jax.lax.min

jax.lax.min(x, y)

Elementwise minimum: \(\mathrm{min}(x, y)\)

This function lowers directly to the stablehlo.minimum operation for non-complex inputs. For complex numbers, this uses a lexicographic comparison on the (real, imaginary) pairs.

Parameters:

• x (ArrayLike) – Input arrays. Must have matching dtypes. If neither is a scalar, x and y must have the same rank and be broadcast compatible.

• y (ArrayLike) – Input arrays. Must have matching dtypes. If neither is a scalar, x and y must have the same rank and be broadcast compatible.

Returns:

An array of the same dtype as x and y containing the elementwise minimum.

Return type:

Array

See also

• jax.numpy.minimum(): more flexibly NumPy-style minimum.

• jax.lax.reduce_min(): minimum along an axis of an array.

• jax.lax.max(): elementwise maximum.