bernoulli¶

paddle. bernoulli ( x: Tensor, p: float | None = None, name: str | None = None ) → Tensor [source]

For each element \(x_i\) in input x, take a sample from the Bernoulli distribution, also called two-point distribution, with success probability \(x_i\). The Bernoulli distribution with success probability \(x_i\) is a discrete probability distribution with probability mass function

\[\begin{split}p(y)=\begin{cases} x_i,&y=1\\ 1-x_i,&y=0 \end{cases}.\end{split}\]

Parameters

x (Tensor) – The input Tensor, it’s data type should be float32, float64.
p (float|None, optional) – If p is given, the success probability will always be p. Default is None, which means to use the success probability specified by input x.
name (str|None, optional) – For details, please refer to Name. Generally, no setting is required. Default: None.

Returns

Tensor, A Tensor filled samples from Bernoulli distribution, whose shape and dtype are same as x.

Examples

>>> import paddle

>>> paddle.set_device('cpu')  # on CPU device
>>> paddle.seed(100)

>>> x = paddle.rand([2,3])
>>> print(x)
>>> 
Tensor(shape=[2, 3], dtype=float32, place=Place(cpu), stop_gradient=True,
[[0.55355281, 0.20714243, 0.01162981],
 [0.51577556, 0.36369765, 0.26091650]])
>>> 

>>> out = paddle.bernoulli(x)
>>> print(out)
>>> 
Tensor(shape=[2, 3], dtype=float32, place=Place(cpu), stop_gradient=True,
[[1., 0., 1.],
 [0., 1., 0.]])
>>> 

>>> out = paddle.bernoulli(x, p=0)
>>> print(out)
Tensor(shape=[2, 3], dtype=float32, place=Place(cpu), stop_gradient=True,
[[0., 0., 0.],
 [0., 0., 0.]])