tf.math.xlog1py

View source on GitHub

Compute x * log1p(y).

tf.math.xlog1py(
    x, y, name=None
)

Given x and y, compute x * log1p(y). This function safely returns zero when x = 0, no matter what the value of y is.

Example:

tf.math.xlog1py(0., 1.) 
<tf.Tensor: shape=(), dtype=float32, numpy=0.> 
tf.math.xlog1py(1., 1.) 
<tf.Tensor: shape=(), dtype=float32, numpy=0.6931472> 
tf.math.xlog1py(2., 2.) 
<tf.Tensor: shape=(), dtype=float32, numpy=2.1972246> 
tf.math.xlog1py(0., -1.) 
<tf.Tensor: shape=(), dtype=float32, numpy=0.> 

Args:

  • x: A tf.Tensor of type bfloat16, half, float32, float64, complex64, complex128
  • y: A tf.Tensor of type bfloat16, half, float32, float64, complex64, complex128
  • name: A name for the operation (optional).

Returns:

x * log1p(y).

Scipy Compatibility

Equivalent to scipy.special.xlog1py