masked_fill¶

paddle. masked_fill ( x, mask, value, name=None ) [source]

Fills elements of self tensor with value where mask is True. The shape of mask must be broadcastable with the shape of the underlying tensor.

Parameters

• x (Tensor) – The Destination Tensor. Supported data types are float, double, int, int64_t,float16 and bfloat16.

• mask (Tensor) – The boolean tensor indicate the position to be filled. The data type of mask must be bool.

• value (Scalar or 0-D Tensor) – The value used to fill the target tensor. Supported data types are float, double, int, int64_t,float16 and bfloat16.

• name (str, optional) – The default value is None. Normally there is no need for user to set this property. For more information, please refer to Name.

Returns

Tensor, same dimension and dtype with x.

Examples

>>> 
>>> import paddle
>>> x = paddle.ones((3, 3), dtype="float32")
>>> mask = paddle.to_tensor([[True, True, False]])
>>> print(mask)
Tensor(shape=[1, 3], dtype=bool, place=Place(gpu:0), stop_gradient=True,
       [[True , True , False]])
>>> out = paddle.masked_fill(x, mask, 2)
>>> print(out)
Tensor(shape=[3, 3], dtype=float32, place=Place(gpu:0), stop_gradient=True,
       [[2., 2., 1.],
        [2., 2., 1.],
        [2., 2., 1.]])