amin : ndarray or scalar

Minimum of a. If axis is None, the result is a scalar value. If axis is given, the result is an array of dimension a.ndim - 1.

See also

amax
The maximum value of an array along a given axis, propagating any NaNs.
nanmin
The minimum value of an array along a given axis, ignoring any NaNs.
minimum
Element-wise minimum of two arrays, propagating any NaNs.
fmin
Element-wise minimum of two arrays, ignoring any NaNs.
argmin
Return the indices of the minimum values.

nanmax, maximum, fmax

Notes

NaN values are propagated, that is if at least one item is NaN, the corresponding min value will be NaN as well. To ignore NaN values (MATLAB behavior), please use nanmin.

Don’t use amin for element-wise comparison of 2 arrays; when a.shape[0] is 2, minimum(a[0], a[1]) is faster than amin(a, axis=0).

Examples

>>> a = np.arange(4).reshape((2,2))
>>> a
array([[0, 1],
       [2, 3]])
>>> np.amin(a)           # Minimum of the flattened array
0
>>> np.amin(a, axis=0)   # Minima along the first axis
array([0, 1])
>>> np.amin(a, axis=1)   # Minima along the second axis
array([0, 2])
>>> b = np.arange(5, dtype=float)
>>> b[2] = np.NaN
>>> np.amin(b)
nan
>>> np.nanmin(b)
0.0
>>> np.min([[-50], [10]], axis=-1, initial=0)
array([-50,   0])

Notice that the initial value is used as one of the elements for which the minimum is determined, unlike for the default argument Python’s max function, which is only used for empty iterables.

Notice that this isn’t the same as Python’s default argument.

>>> np.min([6], initial=5)
5
>>> min([6], default=5)
6
dask.array.minimum(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Element-wise minimum of array elements.

Compare two arrays and returns a new array containing the element-wise minima. If one of the elements being compared is a NaN, then that element is returned. If both elements are NaNs then the first is returned. The latter distinction is important for complex NaNs, which are defined as at least one of the real or imaginary parts being a NaN. The net effect is that NaNs are propagated.

Parameters:
x1, x2 : array_like

The arrays holding the elements to be compared. They must have the same shape, or shapes that can be broadcast to a single shape.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y : ndarray or scalar

The minimum of x1 and x2, element-wise. This is a scalar if both x1 and x2 are scalars.

See also

maximum
Element-wise maximum of two arrays, propagates NaNs.
fmin
Element-wise minimum of two arrays, ignores NaNs.
amin
The minimum value of an array along a given axis, propagates NaNs.
nanmin
The minimum value of an array along a given axis, ignores NaNs.

fmax, amax, nanmax

Notes

The minimum is equivalent to np.where(x1 <= x2, x1, x2) when neither x1 nor x2 are NaNs, but it is faster and does proper broadcasting.

Examples

>>> np.minimum([2, 3, 4], [1, 5, 2])  # doctest: +SKIP
array([1, 3, 2])
>>> np.minimum(np.eye(2), [0.5, 2]) # broadcasting  # doctest: +SKIP
array([[ 0.5,  0. ],
       [ 0. ,  1. ]])
>>> np.minimum([np.nan, 0, np.nan],[0, np.nan, np.nan])  # doctest: +SKIP
array([ NaN,  NaN,  NaN])
>>> np.minimum(-np.Inf, 1)  # doctest: +SKIP
-inf
dask.array.modf(x, [out1, out2, ]/, [out=(None, None), ]*, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Return the fractional and integral parts of an array, element-wise.

The fractional and integral parts are negative if the given number is negative.

Parameters:
x : array_like

Input array.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
y1 : ndarray

Fractional part of x. This is a scalar if x is a scalar.

y2 : ndarray

Integral part of x. This is a scalar if x is a scalar.

See also

divmod
divmod(x, 1) is equivalent to modf with the return values switched, except it always has a positive remainder.

Notes

For integer input the return values are floats.

Examples

>>> np.modf([0, 3.5])  # doctest: +SKIP
(array([ 0. ,  0.5]), array([ 0.,  3.]))
>>> np.modf(-0.5)  # doctest: +SKIP
(-0.5, -0)
dask.array.moment(a, order, axis=None, dtype=None, keepdims=False, ddof=0, split_every=None, out=None)
dask.array.nanargmax(x, axis, **kwargs)
dask.array.nanargmin(x, axis, **kwargs)
dask.array.nancumprod(a, axis=None, dtype=None, out=None)

Return the cumulative product of array elements over a given axis treating Not a Numbers (NaNs) as one. The cumulative product does not change when NaNs are encountered and leading NaNs are replaced by ones.

Ones are returned for slices that are all-NaN or empty.

New in version 1.12.0.

Parameters:
a : array_like

Input array.

axis : int, optional

Axis along which the cumulative product is computed. By default the input is flattened.

dtype : dtype, optional

Type of the returned array, as well as of the accumulator in which the elements are multiplied. If dtype is not specified, it defaults to the dtype of a, unless a has an integer dtype with a precision less than that of the default platform integer. In that case, the default platform integer is used instead.

out : ndarray, optional

Alternative output array in which to place the result. It must have the same shape and buffer length as the expected output but the type of the resulting values will be cast if necessary.

