np.roll, but not toroidal (backfills with 0s), and only 1d.

zroll1d.py

import numpy as np

def zroll1d(arr1d, x):
    """np.roll, but not toroidal (backfills with 0s), and only 1d."""
    if x == 0:
        return arr1d
    elif x > 0:
        return np.pad(
            arr1d[:-x],
            ((x, 0),),
        )
    else:
        return np.pad(
            arr1d[-x:],
            ((0, -x),),
        )