def _class_sizes():
# start with the actual distribution of class sizes from the book
d = {
7: 8,
12: 8,
17: 14,
22: 4,
27: 6,
32: 12,
37: 8,
42: 3,
47: 2,
}
# form the pmf
pmf = _04_Pmf._make_pmf_from_dict(d, 'actual')
print('mean', pmf._mean())
print('var', pmf._var())
# compute the biased pmf
biased_pmf = _bias_pmf(pmf, 'observed')
print('mean', biased_pmf._mean())
print('var', biased_pmf._var())
# unbias the biased pmf
unbiased_pmf = _unbias_pmf(biased_pmf, 'unbiased')
print('mean', unbiased_pmf._mean())
print('var', unbiased_pmf._var())
# plot the Pmfs
_05_myplot._pmfs([pmf, biased_pmf])
_05_myplot._show(xlabel='Class size', ylabel='PMF')
def testRender(self):
t = [1, 2, 2, 3, 5]
pmf = _04_Pmf._make_pmf_from_list(t)
xs, ps = pmf._render()
d = dict(zip(xs, ps))
new_pmf = _04_Pmf._make_pmf_from_dict(d)
self.checkPmf(new_pmf)
def _expected(pa, q, n):
"""
Makes a Pmf with the expected number of trials in each of four bins.
Args:
pa: probability of offering version A
q: probability of success for both A and B
n: number of trials
Returns:
hist that maps (version, outcome) to expected number,
where version is string A or B and outcome is string Y or N.
"""
versions = _04_Pmf._make_pmf_from_dict(dict(A=pa, B=1 - pa))
outcomes = _04_Pmf._make_pmf_from_dict(dict(Y=q, N=1 - q))
hist = _04_Pmf.Hist()
for version, pp in versions._items():
for outcome, qq in outcomes._items():
hist._incr((version, outcome), pp * qq * n)
return hist
def testMakePmf(self):
t = [1, 2, 2, 3, 5]
pmf = _04_Pmf._make_pmf_from_list(t)
self.checkPmf(pmf)
d = pmf._get_dict()
self.assertAlmostEquals(d[2], 0.4)
vals = pmf._values()
self.assertEquals(sorted(vals), [1, 2, 3, 5])
items = pmf._items()
d = dict(items)
new_pmf = _04_Pmf._make_pmf_from_dict(d)
self.checkPmf(new_pmf)
