def testRespondents(self):
resp = brfss.Respondents()
resp.ReadRecords(n=10000)
self.assertEquals(len(resp.records), 10000)
hist = MakeHist(resp, 'wtkg2')
t = hist.Values()
low, high = min(t), max(t)
self.assertAlmostEquals(low, 22.73)
self.assertEquals(hist.Freq('NA'), 343)
hist = MakeHist(resp, 'weight2')
t = hist.Values()
low, high = min(t), max(t)
self.assertAlmostEquals(low, 22.727272727)
self.assertEquals(hist.Freq('NA'), 343)
hist = MakeHist(resp, 'wtyrago')
t = hist.Values()
low, high = min(t), max(t)
self.assertAlmostEquals(low, 27.27272727)
self.assertEquals(hist.Freq('NA'), 616)
hist = MakeHist(resp, 'htm3')
t = hist.Values()
low, high = min(t), max(t)
self.assertAlmostEquals(low, 104)
self.assertEquals(hist.Freq('NA'), 101)
hist = MakeHist(resp, 'sex')
self.assertEquals(hist.Freq(1), 3669)
self.assertEquals(hist.Freq(2), 6331)
def DumpHeights(data_dir='.', n=10000):
"""Read the BRFSS dataset, extract the heights and pickle them."""
resp = brfss.Respondents()
resp.ReadRecords(data_dir, n)
d = {1: [], 2: []}
[d[r.sex].append(r.htm3) for r in resp.records if r.htm3 != 'NA']
fp = open('variability_data.pkl', 'wb')
pickle.dump(d, fp)
fp.close()
def main():
resp = brfss.Respondents()
resp.ReadRecords(data_dir='res')
heights, weights = resp.GetHeightAndWeight()
r1 = Corr(heights, weights)
# 方法2
r2 = correlation.Corr(heights, weights)
print(r1, "vs", r2)
print("E = 1 vs", Corr(heights, heights))
def main():
# test
t = [7, 1, 2, 5]
print(ToRanks(t))
# brfss
resp = brfss.Respondents()
resp.ReadRecords(data_dir='res')
heights, weights = resp.GetHeightAndWeight()
r1 = SpearmanCorr(heights, weights)
r2 = correlation.SpearmanCorr(heights, weights)
print("r1, r2 = ", r1, r2)
def main():
resp = brfss.Respondents()
resp.ReadRecords(data_dir='res')
d = resp.SummarizeHeight()
man_d = d[1]
lady_d = d[2]
# 男性的mu, var, sigma, 变异系数CV
man_mu, man_var = thinkstats.TrimmedMeanVar(man_d)
man_sigma = math.sqrt(man_var)
man_cv = man_sigma/man_mu
print("man: mu = %.3f, var = %.3f, sigma = %.3f, cv = %.3f" % (man_mu, man_var, man_sigma, man_cv))
# 女性的mu, var, sigma, 变异系数CV
lady_mu, lady_var = thinkstats.TrimmedMeanVar(lady_d)
lady_sigma = math.sqrt(lady_var)
lady_cv = lady_sigma/lady_mu
print("lady: mu = %.3f, var = %.3f, sigma = %.3f, cv = %.3f" % (lady_mu, lady_var, lady_sigma, lady_cv))
# 男性, 女性Hist分布
man_hist = Pmf.MakeHistFromList(man_d, name='man hist')
myplot.Hist(man_hist)
myplot.Show()
myplot.Clf()
lady_hist = Pmf.MakeHistFromList(lady_d, name='lady hist')
myplot.Hist(lady_hist)
myplot.Show()
myplot.Clf()
# 男性, 女性Pmf分布
man_pmf = Pmf.MakePmfFromHist(man_hist, name='man pmf')
myplot.Pmf(man_pmf)
myplot.Show()
myplot.Clf()
lady_pmf = Pmf.MakePmfFromHist(lady_hist, name='lady pmf')
myplot.Pmf(lady_pmf)
myplot.Show()
myplot.Clf()
# 男性/女性Cdf累积分布
man_cdf = Cdf.MakeCdfFromPmf(man_pmf, name='man cdf')
lady_cdf = Cdf.MakeCdfFromPmf(lady_pmf, name='lady cdf')
myplot.Cdfs((man_cdf, lady_cdf), complement=False, transform=None)
myplot.Show()
def main():
resp = brfss.Respondents()
resp.ReadRecords(data_dir='res')
d = resp.SummarizeHeight()
man_l = d[1]
cdf = Cdf.MakeCdfFromList(man_l, name='man height')
mu = cdf.Mean()
var = cdf.Var(mu=mu)
sigma = math.sqrt(var)
print("man height: mean = %.3f var = %.3f sigma = %.3f" % (mu, var, sigma))
myplot.Cdf(cdf, complement=False, transform=None)
myplot.Show()
v = (cdf.Prob(185) - cdf.Prob(178)) * 100
print("178 - 185: %.3f%%" % v)
def main():
resp = brfss.Respondents()
resp.ReadRecords(data_dir='res')
heights, weights = resp.GetHeightAndWeight()
c1 = Cov(heights, weights)
c2 = Cov(heights, heights)
_, var = thinkstats.MeanVar(heights)
print(c1, c2, var)
print("-------------- ")
# 官方方法2
c3 = correlation.Cov(heights, weights)
c4 = correlation.Cov(heights, heights)
print(c3, c4, var)
import Cdf
import math
import myplot
import numpy
import brfss
resp = brfss.Respondents()
resp.ReadRecords('.')
filtered_records = filter(lambda (x): x.wtkg2 != 'NA', resp.records)
xs = sorted(numpy.random.normal(0, 1, len(filtered_records)))
ys = sorted(map(lambda (y): y.wtkg2, filtered_records))
logys = sorted(map(lambda (y): math.log10(y.wtkg2), filtered_records))
myplot.Plot(xs, ys)
myplot.Show(title="Normal plot for weight")
myplot.Plot(xs, logys)
myplot.Show(title="Normal plot for log(weight)")