Returns:
nancumprod : ndarray

A new array holding the result is returned unless out is specified, in which case it is returned.

See also

numpy.cumprod
Cumulative product across array propagating NaNs.
isnan
Show which elements are NaN.

Examples

>>> np.nancumprod(1)
array([1])
>>> np.nancumprod([1])
array([1])
>>> np.nancumprod([1, np.nan])
array([ 1.,  1.])
>>> a = np.array([[1, 2], [3, np.nan]])
>>> np.nancumprod(a)
array([ 1.,  2.,  6.,  6.])
>>> np.nancumprod(a, axis=0)
array([[ 1.,  2.],
       [ 3.,  2.]])
>>> np.nancumprod(a, axis=1)
array([[ 1.,  2.],
       [ 3.,  3.]])
dask.array.nancumsum(a, axis=None, dtype=None, out=None)

Return the cumulative sum of array elements over a given axis treating Not a Numbers (NaNs) as zero. The cumulative sum does not change when NaNs are encountered and leading NaNs are replaced by zeros.

Zeros are returned for slices that are all-NaN or empty.

New in version 1.12.0.

Parameters:
a : array_like

Input array.

axis : int, optional

Axis along which the cumulative sum is computed. The default (None) is to compute the cumsum over the flattened array.

dtype : dtype, optional

Type of the returned array and of the accumulator in which the elements are summed. If dtype is not specified, it defaults to the dtype of a, unless a has an integer dtype with a precision less than that of the default platform integer. In that case, the default platform integer is used.

out : ndarray, optional

Alternative output array in which to place the result. It must have the same shape and buffer length as the expected output but the type will be cast if necessary. See doc.ufuncs (Section “Output arguments”) for more details.

Returns:
nancumsum : ndarray.

A new array holding the result is returned unless out is specified, in which it is returned. The result has the same size as a, and the same shape as a if axis is not None or a is a 1-d array.

See also

numpy.cumsum
Cumulative sum across array propagating NaNs.
isnan
Show which elements are NaN.

Examples

>>> np.nancumsum(1)
array([1])
>>> np.nancumsum([1])
array([1])
>>> np.nancumsum([1, np.nan])
array([ 1.,  1.])
>>> a = np.array([[1, 2], [3, np.nan]])
>>> np.nancumsum(a)
array([ 1.,  3.,  6.,  6.])
>>> np.nancumsum(a, axis=0)
array([[ 1.,  2.],
       [ 4.,  2.]])
>>> np.nancumsum(a, axis=1)
array([[ 1.,  3.],
       [ 3.,  3.]])
dask.array.nanmax(a, axis=None, out=None, keepdims=<no value>)

Return the maximum of an array or maximum along an axis, ignoring any NaNs. When all-NaN slices are encountered a RuntimeWarning is raised and NaN is returned for that slice.

Parameters:
a : array_like

Array containing numbers whose maximum is desired. If a is not an array, a conversion is attempted.

axis : {int, tuple of int, None}, optional

Axis or axes along which the maximum is computed. The default is to compute the maximum of the flattened array.

out : ndarray, optional

Alternate output array in which to place the result. The default is None; if provided, it must have the same shape as the expected output, but the type will be cast if necessary. See doc.ufuncs for details.

New in version 1.8.0.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the original a.

If the value is anything but the default, then keepdims will be passed through to the max method of sub-classes of ndarray. If the sub-classes methods does not implement keepdims any exceptions will be raised.

New in version 1.8.0.

Returns:
nanmax : ndarray

An array with the same shape as a, with the specified axis removed. If a is a 0-d array, or if axis is None, an ndarray scalar is returned. The same dtype as a is returned.

See also

nanmin
The minimum value of an array along a given axis, ignoring any NaNs.
amax
The maximum value of an array along a given axis, propagating any NaNs.
fmax
Element-wise maximum of two arrays, ignoring any NaNs.
maximum
Element-wise maximum of two arrays, propagating any NaNs.
isnan
Shows which elements are Not a Number (NaN).
isfinite
Shows which elements are neither NaN nor infinity.

amin, fmin, minimum

Notes

NumPy uses the IEEE Standard for Binary Floating-Point for Arithmetic (IEEE 754). This means that Not a Number is not equivalent to infinity. Positive infinity is treated as a very large number and negative infinity is treated as a very small (i.e. negative) number.

If the input has a integer type the function is equivalent to np.max.

Examples

>>> a = np.array([[1, 2], [3, np.nan]])
>>> np.nanmax(a)
3.0
>>> np.nanmax(a, axis=0)
array([ 3.,  2.])
>>> np.nanmax(a, axis=1)
array([ 2.,  3.])

When positive infinity and negative infinity are present:

>>> np.nanmax([1, 2, np.nan, np.NINF])
2.0
>>> np.nanmax([1, 2, np.nan, np.inf])
inf
dask.array.nanmean(a, axis=None, dtype=None, out=None, keepdims=<no value>)

Compute the arithmetic mean along the specified axis, ignoring NaNs.

