jax.numpy.fmod#

jax.numpy.fmod(x1, x2, /)[source]#

Calculate element-wise floating-point modulo operation.

JAX implementation of numpy.fmod.

Parameters:
  • x1 (ArrayLike) – scalar or array. Specifies the dividend.

  • x2 (ArrayLike) – scalar or array. Specifies the divisor. x1 and x2 should either have same shape or be broadcast compatible.

Returns:

An array containing the result of the element-wise floating-point modulo operation of x1 and x2 with same sign as the elements of x1.

Return type:

Array

Note

The result of jnp.fmod is equivalent to x1 - x2 * jnp.fix(x1 / x2).

See also

Examples

>>> x1 = jnp.array([[3, -1, 4],
...                 [8, 5, -2]])
>>> x2 = jnp.array([2, 3, -5])
>>> jnp.fmod(x1, x2)
Array([[ 1, -1,  4],
       [ 0,  2, -2]], dtype=int32)
>>> x1 - x2 * jnp.fix(x1 / x2)
Array([[ 1., -1.,  4.],
       [ 0.,  2., -2.]], dtype=float32)