tf.math.nextafter

Defined in generated file: python/ops/gen_math_ops.py

Returns the next representable value of x1 in the direction of x2, element-wise.

Aliases:

tf.math.nextafter(
    x1,
    x2,
    name=None
)

This operation returns the same result as the C++ std::nextafter function.

It can also return a subnormal number.

Args:

  • x1: A Tensor. Must be one of the following types: float64, float32.
  • x2: A Tensor. Must have the same type as x1.
  • name: A name for the operation (optional).

Returns:

A Tensor. Has the same type as x1.

Cpp Compatibility

Equivalent to C++ std::nextafter function.