Equivalent of numpy.where() value choice parameters

I am trying to emulate the behavior of the Numpy function numpy.where() in MATLAB, specifically the optional 2 parameters that allow picking from from other arrays.

I have the following Python code which selects

>>> import numpy as np
>>> x = np.arange(10)
>>> y = np.arange(10) * -0.1
>>> z = np.where(x % 2 == 0, x, y)
>>> print(z)
[ 0.  -0.1  2.  -0.3  4.  -0.5  6.  -0.7  8.  -0.9]

The function numpy.where() selects between the two arrays, x and y, based on the condition in the first argument.

The best I have been able to do in MATLAB is:

x = 0:9
y = (0:9) * -0.1

z = zeros(size(x))
z(mod(x, 2)!=0) = y(mod(x,2)==0)
z(mod(x, 2)~=0) = y(mod(x,2)~=0)

Is there a better way of doing this?

Answer 1

[Jan]

Your code looks almost fine:

x = 0:9
y = (0:9) * -0.1
z = zeros(size(x));

z(mod(x, 2) == 0) = x(mod(x,2) == 0);
%           ^ not ! ^ not y
z(mod(x, 2) ~= 0) = y(mod(x,2) ~= 0);

More efficient:

z = x;
m = (mod(x, 2) ~= 0);
z(m) = y(m);

[Andrew Glick]

Thanks!

(I'm a bit embarrassed that I made two obvious typos in the Matlab code. Oops! Thanks for catching those!)

[Jan]

Typos are then beloved companion of the programmer. They emerge from expressing thoughts by fingers. As soon as we omit the keyboards by developping a mind controlled input method, the new device will get a bi-directional interface and monopolistic internet services will project commercials directly in my subconciousness. Therefore I like typos.