Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

BUG: Scipy stats probability greater than 1

See original GitHub issue

Describe your issue.

I am computing values of a binomial probability mass function and find that their sum exceeds 1 when passing extreme parameters to the pmf. stackoverflow question. No error is thrown, meaning the problem could easily go unnoticed.

Reproducing Code Example

from scipy.stats import binom
n = 1541096362225563.0
p = 1.0477878413173978e-18
vals = [binom.pmf(n=n, p=p, k=i) for i in range(3)]
print(vals)  # -> [0.9983865609638467, 1.0, 0.060730098812886574]

Error message

SciPy/NumPy/Python version information

1.5.3 1.18.5 sys.version_info(major=3, minor=6, micro=9, releaselevel=‘final’, serial=0)

Issue Analytics

State:
Created a year ago
Comments:7 (2 by maintainers)

Top GitHub Comments

3reactions

mborlandcommented, Oct 16, 2022

See: https://github.com/boostorg/math/pull/847.

I used @mdhaber suggested test values, and now we have the correct values and a sum <= 1.

1reaction

mborlandcommented, Oct 18, 2022

Fix has been merged into Boost.Math. Let me know if you need anything else.

Top Results From Across the Web

Scipy stats probability greater than 1 - python - Stack Overflow

For such extremely small p and extremely large n I'd suggest trying a Poisson approximation to the binomial, where the rate λ is...

Statistics (scipy.stats) — SciPy v1.9.3 Manual

Let's check the number and name of the shape parameters of the gamma distribution. (We know from the above that this should be...

The dirichlet distribution unnecessarily rejects probabilities ...

I think that there is a bug in scipy.stats: the dirichlet ... but the issue remains that x can be zero if alpha...

Distributions

The probability distribution (for a fair dice) is constant: 1/6; Scipy.stats documentation on discrete statistical distributions. Continuous.

Some Basic Statistics - Jupyter Notebooks Gallery

One -way analysis of variance (ANOVA); Student's t-test; F-test; Coefficient of determination; Pearson's correlation coefficient; Probability Distribution ...