CODEWITHSUNDEEP
pythonscipysympynumerical-methods
bettersayhello
bettersayhello

Reputation: 2597

Difference in output of sympy.binomial vs scipy.special.comb

Why does the following code that calls comb produce the wrong output?

from scipy.special import comb
from sympy import binomial


def coup(n=100, m=10):
    expectation = 0
    for j in range(1, n + 1):
        coeff = -1**(j-1) * binomial(n, j)
        denominator = 1 - (binomial(n - j, m) / binomial(n, m))
        expectation += coeff / denominator
    
   print(expectation)

   # output: 1764921496446.9807
   # correct output when using binomial: 49.9445660566641

Upvotes: 0

Views: 45

Answers (1)

Karan Shishoo
Karan Shishoo

Reputation: 2842

You need to include the exact parameter as True for the comb function if you want it to have full precision and not just floating point precision. (It defaults to False) for example -

comb(100,15) - binomial(100,15)

would give an output of -32, but

comb(100,15, exact=True) - binomial(100,15)

gives an output of 0

This is as per SciPy docs - https://docs.scipy.org/doc/scipy/reference/generated/scipy.special.comb.html

Upvotes: 4

Related Questions