select random key that is weighted by it's value

gistfile1.py

from random import choice

d = {
  'foo1': 1,
  'foo2': 5,
  'foo3': 3
}

choice([k for k in d for x in range(d[k])])

## think of it as
L = []
for k in d:
    for x in range(d[k]):
        L.append(k)
choice (L)

# L ends up end up as this list:
['foo1', 'foo2', 'foo2', 'foo2', 'foo2', 'foo2', 'foo3', 'foo3', 'foo3']

"""
By using choice you obviously have the most chance of it selecting 'foo2' in the region of 5 in 9 chance.
Memory usage is high for higher numbers so don't use this method for lots of choices or more precise chances of probability.
It will also only work with integers.
"""

By using choice you obviously have the most chance of it selecting 'foo2' in the region of 5 in 9 chance.
Memory usage is high for higher numbers so don't use this method for lots of choices or more precise chances of probability.
It will also only work with integers.

Heavily stolen from: http://stackoverflow.com/questions/1056151/random-python-dictionary-key-weighted-by-values