def PlotResampledByDecade(resps, iters=11, predict_flag=False, omit=None):
"""Plots survival curves for resampled data.
resps: list of DataFrames
iters: number of resamples to plot
predict_flag: whether to also plot predictions
"""
for i in range(iters):
samples = [thinkstats2.ResampleRowsWeighted(resp)
for resp in resps]
sample = pd.concat(samples, ignore_index=True)
groups = sample.groupby('decade')
if omit:
groups = [(name, group) for name, group in groups
if name not in omit]
# TODO: refactor this to collect resampled estimates and
# plot shaded areas
if i == 0:
AddLabelsByDecade(groups, alpha=0.7)
if predict_flag:
PlotPredictionsByDecade(groups, alpha=0.1)
EstimateMarriageSurvivalByDecade(groups, alpha=0.1)
else:
EstimateMarriageSurvivalByDecade(groups, alpha=0.2)
def ResampleResps(resps):
"""Resamples each dataframe and then concats them.
resps: list of DataFrame
returns: DataFrame
"""
# we have to resample the data from each cycles separately
samples = [thinkstats2.ResampleRowsWeighted(resp, column='finalwgt')
for resp in resps]
# then join the cycles into one big sample
sample = pd.concat(samples, ignore_index=True)
# remove married people with unknown marriage dates
sample['missing'] = (sample.evrmarry & sample.agemarry.isnull())
sample = sample[~sample.missing]
# TODO: fill missing values
#DigitizeResp(sample)
#grouped = sample.groupby('birth_index')
#for name, group in iter(grouped):
# cdf = thinkstats2.Cdf(group.agemarry)
# print(name, cdf.Mean())
JitterResp(sample, 'age', jitter=1)
JitterResp(sample, 'agemarry', jitter=1)
DigitizeResp(sample)
return sample
def ResampleSurvival(resp, iters=101):
"""Resamples respondents and estimates the survival function.
resp: DataFrame of respondents
iters: number of resamples
"""
_, sf = EstimateMarriageSurvival(resp)
thinkplot.Plot(sf)
low, high = resp.min(), resp.max()
ts = np.arange(low, high, 1/12.0)
ss_seq = []
for _ in range(iters):
sample = thinkstats2.ResampleRowsWeighted(resp)
_, sf = EstimateMarriageSurvival(sample)
ss_seq.append(sf.Probs(ts))
low, high = thinkstats2.PercentileRows(ss_seq, [5, 95])
thinkplot.FillBetween(ts, low, high, color='gray', label='90% CI')
thinkplot.Save(root='survival3',
xlabel='age (years)',
ylabel='prob unmarried',
xlim=[12, 46],
ylim=[0, 1],
formats=FORMATS)
def PlotDivorceResampledByDecade(resps,
iters=11,
group_str='decade',
predict_flag=False,
omit=None):
"""Plots survival curves for resampled data.
resps: list of DataFrames
iters: number of resamples to plot
predict_flag: whether to also plot predictions
"""
for i in range(iters):
samples = [thinkstats2.ResampleRowsWeighted(resp) for resp in resps]
sample = pd.concat(samples, ignore_index=True)
groups = sample.groupby(group_str)
if omit:
groups = [(name, group) for name, group in groups
if name not in omit]
if i == 0:
AddLabelsByDecade(groups, alpha=0.7)
if predict_flag:
PlotDivorcePredictionsByDecade(groups, alpha=0.1)
EstimateDivorceSurvivalByDecade(groups, alpha=0.1)
else:
EstimateDivorceSurvivalByDecade(groups, alpha=0.2)
def ResampleDivorceCurveByDecade(resps):
for i in range(41):
samples = [thinkstats2.ResampleRowsWeighted(resp) for resp in resps]
sample = pandas.concat(samples, ignore_index=True)
groups = sample.groupby('decade')
if i == 0:
survival.AddLabelsByDecade(groups, alpha=0.7)
EstimateSurvivalByDecade(groups, alpha=0.1)
thinkplot.Save(root='survival7', xlabel='years', axis=[0, 28, 0, 1])
def ResampleDivorceCurve(resps):
"""Plots divorce curves based on resampled data.
