Array Operations¶

Array creation routines¶

`chainerx.empty`	Returns an array without initializing the elements.
`chainerx.empty_like`	Returns a new array with same shape and dtype of a given array.
`chainerx.eye`	Returns a 2-D array with ones on the diagonals and zeros elsewhere.
`chainerx.identity`	Returns a 2-D identity array.
`chainerx.ones`	Returns a new array of given shape and dtype, filled with ones.
`chainerx.ones_like`	Returns an array of ones with same shape and dtype as a given array.
`chainerx.zeros`	Returns a new array of given shape and dtype, filled with zeros.
`chainerx.zeros_like`	Returns an array of zeros with same shape and dtype as a given array.
`chainerx.full`	Returns a new array of given shape and dtype, filled with a given value.
`chainerx.full_like`	Returns a full array with same shape and dtype as a given array.
`chainerx.array`	Creates an array.
`chainerx.asarray`	Converts an object to an array.
`chainerx.asanyarray`	Converts an object to an array.
`chainerx.ascontiguousarray`	Returns a C-contiguous array.
`chainerx.copy`	Creates a copy of a given array.
`chainerx.frombuffer`	Returns a 1-D array interpretation of a buffer.
`chainerx.fromfile`	Constructs an array from data in a text or binary file.
`chainerx.fromfunction`	Constructs an array by executing a function over each coordinate.
`chainerx.fromiter`	Constructs a new 1-D array from an iterable object.
`chainerx.fromstring`	Constructs a new 1-D array initialized from text data in a string.
`chainerx.loadtxt`	Constructs an array by loading data from a text file.
`chainerx.arange`	Returns an array with evenly spaced values within a given interval.
`chainerx.linspace`	Returns an array with evenly spaced numbers over a specified interval.
`chainerx.diag`	Returns a diagonal or a diagonal array.
`chainerx.diagflat`	Creates a diagonal array from the flattened input.

Activation functions¶

`chainerx.log_softmax`	The log of the softmax of input array.
`chainerx.tanh`	Hyperbolic tangent, element-wise
`chainerx.relu`	Rectified Linear Unit function.
`chainerx.sigmoid`

Array manipulation routines¶

`chainerx.reshape`	Returns a reshaped array.
`chainerx.ravel`	Returns a flattened array.
`chainerx.transpose`	Permutes the dimensions of an array.
`chainerx.broadcast_to`	Broadcasts an array to a given shape.
`chainerx.squeeze`	Removes size-one axes from the shape of an array.
`chainerx.asarray`	Converts an object to an array.
`chainerx.ascontiguousarray`	Returns a C-contiguous array.
`chainerx.concatenate`	Joins arrays along an axis.
`chainerx.stack`	Stacks arrays along a new axis.
`chainerx.split`	Splits an array into multiple sub arrays along a given axis.

Indexing routines¶

chainerx.take

Takes elements from an array along an axis.

Linear algebra¶

chainerx.dot

Returns a dot product of two arrays.

Logic functions¶

`chainerx.isinf`	Test element-wise for positive or negative infinity.
`chainerx.isnan`	Test element-wise for NaN and return result as a boolean array.
`chainerx.logical_and`	Returns an array of x1 AND x2 element-wise.
`chainerx.logical_or`	Returns an array of x1 OR x2 element-wise.
`chainerx.logical_not`	Returns an array of NOT x element-wise.
`chainerx.greater`	Returns an array of (x1 > x2) element-wise.
`chainerx.greater_equal`	Returns an array of (x1 >= x2) element-wise.
`chainerx.less`	Returns an array of (x1 < x2) element-wise.
`chainerx.less_equal`	Returns an array of (x1 <= x2) element-wise.
`chainerx.equal`	Returns an array of (x1 == x2) element-wise.
`chainerx.not_equal`	Returns an array of (x1 != x2) element-wise.

Mathematical functions¶

`chainerx.negative`	Numerical negative, element-wise.
`chainerx.add`	Add arguments, element-wise.
`chainerx.subtract`	Subtract arguments, element-wise.
`chainerx.multiply`	Multiply arguments, element-wise.
`chainerx.divide`	Divide arguments, element-wise.
`chainerx.sum`	Sum of array elements over a given axis.
`chainerx.maximum`	Maximum arguments, element-wise.
`chainerx.exp`	Numerical exponential, element-wise.
`chainerx.log`	Natural logarithm, element-wise.
`chainerx.logsumexp`	The log of the sum of exponentials of input array.
`chainerx.log_softmax`	The log of the softmax of input array.
`chainerx.sqrt`	Non-negative square-root, element-wise
`chainerx.sin`	Sine, element-wise
`chainerx.cos`	Cosine, element-wise
`chainerx.tan`	Tangent, element-wise
`chainerx.arcsin`	Inverse sine, element-wise
`chainerx.arccos`	Trigonometric inverse cosine, element-wise
`chainerx.arctan`	Trigonometric inverse tangent, element-wise
`chainerx.tanh`	Hyperbolic tangent, element-wise
`chainerx.square`	Returns the element-wise square of the input.
`chainerx.clip`	Clips the values of an array to a given interval.
`chainerx.ceil`	Return the ceiling of the input, element-wise..

Random sampling¶

`chainerx.random.normal`	Draws random samples from a normal (Gaussian) distribution.
`chainerx.random.uniform`	Draws samples from a uniform distribution.

Sorting, searching, and counting¶

chainerx.argmax

Returns the indices of the maximum along an axis.

Statistics¶

chainerx.amax

Returns the maximum of an array or the maximum along an axis.

Connection¶

`chainerx.conv`	N-dimensional convolution.
`chainerx.conv_transpose`	N-dimensional transposed convolution.
`chainerx.linear`	Linear function, or affine transformation.

Normalization¶

`chainerx.batch_norm`	Batch normalization function.
`chainerx.fixed_batch_norm`	Batch normalization function with fixed statistics.

Pooling¶

`chainerx.max_pool`	Spatial max pooling function.
`chainerx.average_pool`	Spatial average pooling function.