Returns the average of the array elements. The average is taken over the flattened array by default, otherwise over the specified axis. float64 intermediate and return values are used for integer inputs.

For all-NaN slices, NaN is returned and a RuntimeWarning is raised.

New in version 1.8.0.

Parameters:
a : array_like

Array containing numbers whose mean is desired. If a is not an array, a conversion is attempted.

axis : {int, tuple of int, None}, optional

Axis or axes along which the means are computed. The default is to compute the mean of the flattened array.

dtype : data-type, optional

Type to use in computing the mean. For integer inputs, the default is float64; for inexact inputs, it is the same as the input dtype.

out : ndarray, optional

Alternate output array in which to place the result. The default is None; if provided, it must have the same shape as the expected output, but the type will be cast if necessary. See doc.ufuncs for details.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the original a.

If the value is anything but the default, then keepdims will be passed through to the mean or sum methods of sub-classes of ndarray. If the sub-classes methods does not implement keepdims any exceptions will be raised.

Returns:
m : ndarray, see dtype parameter above

If out=None, returns a new array containing the mean values, otherwise a reference to the output array is returned. Nan is returned for slices that contain only NaNs.

See also

average
Weighted average
mean
Arithmetic mean taken while not ignoring NaNs

var, nanvar

Notes

The arithmetic mean is the sum of the non-NaN elements along the axis divided by the number of non-NaN elements.

Note that for floating-point input, the mean is computed using the same precision the input has. Depending on the input data, this can cause the results to be inaccurate, especially for float32. Specifying a higher-precision accumulator using the dtype keyword can alleviate this issue.

Examples

>>> a = np.array([[1, np.nan], [3, 4]])
>>> np.nanmean(a)
2.6666666666666665
>>> np.nanmean(a, axis=0)
array([ 2.,  4.])
>>> np.nanmean(a, axis=1)
array([ 1.,  3.5])
dask.array.nanmin(a, axis=None, out=None, keepdims=<no value>)

Return minimum of an array or minimum along an axis, ignoring any NaNs. When all-NaN slices are encountered a RuntimeWarning is raised and Nan is returned for that slice.

Parameters:
a : array_like

Array containing numbers whose minimum is desired. If a is not an array, a conversion is attempted.

axis : {int, tuple of int, None}, optional

Axis or axes along which the minimum is computed. The default is to compute the minimum of the flattened array.

out : ndarray, optional

Alternate output array in which to place the result. The default is None; if provided, it must have the same shape as the expected output, but the type will be cast if necessary. See doc.ufuncs for details.

New in version 1.8.0.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the original a.

If the value is anything but the default, then keepdims will be passed through to the min method of sub-classes of ndarray. If the sub-classes methods does not implement keepdims any exceptions will be raised.

New in version 1.8.0.

Returns:
nanmin : ndarray

An array with the same shape as a, with the specified axis removed. If a is a 0-d array, or if axis is None, an ndarray scalar is returned. The same dtype as a is returned.

See also

nanmax
The maximum value of an array along a given axis, ignoring any NaNs.
amin
The minimum value of an array along a given axis, propagating any NaNs.
fmin
Element-wise minimum of two arrays, ignoring any NaNs.
minimum
Element-wise minimum of two arrays, propagating any NaNs.
isnan
Shows which elements are Not a Number (NaN).
isfinite
Shows which elements are neither NaN nor infinity.

amax, fmax, maximum

Notes

NumPy uses the IEEE Standard for Binary Floating-Point for Arithmetic (IEEE 754). This means that Not a Number is not equivalent to infinity. Positive infinity is treated as a very large number and negative infinity is treated as a very small (i.e. negative) number.

If the input has a integer type the function is equivalent to np.min.

Examples

>>> a = np.array([[1, 2], [3, np.nan]])
>>> np.nanmin(a)
1.0
>>> np.nanmin(a, axis=0)
array([ 1.,  2.])
>>> np.nanmin(a, axis=1)
array([ 1.,  3.])

When positive infinity and negative infinity are present:

>>> np.nanmin([1, 2, np.nan, np.inf])
1.0
>>> np.nanmin([1, 2, np.nan, np.NINF])
-inf
dask.array.nanprod(a, axis=None, dtype=None, out=None, keepdims=<no value>)

Return the product of array elements over a given axis treating Not a Numbers (NaNs) as ones.

One is returned for slices that are all-NaN or empty.

New in version 1.10.0.

Parameters:
a : array_like

Array containing numbers whose product is desired. If a is not an array, a conversion is attempted.

axis : {int, tuple of int, None}, optional

Axis or axes along which the product is computed. The default is to compute the product of the flattened array.

dtype : data-type, optional

The type of the returned array and of the accumulator in which the elements are summed. By default, the dtype of a is used. An exception is when a has an integer type with less precision than the platform (u)intp. In that case, the default will be either (u)int32 or (u)int64 depending on whether the platform is 32 or 64 bits. For inexact inputs, dtype must be inexact.

out : ndarray, optional

Alternate output array in which to place the result. The default is None. If provided, it must have the same shape as the expected output, but the type will be cast if necessary. See doc.ufuncs for details. The casting of NaN to integer can yield unexpected results.

keepdims : bool, optional

If True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the original arr.

