Return to Answer

replaced http://stackoverflow.com/ with https://stackoverflow.com/

Source Link

edited May 23, 2017 at 12:00

URL Rewriter Bot

For this special case, you could also write a function that operates on both, NumPy arrays and plain Python floats:

def func2d(x, y):
    z = 2.0 * (x > y) - 1.0
    z *= y
    return x + z

This version is also more than four times as fast as unutbu's func2a()unutbu's func2a() (tested with N = 100).

For this special case, you could also write a function that operates on both, NumPy arrays and plain Python floats:

def func2d(x, y):
    z = 2.0 * (x > y) - 1.0
    z *= y
    return x + z

This version is also more than four times as fast as unutbu's func2a() (tested with N = 100).

For this special case, you could also write a function that operates on both, NumPy arrays and plain Python floats:

def func2d(x, y):
    z = 2.0 * (x > y) - 1.0
    z *= y
    return x + z

This version is also more than four times as fast as unutbu's func2a() (tested with N = 100).

Source Link

answered Mar 5, 2011 at 13:25

For this special case, you could also write a function that operates on both, NumPy arrays and plain Python floats:

def func2d(x, y):
    z = 2.0 * (x > y) - 1.0
    z *= y
    return x + z

This version is also more than four times as fast as unutbu's func2a() (tested with N = 100).