resps: list of respondent DataFrames
"""
for _ in range(41):
samples = [thinkstats2.ResampleRowsWeighted(resp) for resp in resps]
sample = pandas.concat(samples, ignore_index=True)
PlotDivorceCurveByDecade(sample, color='#225EA8', alpha=0.1)
thinkplot.Show(xlabel='years', axis=[0, 28, 0, 1])
def ResampleByYear(df, column='wtssall'):
"""Resample rows within each year.
df: DataFrame
column: string name of weight variable
returns DataFrame
"""
grouped = df.groupby('year')
samples = [
thinkstats2.ResampleRowsWeighted(group, column) for _, group in grouped
]
sample = pd.concat(samples, ignore_index=True)
return sample
def ResampleDivorceCurveByDecade(resps):
"""Plots divorce curves for each birth cohort.
resps: list of respondent DataFrames
"""
for i in range(41):
samples = [thinkstats2.ResampleRowsWeighted(resp) for resp in resps]
sample = pandas.concat(samples, ignore_index=True)
groups = sample.groupby('decade')
if i == 0:
survival.AddLabelsByDecade(groups, alpha=0.7)
EstimateSurvivalByDecade(groups, alpha=0.1)
thinkplot.Save(root='survival6', xlabel='years', axis=[0, 28, 0, 1])
def DivorceCurveDecade(dfs):
""" plot divorce curves
@param: dfs - list of dataframes (hitched_)
"""
for i in range(50):
samples = [thinkstats2.ResampleRowsWeighted(df) for df in dfs]
sample = pd.concat(samples, ignore_index=True, sort=True)
groups = sample.groupby('decade')
if i == 0:
survival.AddLabelsByDecade(groups, alpha=0.5)
EstSurvivalDecade(groups, alpha=0.1)
thinkplot.Config(xlabel='Years',
ylabel='Fraction undivorced',
axis=[0, 28, 0, 1])
def ResampleSurvival(dados, limiar, iters=101):
"""Resamples respondents and estimates the survival function.
resp: DataFrame of respondents
iters: number of resamples
"""
_, sf = EstimateMarriageSurvival(dados, limiar)
thinkplot.Plot(sf)
low, high = dados.min(), dados.max()
ts = np.arange(low, high, 1)
ss_seq = []
for _ in range(iters):
sample = thinkstats2.ResampleRowsWeighted(pd.DataFrame(dados), column='MANSO')
_, sf = EstimateMarriageSurvival(sample['MANSO'], limiar)
ss_seq.append(sf.Probs(ts))
low, high = thinkstats2.PercentileRows(ss_seq, [5, 95])
thinkplot.FillBetween(ts, low, high, color='gray', label='90% CI')
def PlotResampledByAge(resps, **options):
samples = [thinkstats2.ResampleRowsWeighted(resp) for resp in resps]
sample = pandas.concat(samples, ignore_index=True)
groups = sample.groupby('decade')
#number of group divisions
n = 6
#number of years per group if there are n groups
group_size = 30 / n
#labels age brackets depending on # divs
labels = [
'{} to {}'.format(int(15 + group_size * i),
int(15 + (i + 1) * group_size)) for i in range(n)
]
# 0 representing 15-24, 1 being 25-34, and 2 being 35-44
#initilize dictionary of size n, with empty lists
prob_dict = {i: [] for i in range(n)}
#TODO: Look into not hardcoding this
decades = [30, 40, 50, 60, 70, 80, 90]
for _, group in groups:
#calcualates the survival function for each decade
_, sf = survival.EstimateSurvival(group)
if len(sf.ts) > 1:
#iterates through all n age groups to find the probability of marriage for that group
for group_num in range(0, n):
temp_prob_list = sf.Probs([
t for t in sf.ts if (15 + group_size * group_num) <= t <= (
15 + (group_num + 1) * group_size)
])
if len(temp_prob_list) != 0:
prob_dict[group_num].append(
sum(temp_prob_list) / len(temp_prob_list))
else:
pass
for key in prob_dict:
xs = decades[0:len(prob_dict[key])]
thinkplot.plot(xs, prob_dict[key], label=labels[key], **options)
def ResampleDivorceCurve(resps):
for _ in range(41):
samples = [thinkstats2.ResampleRowsWeighted(resp) for resp in resps]
sample = pandas.concat(samples, ignore_index=True)
PlotDivorceCurveByDecade(sample, color='#225EA8', alpha=0.1)