Returns:
nanprod : ndarray

A new array holding the result is returned unless out is specified, in which case it is returned.

See also

numpy.prod
Product across array propagating NaNs.
isnan
Show which elements are NaN.

Examples

>>> np.nanprod(1)
1
>>> np.nanprod([1])
1
>>> np.nanprod([1, np.nan])
1.0
>>> a = np.array([[1, 2], [3, np.nan]])
>>> np.nanprod(a)
6.0
>>> np.nanprod(a, axis=0)
array([ 3.,  2.])
dask.array.nanstd(a, axis=None, dtype=None, out=None, ddof=0, keepdims=<no value>)

Compute the standard deviation along the specified axis, while ignoring NaNs.

Returns the standard deviation, a measure of the spread of a distribution, of the non-NaN array elements. The standard deviation is computed for the flattened array by default, otherwise over the specified axis.

For all-NaN slices or slices with zero degrees of freedom, NaN is returned and a RuntimeWarning is raised.

New in version 1.8.0.

Parameters:
a : array_like

Calculate the standard deviation of the non-NaN values.

axis : {int, tuple of int, None}, optional

Axis or axes along which the standard deviation is computed. The default is to compute the standard deviation of the flattened array.

dtype : dtype, optional

Type to use in computing the standard deviation. For arrays of integer type the default is float64, for arrays of float types it is the same as the array type.

out : ndarray, optional

Alternative output array in which to place the result. It must have the same shape as the expected output but the type (of the calculated values) will be cast if necessary.

ddof : int, optional

Means Delta Degrees of Freedom. The divisor used in calculations is N - ddof, where N represents the number of non-NaN elements. By default ddof is zero.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the original a.

If this value is anything but the default it is passed through as-is to the relevant functions of the sub-classes. If these functions do not have a keepdims kwarg, a RuntimeError will be raised.

Returns:
standard_deviation : ndarray, see dtype parameter above.

If out is None, return a new array containing the standard deviation, otherwise return a reference to the output array. If ddof is >= the number of non-NaN elements in a slice or the slice contains only NaNs, then the result for that slice is NaN.

See also

var, mean, std, nanvar, nanmean

numpy.doc.ufuncs
Section “Output arguments”

Notes

The standard deviation is the square root of the average of the squared deviations from the mean: std = sqrt(mean(abs(x - x.mean())**2)).

The average squared deviation is normally calculated as x.sum() / N, where N = len(x). If, however, ddof is specified, the divisor N - ddof is used instead. In standard statistical practice, ddof=1 provides an unbiased estimator of the variance of the infinite population. ddof=0 provides a maximum likelihood estimate of the variance for normally distributed variables. The standard deviation computed in this function is the square root of the estimated variance, so even with ddof=1, it will not be an unbiased estimate of the standard deviation per se.

Note that, for complex numbers, std takes the absolute value before squaring, so that the result is always real and nonnegative.

For floating-point input, the std is computed using the same precision the input has. Depending on the input data, this can cause the results to be inaccurate, especially for float32 (see example below). Specifying a higher-accuracy accumulator using the dtype keyword can alleviate this issue.

Examples

>>> a = np.array([[1, np.nan], [3, 4]])
>>> np.nanstd(a)
1.247219128924647
>>> np.nanstd(a, axis=0)
array([ 1.,  0.])
>>> np.nanstd(a, axis=1)
array([ 0.,  0.5])
dask.array.nansum(a, axis=None, dtype=None, out=None, keepdims=<no value>)

Return the sum of array elements over a given axis treating Not a Numbers (NaNs) as zero.

In NumPy versions <= 1.9.0 Nan is returned for slices that are all-NaN or empty. In later versions zero is returned.

Parameters:
a : array_like

Array containing numbers whose sum is desired. If a is not an array, a conversion is attempted.

axis : {int, tuple of int, None}, optional

Axis or axes along which the sum is computed. The default is to compute the sum of the flattened array.

dtype : data-type, optional

The type of the returned array and of the accumulator in which the elements are summed. By default, the dtype of a is used. An exception is when a has an integer type with less precision than the platform (u)intp. In that case, the default will be either (u)int32 or (u)int64 depending on whether the platform is 32 or 64 bits. For inexact inputs, dtype must be inexact.

New in version 1.8.0.

out : ndarray, optional

Alternate output array in which to place the result. The default is None. If provided, it must have the same shape as the expected output, but the type will be cast if necessary. See doc.ufuncs for details. The casting of NaN to integer can yield unexpected results.

New in version 1.8.0.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the original a.

If the value is anything but the default, then keepdims will be passed through to the mean or sum methods of sub-classes of ndarray. If the sub-classes methods does not implement keepdims any exceptions will be raised.

New in version 1.8.0.

Returns:
nansum : ndarray.

A new array holding the result is returned unless out is specified, in which it is returned. The result has the same size as a, and the same shape as a if axis is not None or a is a 1-d array.

See also

numpy.sum
Sum across array propagating NaNs.
isnan
Show which elements are NaN.
isfinite
Show which elements are not NaN or +/-inf.

