def RunEstimate(update_func, num_points=31, median_flag=False):
"""Runs the whole analysis.
update_func: which of the update functions to use
num_points: number of points in the Suite (in each dimension)
"""
d = ReadHeights(nrows=None)
labels = {1: 'male', 2: 'female'}
# PlotCdfs(d, labels)
suites = {}
for key, xs in d.items():
label = labels[key]
print(label, len(xs))
Summarize(xs)
xs = thinkbayes2.Jitter(xs, 1.3)
mus, sigmas = FindPriorRanges(xs, num_points, median_flag=median_flag)
suite = Height(mus, sigmas, label)
suites[label] = suite
update_func(suite, xs)
print('MLE', suite.MaximumLikelihood())
PlotPosterior(suite)
pmf_m = suite.Marginal(0)
pmf_s = suite.Marginal(1)
print('marginal mu', pmf_m.Mean(), pmf_m.Var())
print('marginal sigma', pmf_s.Mean(), pmf_s.Var())
# PlotMarginals(suite)
PlotCoefVariation(suites)
def RunEstimate(update_func, num_points=31, median_flag=False):
"""Runs the whole analysis.
update_func: which of the update functions to use
num_points: number of points in the Suite (in each dimension)
"""
d = ReadHeights(nrows=None)
labels = {1: 'male', 2: 'female'}
suites = {}
for key, xs in d.items():
label = labels[key]
print(label, len(xs))
Summarize(xs)
xs = thinkbayes2.Jitter(xs, 1.3)
mus, sigmas = FindPriorRanges(xs, num_points, median_flag=median_flag)
suite = Height(mus, sigmas, label)
suites[label] = suite
update_func(suite, xs)
print('MAP', suite.MaximumLikelihood())
# joint distributions of mu and sigma for men and women
suite1 = suites['male']
suite2 = suites['female']
def PlotCdfs(d, labels):
"""Plot CDFs for each sequence in a dictionary.
Jitters the data and subtracts away the mean.
d: map from key to sequence of values
labels: map from key to string label
"""
thinkplot.Clf()
for key, xs in d.items():
mu = thinkbayes2.Mean(xs)
xs = thinkbayes2.Jitter(xs, 1.3)
xs = [x - mu for x in xs]
cdf = thinkbayes2.MakeCdfFromList(xs)
thinkplot.Cdf(cdf, label=labels[key])
thinkplot.Show()
def RunEstimate(update_func, num_points=31, median_flag=False):
"""Runs the whole analysis.
update_func: which of the update functions to use
num_points: number of points in the Suite (in each dimension)
"""
d = ReadHeights(nrows=None)
labels = {1: 'male', 2: 'female'}
suites = {}
for key, xs in d.items():
label = labels[key]
print(label, len(xs))
Summarize(xs)
xs = thinkbayes2.Jitter(xs, 1.3)
mus, sigmas = FindPriorRanges(xs, num_points, median_flag=median_flag)
suite = Height(mus, sigmas, label)
suites[label] = suite
update_func(suite, xs)
print('MAP', suite.MaximumLikelihood())
suite1 = suites['male']
suite2 = suites['female']
mu1 = suite1.Marginal(0)
sigma1 = suite1.Marginal(1)
mu2 = suite2.Marginal(0)
sigma2 = suite2.Marginal(1)
diff = mu1 - mu2
sigma = (sigma1 + sigma2) / 2
pmf_d = diff / sigma
thinkplot.Cdf(pmf_d.MakeCdf())
thinkplot.Show(xlabel='# stddev between means', ylabel='PMF')