Getting quantiles from a beta distribution using python

Question

I need to get the Nth quantile of a beta distribution, or equivalently, the 95% or 99% percentile. This is so much easier in Maple, which allows symbolic input -- but how is this done in Python?

I've searched stackoverflow, and it seems that people are often concerned with the normal distribution only.

I am confused. Are you asking for a way to tell, for example, whether `x=1` is in the 95% percentile? You could use `scipy.stats.beta` for this. — Julien, Oct 26 '15 at 01:50

score 6 · Accepted Answer · answered Oct 27 '15 at 03:05

6

I ended up with the ppf instead:

scipy.stats.beta.ppf(prob,2,N-2)

answered Oct 27 '15 at 03:05

irene

2,085
1
22
36

Much better. 123 – Petter Oct 16 '17 at 13:44

score 1 · Answer 2 · answered Oct 26 '15 at 02:14

You can compute the quantile of a beta distribution with the following function:

from scipy.stats import beta
import numpy as np
a, b = 2.31, 0.627
x = np.linspace(beta.ppf(0.01, a, b), beta.ppf(0.99, a, b), 100)
distribution=beta.pdf(x, a, b)
def quantile(x,quantiles):
    xsorted = sorted(x)
    qvalues = [xsorted[int(q * len(xsorted))] for q in quantiles]
    return zip(quantiles,qvalues)
quantiles = quantile(distribution,[0.05,0.16,.5,.84, 0.95])

Thanks a lot! I ended up using the ppf but this is really helpful. — irene, Oct 27 '15 at 03:08

Getting quantiles from a beta distribution using python

2 Answers2