Notes

If both positive and negative infinity are present, the sum will be Not A Number (NaN).

Examples

>>> np.nansum(1)
1
>>> np.nansum([1])
1
>>> np.nansum([1, np.nan])
1.0
>>> a = np.array([[1, 1], [1, np.nan]])
>>> np.nansum(a)
3.0
>>> np.nansum(a, axis=0)
array([ 2.,  1.])
>>> np.nansum([1, np.nan, np.inf])
inf
>>> np.nansum([1, np.nan, np.NINF])
-inf
>>> np.nansum([1, np.nan, np.inf, -np.inf]) # both +/- infinity present
nan
dask.array.nanvar(a, axis=None, dtype=None, out=None, ddof=0, keepdims=<no value>)

Compute the variance along the specified axis, while ignoring NaNs.

Returns the variance of the array elements, a measure of the spread of a distribution. The variance is computed for the flattened array by default, otherwise over the specified axis.

For all-NaN slices or slices with zero degrees of freedom, NaN is returned and a RuntimeWarning is raised.

New in version 1.8.0.

Parameters:
a : array_like

Array containing numbers whose variance is desired. If a is not an array, a conversion is attempted.

axis : {int, tuple of int, None}, optional

Axis or axes along which the variance is computed. The default is to compute the variance of the flattened array.

dtype : data-type, optional

Type to use in computing the variance. For arrays of integer type the default is float32; for arrays of float types it is the same as the array type.

out : ndarray, optional

Alternate output array in which to place the result. It must have the same shape as the expected output, but the type is cast if necessary.

ddof : int, optional

“Delta Degrees of Freedom”: the divisor used in the calculation is N - ddof, where N represents the number of non-NaN elements. By default ddof is zero.

keepdims : bool, optional

If this is set to True, the axes which are reduced are left in the result as dimensions with size one. With this option, the result will broadcast correctly against the original a.

Returns:
variance : ndarray, see dtype parameter above

If out is None, return a new array containing the variance, otherwise return a reference to the output array. If ddof is >= the number of non-NaN elements in a slice or the slice contains only NaNs, then the result for that slice is NaN.

See also

std
Standard deviation
mean
Average
var
Variance while not ignoring NaNs

nanstd, nanmean

numpy.doc.ufuncs
Section “Output arguments”

Notes

The variance is the average of the squared deviations from the mean, i.e., var = mean(abs(x - x.mean())**2).

The mean is normally calculated as x.sum() / N, where N = len(x). If, however, ddof is specified, the divisor N - ddof is used instead. In standard statistical practice, ddof=1 provides an unbiased estimator of the variance of a hypothetical infinite population. ddof=0 provides a maximum likelihood estimate of the variance for normally distributed variables.

Note that for complex numbers, the absolute value is taken before squaring, so that the result is always real and nonnegative.

For floating-point input, the variance is computed using the same precision the input has. Depending on the input data, this can cause the results to be inaccurate, especially for float32 (see example below). Specifying a higher-accuracy accumulator using the dtype keyword can alleviate this issue.

For this function to work on sub-classes of ndarray, they must define sum with the kwarg keepdims

Examples

>>> a = np.array([[1, np.nan], [3, 4]])
>>> np.var(a)
1.5555555555555554
>>> np.nanvar(a, axis=0)
array([ 1.,  0.])
>>> np.nanvar(a, axis=1)
array([ 0.,  0.25])
dask.array.nan_to_num(*args, **kwargs)

Replace NaN with zero and infinity with large finite numbers.

If x is inexact, NaN is replaced by zero, and infinity and -infinity replaced by the respectively largest and most negative finite floating point values representable by x.dtype.

For complex dtypes, the above is applied to each of the real and imaginary components of x separately.

If x is not inexact, then no replacements are made.

Parameters:
x : scalar or array_like

Input data.

copy : bool, optional

Whether to create a copy of x (True) or to replace values in-place (False). The in-place operation only occurs if casting to an array does not require a copy. Default is True.

New in version 1.13.

Returns:
out : ndarray

x, with the non-finite values replaced. If copy is False, this may be x itself.

See also

isinf
Shows which elements are positive or negative infinity.
isneginf
Shows which elements are negative infinity.
isposinf
Shows which elements are positive infinity.
isnan
Shows which elements are Not a Number (NaN).
isfinite
Shows which elements are finite (not NaN, not infinity)

Notes

NumPy uses the IEEE Standard for Binary Floating-Point for Arithmetic (IEEE 754). This means that Not a Number is not equivalent to infinity.

