CODEWITHSUNDEEP
pythonnumpyoverflow
Window Licker
Window Licker

Reputation: 13

Numpy: How do I get inside an "overflow encountered in exp"?

First post, please be gentle!

I am trying to do root finding on an exponential function 1/(np.exp((j-w)/x)-1), and I find I keep getting the RuntimeWarning: overflow encountered in exp.

I understand this comes from the limitations on float64, but I can't seem to insert code to help understand where this is coming from. I have tried to include some try...except OverflowError blocks to give me a peek inside where this keeps happening, but it never seems to catch this.

I have ideas on how to handle it, as long as I can get my code to do something different when it encounters overflows.

Thanks in advance!

Upvotes: 1

Views: 782

Answers (1)

rpoleski
rpoleski

Reputation: 998

I'm guessing the problem is that you cannot figure out when warning happens. To do that use try/except but in slightly modified way:

import warnings
warnings.filterwarnings("error")
...
try:
    1/(np.exp((j-w)/x)-1)
except RuntimeWarning:
    print(j)
    print(w)
    print(x)
    print((j-w)/x)

This should give a hint, there the problem lies

Upvotes: 1

Related Questions