Calculate element-wise inverse of hyperbolic sine of input.
JAX implementation of numpy.arcsinh.
The inverse of hyperbolic sine is defined by:
\[arcsinh(x) = \ln(x + \sqrt{1 + x^2})\]
x (ArrayLike) – input array or scalar.
An array of same shape as x containing the inverse of hyperbolic sine of each element of x, promoting to inexact dtype.
Note
jnp.arcsinh returns nan for values outside the range (-inf, inf).
jnp.arcsinh follows the branch cut convention of numpy.arcsinh for complex inputs.
Examples
>>> x = jnp.array([[-2, 3, 1],
... [4, 9, -5]])
>>> with jnp.printoptions(precision=3, suppress=True):
... jnp.arcsinh(x)
Array([[-1.444, 1.818, 0.881],
[ 2.095, 2.893, -2.312]], dtype=float32)
For complex-valued inputs:
>>> x1 = jnp.array([4-3j, 2j]) >>> with jnp.printoptions(precision=3, suppress=True): ... jnp.arcsinh(x1) Array([2.306-0.634j, 1.317+1.571j], dtype=complex64)
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4