Examples

>>> np.nan_to_num(np.inf)  # doctest: +SKIP
1.7976931348623157e+308
>>> np.nan_to_num(-np.inf)  # doctest: +SKIP
-1.7976931348623157e+308
>>> np.nan_to_num(np.nan)  # doctest: +SKIP
0.0
>>> x = np.array([np.inf, -np.inf, np.nan, -128, 128])  # doctest: +SKIP
>>> np.nan_to_num(x)  # doctest: +SKIP
array([  1.79769313e+308,  -1.79769313e+308,   0.00000000e+000,
        -1.28000000e+002,   1.28000000e+002])
>>> y = np.array([complex(np.inf, np.nan), np.nan, complex(np.nan, np.inf)])  # doctest: +SKIP
>>> np.nan_to_num(y)  # doctest: +SKIP
array([  1.79769313e+308 +0.00000000e+000j,
         0.00000000e+000 +0.00000000e+000j,
         0.00000000e+000 +1.79769313e+308j])
dask.array.nextafter(x1, x2, /, out=None, *, where=True, casting='same_kind', order='K', dtype=None, subok=True[, signature, extobj])

Return the next floating-point value after x1 towards x2, element-wise.

Parameters:
x1 : array_like

Values to find the next representable value of.

x2 : array_like

The direction where to look for the next representable value of x1.

out : ndarray, None, or tuple of ndarray and None, optional

A location into which the result is stored. If provided, it must have a shape that the inputs broadcast to. If not provided or None, a freshly-allocated array is returned. A tuple (possible only as a keyword argument) must have length equal to the number of outputs.

where : array_like, optional

Values of True indicate to calculate the ufunc at that position, values of False indicate to leave the value in the output alone.

**kwargs

For other keyword-only arguments, see the ufunc docs.

Returns:
out : ndarray or scalar

The next representable values of x1 in the direction of x2. This is a scalar if both x1 and x2 are scalars.

Examples

>>> eps = np.finfo(np.float64).eps  # doctest: +SKIP
>>> np.nextafter(1, 2) == eps + 1  # doctest: +SKIP
True
>>> np.nextafter([1, 2], [2, 1]) == [eps + 1, 2 - eps]  # doctest: +SKIP
array([ True,  True])
dask.array.nonzero(a)

Return the indices of the elements that are non-zero.

This docstring was copied from numpy.nonzero.

Some inconsistencies with the Dask version may exist.

Returns a tuple of arrays, one for each dimension of a, containing the indices of the non-zero elements in that dimension. The values in a are always tested and returned in row-major, C-style order. The corresponding non-zero values can be obtained with:

a[nonzero(a)]

To group the indices by element, rather than dimension, use:

transpose(nonzero(a))

The result of this is always a 2-D array, with a row for each non-zero element.

Parameters:
a : array_like

Input array.

Returns:
tuple_of_arrays : tuple

Indices of elements that are non-zero.

See also

flatnonzero
Return indices that are non-zero in the flattened version of the input array.
ndarray.nonzero
Equivalent ndarray method.
count_nonzero
Counts the number of non-zero elements in the input array.

Examples

>>> x = np.array([[3, 0, 0], [0, 4, 0], [5, 6, 0]])  # doctest: +SKIP
>>> x  # doctest: +SKIP
array([[3, 0, 0],
       [0, 4, 0],
       [5, 6, 0]])
>>> np.nonzero(x)  # doctest: +SKIP
(array([0, 1, 2, 2]), array([0, 1, 0, 1]))
>>> x[np.nonzero(x)]  # doctest: +SKIP
array([3, 4, 5, 6])
>>> np.transpose(np.nonzero(x))  # doctest: +SKIP
array([[0, 0],
       [1, 1],
       [2, 0],
       [2, 1])

A common use for nonzero is to find the indices of an array, where a condition is True. Given an array a, the condition a > 3 is a boolean array and since False is interpreted as 0, np.nonzero(a > 3) yields the indices of the a where the condition is true.

>>> a = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])  # doctest: +SKIP
>>> a > 3  # doctest: +SKIP
array([[False, False, False],
       [ True,  True,  True],
       [ True,  True,  True]])
>>> np.nonzero(a > 3)  # doctest: +SKIP
(array([1, 1, 1, 2, 2, 2]), array([0, 1, 2, 0, 1, 2]))

Using this result to index a is equivalent to using the mask directly:

>>> a[np.nonzero(a > 3)]  # doctest: +SKIP
array([4, 5, 6, 7, 8, 9])
>>> a[a > 3]  # prefer this spelling  # doctest: +SKIP
array([4, 5, 6, 7, 8, 9])

nonzero can also be called as a method of the array.

>>> (a > 3).nonzero()  # doctest: +SKIP
(array([1, 1, 1, 2, 2, 2]), array([0, 1, 2, 0, 1, 2]))
dask.array.notnull(values)

pandas.notnull for dask arrays

dask.array.ones(*args, **kwargs)

Blocked variant of ones

Follows the signature of ones exactly except that it also requires a keyword argument chunks=(…)

Original signature follows below.

Return a new array of given shape and type, filled with ones.

Parameters:
shape : int or sequence of ints

Shape of the new array, e.g., (2, 3) or 2.

dtype : data-type, optional

The desired data-type for the array, e.g., numpy.int8. Default is numpy.float64.

order : {‘C’, ‘F’}, optional, default: C

Whether to store multi-dimensional data in row-major (C-style) or column-major (Fortran-style) order in memory.

Returns:
out : ndarray

Array of ones with the given shape, dtype, and order.

