mygrad.nnet.activations.tanh#

class mygrad.nnet.activations.tanh(x: ArrayLike, out: Tensor | ndarray | None = None, *, where: Mask = True, dtype: DTypeLikeReals = None, constant: bool | None = None)#

Hyperbolic tangent, element-wise.

This docstring was adapted from that of numpy.tanh [3]

Parameters:

xArrayLike

Input tensor.

outOptional[Union[Tensor, ndarray]]

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 tensor is returned.

constantOptional[bool]

If True, this tensor is treated as a constant, and thus does not facilitate back propagation (i.e. constant.grad will always return None).

Defaults to False for float-type data. Defaults to True for integer-type data.

Integer-type tensors must be constant.

whereMask

This condition is broadcast over the input. At locations where the condition is True, the out tensor will be set to the ufunc result. Elsewhere, the out tensor will retain its original value. Note that if an uninitialized out tensor is created via the default out=None, locations within it where the condition is False will remain uninitialized.

dtypeOptional[DTypeLikeReals]

The dtype of the resulting tensor.

Returns:

yTensor: The corresponding hyperbolic tangent values.

References

[1]

M. Abramowitz and I. A. Stegun, Handbook of Mathematical Functions. New York, NY: Dover, 1972, pg. 83. http://www.math.sfu.ca/~cbm/aands/

[2]

Wikipedia, “Hyperbolic function”, https://en.wikipedia.org/wiki/Hyperbolic_function

[3]

Retrieved from https://numpy.org/doc/stable/reference/generated/numpy.tanh.html

Examples

>>> import mygrad as mg
>>> x = mg.linspace(-2, 2, 10)
>>> y = mg.tanh(x); y
Tensor([-0.96402758, -0.9146975 , -0.8044548 , -0.58278295, -0.21863508,
     0.21863508,  0.58278295,  0.8044548 ,  0.9146975 ,  0.96402758])

>>> y.backward()  # compute d(tanh)/dx
>>> x.grad
array([0.07065082, 0.16332849, 0.35285247, 0.66036404, 0.9521987 ,
       0.9521987 , 0.66036404, 0.35285247, 0.16332849, 0.07065082])

(Source code, png, hires.png, pdf)

../_images/mygrad-nnet-activations-tanh-1.png

Attributes:

identity
signature

Methods

`accumulate`([axis, dtype, out, constant])	Not implemented
`at`(indices[, b, constant])	Not implemented
`outer`(b, *[, dtype, out])	Not Implemented
`reduce`([axis, dtype, out, keepdims, ...])	Not Implemented
`reduceat`(indices[, axis, dtype, out])	Not Implemented
`resolve_dtypes`(dtypes, *[, signature, ...])	Find the dtypes NumPy will use for the operation.

__init__(*args, **kwargs)#

Methods

`__init__`(args, *kwargs)
`accumulate`([axis, dtype, out, constant])	Not implemented
`at`(indices[, b, constant])	Not implemented
`outer`(b, *[, dtype, out])	Not Implemented
`reduce`([axis, dtype, out, keepdims, ...])	Not Implemented
`reduceat`(indices[, axis, dtype, out])	Not Implemented
`resolve_dtypes`(dtypes, *[, signature, ...])	Find the dtypes NumPy will use for the operation.

Attributes

`identity`
`nargs`
`nin`
`nout`
`ntypes`
`signature`
`types`