How do I simulate flip of biased coin?

random.random() returns a uniformly distributed pseudo-random floating point number in the range [0, 1). This number is less than a given number p in the range [0,1) with probability p. Thus:

def flip(p):
    return 'H' if random.random() < p else 'T'

Some experiments:

>>> N = 100
>>> flips = [flip(0.2) for i in xrange(N)]
>>> float(flips.count('H'))/N
0.17999999999999999  # Approximately 20% of the coins are heads

>>> N = 10000
>>> flips = [flip(0.2) for i in xrange(N)]
>>> float(flips.count('H'))/N
0.20549999999999999  # Better approximation 

Leave a Comment