See also

ones_like
Return an array of ones with shape and type of input.
empty
Return a new uninitialized array.
zeros
Return a new array setting values to zero.
full
Return a new array of given shape filled with value.

Examples

>>> np.ones(5)
array([ 1.,  1.,  1.,  1.,  1.])
>>> np.ones((5,), dtype=int)
array([1, 1, 1, 1, 1])
>>> np.ones((2, 1))
array([[ 1.],
       [ 1.]])
>>> s = (2,2)
>>> np.ones(s)
array([[ 1.,  1.],
       [ 1.,  1.]])
dask.array.ones_like(a, dtype=None, chunks=None)

Return an array of ones with the same shape and type as a given array.

Parameters:
a : array_like

The shape and data-type of a define these same attributes of the returned array.

dtype : data-type, optional

Overrides the data type of the result.

chunks : sequence of ints

The number of samples on each block. Note that the last block will have fewer samples if len(array) % chunks != 0.

Returns:
out : ndarray

Array of ones with the same shape and type as a.

See also

zeros_like
Return an array of zeros with shape and type of input.
empty_like
Return an empty array with shape and type of input.
zeros
Return a new array setting values to zero.
ones
Return a new array setting values to one.
empty
Return a new uninitialized array.
dask.array.outer(a, b)

Compute the outer product of two vectors.

This docstring was copied from numpy.outer.

Some inconsistencies with the Dask version may exist.

Given two vectors, a = [a0, a1, ..., aM] and b = [b0, b1, ..., bN], the outer product [1] is:

[[a0*b0  a0*b1 ... a0*bN ]
 [a1*b0    .
 [ ...          .
 [aM*b0            aM*bN ]]
Parameters:
a : (M,) array_like

First input vector. Input is flattened if not already 1-dimensional.

b : (N,) array_like

Second input vector. Input is flattened if not already 1-dimensional.

out : (M, N) ndarray, optional

A location where the result is stored

New in version 1.9.0.

Returns:
out : (M, N) ndarray

out[i, j] = a[i] * b[j]

See also

inner

einsum
einsum('i,j->ij', a.ravel(), b.ravel()) is the equivalent.
ufunc.outer
A generalization to N dimensions and other operations. np.multiply.outer(a.ravel(), b.ravel()) is the equivalent.

References

[1](1, 2) : G. H. Golub and C. F. Van Loan, Matrix Computations, 3rd ed., Baltimore, MD, Johns Hopkins University Press, 1996, pg. 8.

Examples

Make a (very coarse) grid for computing a Mandelbrot set:

>>> rl = np.outer(np.ones((5,)), np.linspace(-2, 2, 5))  # doctest: +SKIP
>>> rl  # doctest: +SKIP
array([[-2., -1.,  0.,  1.,  2.],
       [-2., -1.,  0.,  1.,  2.],
       [-2., -1.,  0.,  1.,  2.],
       [-2., -1.,  0.,  1.,  2.],
       [-2., -1.,  0.,  1.,  2.]])
>>> im = np.outer(1j*np.linspace(2, -2, 5), np.ones((5,)))  # doctest: +SKIP
>>> im  # doctest: +SKIP
array([[ 0.+2.j,  0.+2.j,  0.+2.j,  0.+2.j,  0.+2.j],
       [ 0.+1.j,  0.+1.j,  0.+1.j,  0.+1.j,  0.+1.j],
       [ 0.+0.j,  0.+0.j,  0.+0.j,  0.+0.j,  0.+0.j],
       [ 0.-1.j,  0.-1.j,  0.-1.j,  0.-1.j,  0.-1.j],
       [ 0.-2.j,  0.-2.j,  0.-2.j,  0.-2.j,  0.-2.j]])
>>> grid = rl + im  # doctest: +SKIP
>>> grid  # doctest: +SKIP
array([[-2.+2.j, -1.+2.j,  0.+2.j,  1.+2.j,  2.+2.j],
       [-2.+1.j, -1.+1.j,  0.+1.j,  1.+1.j,  2.+1.j],
       [-2.+0.j, -1.+0.j,  0.+0.j,  1.+0.j,  2.+0.j],
       [-2.-1.j, -1.-1.j,  0.-1.j,  1.-1.j,  2.-1.j],
       [-2.-2.j, -1.-2.j,  0.-2.j,  1.-2.j,  2.-2.j]])

An example using a “vector” of letters:

>>> x = np.array(['a', 'b', 'c'], dtype=object)  # doctest: +SKIP
>>> np.outer(x, [1, 2, 3])  # doctest: +SKIP
array([[a, aa, aaa],
       [b, bb, bbb],
       [c, cc, ccc]], dtype=object)
dask.array.pad(array, pad_width, mode, **kwargs)

Pads an array.

This docstring was copied from numpy.pad.

Some inconsistencies with the Dask version may exist.

Parameters:
array : array_like of rank N

Input array

pad_width : {sequence, array_like, int}

Number of values padded to the edges of each axis. ((before_1, after_1), … (before_N, after_N)) unique pad widths for each axis. ((before, after),) yields same before and after pad for each axis. (pad,) or int is a shortcut for before = after = pad width for all axes.

mode : str or function

One of the following string values or a user supplied function.

