Computes exp(x) - 1
element-wise.
tf.math.expm1(
x: Annotated[Any, tf.raw_ops.Any
],
name=None
) -> Annotated[Any, tf.raw_ops.Any
]
i.e. exp(x) - 1
or e^(x) - 1
, where x
is the input tensor.
e
denotes Euler's number and is approximately equal to 2.718281.
x = tf.constant(2.0)
tf.math.expm1(x) ==> 6.389056
x = tf.constant([2.0, 8.0])
tf.math.expm1(x) ==> array([6.389056, 2979.958], dtype=float32)
x = tf.constant(1 + 1j)
tf.math.expm1(x) ==> (0.46869393991588515+2.2873552871788423j)
Args | |
---|---|
x
|
A Tensor . Must be one of the following types: bfloat16 , half , float32 , float64 , complex64 , complex128 .
|
name
|
A name for the operation (optional). |
Returns | |
---|---|
A Tensor . Has the same type as x .
|