"If" function used for array in Python

Question

I have this function:

def q(x,y):
    r2 = (x/2.)**2 + (2.0*y)**2
    if (r2 > 0.04): 
        return 0.
    else:
        return 1.5

and I want to call it using something like

from pylab import *
dl=0.025
X, Y = mgrid[-0.5:0.5:dl, -0.5:0.5:dl]
g(X,Y)

but obviously this gives an error in the comparison.

Can this be done without having to make a for cycle for X and Y? Because if I want to call q(x,y) for two doubles, the for cycle must be reimplemented for that case...

Edit: (Adding this to the question as it is too long for a comment and not an answer, but it may help others answer.)

It appears pylab.mgrid is the same as numpy.mgrid.

Adjusted for numpy, this code

import numpy
def q(x,y):
    r2 = (x/2.)**2 + (2.0*y)**2
    if (r2 > 0.04): 
        return 0.
    else:
        return 1.5

dl=0.025
X, Y = numpy.mgrid[-0.5:0.5:dl, -0.5:0.5:dl]
q(X,Y)

gives this error

    if (r2 > 0.04):
ValueError: The truth value of an array with more than one element is
ambiguous. Use a.any() or a.all()

I don't know pylab so.. what exactly is in r2 and what do you want to do with it? If it is an array - what is it you want to compare? All elements must satisfy the condition or ..? — rplnt
– rplnt, Commented Oct 27, 2011 at 15:09
r2 is r^2 of an ellipsis... g(X,Y) should be an mgrid like but with values evaluated on every point of X,Y. — Jorge Leitao
– Jorge Leitao, Commented Oct 27, 2011 at 15:12
Please tell us what object mgrid is: print repr(mgrid) for a small example. — hochl
– hochl, Commented Oct 27, 2011 at 15:20
mgrid is: docs.scipy.org/doc/numpy/reference/generated/numpy.mgrid.html — Jorge Leitao
– Jorge Leitao, Commented Oct 27, 2011 at 15:45

Sven Marnach · Accepted Answer · 2011-10-27 15:42:34Z

3

The expression r2 > 0.04 evaluates to a NumPy array of Boolean values, so you can't use it in an if statement. You can have it automatically reinterpreted as numbers, though:

def q(x,y):
    r2 = (x/2.)**2 + (2.0*y)**2
    return 1.5 * (r2 <= 0.04)

answered Oct 27, 2011 at 15:42

Sven Marnach

608k123 gold badges966 silver badges865 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

Jorge Leitao Over a year ago

It is not the most general solution, but it solves the problem for this specific problem. Do you know a way of avoiding this for more cases (more if's else's etc.)

Sven Marnach Over a year ago

@J.C.Leitão: The specifics of how to vectorize a particular piece of code depend on the specifics of the code. In this particular example, it is comparitively simple, in other cases it is harder and sometimes it is impossible. The given technique can usually be extended to multiple if statements, possibly by using the & and | operators or partial assigment r2[r2 > 0.04] = ....

Jorge Leitao Over a year ago

Ok. Thank you Sven for the idea and knowledge shared.

Don · Accepted Answer · 2011-10-27 15:50:36Z

0

You may use map to apply your function to each couple of elements.

edited Oct 27, 2011 at 15:50

answered Oct 27, 2011 at 15:36

Don

17.7k13 gold badges67 silver badges106 bronze badges

3 Comments

Sven Marnach Over a year ago

In the given context, this answer is wrong. X and Y are two-dimensional arrays, so the zip() would produce pairs of rows, so the problem still isn't solved. Moreover, the whole point of NumPy is to avoid this kind of Python loop over all elements.

Don Over a year ago

Ok, I removed the non-working code. But the spirit of my answer was just to suggest 'map' as a possible way of resolving the issue; it didn't meant to be the most efficient solution.

Sven Marnach Over a year ago

Removed my downvote because the answer is no longer wrong. Honsetly, I think it's still not particularly helpful, though.

Collectives™ on Stack Overflow

"If" function used for array in Python

2 Answers 2

3 Comments

3 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

3 Comments

3 Comments

Your Answer

Sign up or log in

Post as a guest

Related