‘constant’

Pads with a constant value.

‘edge’

Pads with the edge values of array.

‘linear_ramp’

Pads with the linear ramp between end_value and the array edge value.

‘maximum’

Pads with the maximum value of all or part of the vector along each axis.

‘mean’

Pads with the mean value of all or part of the vector along each axis.

‘median’

Pads with the median value of all or part of the vector along each axis.

‘minimum’

Pads with the minimum value of all or part of the vector along each axis.

‘reflect’

Pads with the reflection of the vector mirrored on the first and last values of the vector along each axis.

‘symmetric’

Pads with the reflection of the vector mirrored along the edge of the array.

‘wrap’

Pads with the wrap of the vector along the axis. The first values are used to pad the end and the end values are used to pad the beginning.

<function>

Padding function, see Notes.

stat_length : sequence or int, optional

Used in ‘maximum’, ‘mean’, ‘median’, and ‘minimum’. Number of values at edge of each axis used to calculate the statistic value.

((before_1, after_1), … (before_N, after_N)) unique statistic lengths for each axis.

((before, after),) yields same before and after statistic lengths for each axis.

(stat_length,) or int is a shortcut for before = after = statistic length for all axes.

Default is None, to use the entire axis.

constant_values : sequence or int, optional

Used in ‘constant’. The values to set the padded values for each axis.

((before_1, after_1), … (before_N, after_N)) unique pad constants for each axis.

((before, after),) yields same before and after constants for each axis.

(constant,) or int is a shortcut for before = after = constant for all axes.

Default is 0.

end_values : sequence or int, optional

Used in ‘linear_ramp’. The values used for the ending value of the linear_ramp and that will form the edge of the padded array.

((before_1, after_1), … (before_N, after_N)) unique end values for each axis.

((before, after),) yields same before and after end values for each axis.

(constant,) or int is a shortcut for before = after = end value for all axes.

Default is 0.

reflect_type : {‘even’, ‘odd’}, optional

Used in ‘reflect’, and ‘symmetric’. The ‘even’ style is the default with an unaltered reflection around the edge value. For the ‘odd’ style, the extended part of the array is created by subtracting the reflected values from two times the edge value.

Returns:
pad : ndarray

Padded array of rank equal to array with shape increased according to pad_width.

Notes

New in version 1.7.0.

For an array with rank greater than 1, some of the padding of later axes is calculated from padding of previous axes. This is easiest to think about with a rank 2 array where the corners of the padded array are calculated by using padded values from the first axis.

The padding function, if used, should return a rank 1 array equal in length to the vector argument with padded values replaced. It has the following signature:

padding_func(vector, iaxis_pad_width, iaxis, kwargs)

where

vector : ndarray
A rank 1 array already padded with zeros. Padded values are vector[:pad_tuple[0]] and vector[-pad_tuple[1]:].
iaxis_pad_width : tuple
A 2-tuple of ints, iaxis_pad_width[0] represents the number of values padded at the beginning of vector where iaxis_pad_width[1] represents the number of values padded at the end of vector.
iaxis : int
The axis currently being calculated.
kwargs : dict
Any keyword arguments the function requires.

Examples

>>> a = [1, 2, 3, 4, 5]  # doctest: +SKIP
>>> np.pad(a, (2,3), 'constant', constant_values=(4, 6))  # doctest: +SKIP
array([4, 4, 1, 2, 3, 4, 5, 6, 6, 6])
>>> np.pad(a, (2, 3), 'edge')  # doctest: +SKIP
array([1, 1, 1, 2, 3, 4, 5, 5, 5, 5])
>>> np.pad(a, (2, 3), 'linear_ramp', end_values=(5, -4))  # doctest: +SKIP
array([ 5,  3,  1,  2,  3,  4,  5,  2, -1, -4])
>>> np.pad(a, (2,), 'maximum')  # doctest: +SKIP
array([5, 5, 1, 2, 3, 4, 5, 5, 5])
>>> np.pad(a, (2,), 'mean')  # doctest: +SKIP
array([3, 3, 1, 2, 3, 4, 5, 3, 3])
>>> np.pad(a, (2,), 'median')  # doctest: +SKIP
array([3, 3, 1, 2, 3, 4, 5, 3, 3])
>>> a = [[1, 2], [3, 4]]  # doctest: +SKIP
>>> np.pad(a, ((3, 2), (2, 3)), 'minimum')  # doctest: +SKIP
array([[1, 1, 1, 2, 1, 1, 1],
       [1, 1, 1, 2, 1, 1, 1],
       [1, 1, 1, 2, 1, 1, 1],
       [1, 1, 1, 2, 1, 1, 1],
       [3, 3, 3, 4, 3, 3, 3],
       [1, 1, 1, 2, 1, 1, 1],
       [1, 1, 1, 2, 1, 1, 1]])
>>> a = [1, 2, 3, 4, 5]  # doctest: +SKIP
>>> np.pad(a, (2, 3), 'reflect')  # doctest: +SKIP
array([3, 2, 1, 2, 3, 4, 5, 4, 3, 2])
>>> np.pad(a, (2,