# Start with our normal batch of imports and settings
from __future__ import print_function, division

%matplotlib inline
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

import seaborn as sns; sns.set()

females = pd.read_csv('../data/femaleVisitsToPhysician.csv')
females.head()

females = pd.read_csv('../data/femaleVisitsToPhysician.csv',
                      index_col=0)
females.head()

males = pd.read_csv('../data/maleVisitsToPhysician.csv',
                    index_col=0)
males.head()

data = pd.concat([males, females]).sort_index()
data.head()

by_age = data.pivot_table('perCapita', index=['sex', 'age'], columns='year')
by_age.head()

fig, ax = plt.subplots(1, 2, figsize=(14, 5))
by_age.loc['f'].plot(ax=ax[0], title='Visits Per Capita: Females')
by_age.loc['m'].plot(ax=ax[1], title='Visits Per Capita: Males');

sns.set_palette('winter', 8)
fig, ax = plt.subplots(1, 2, figsize=(14, 5))
by_age.loc['f'].plot(ax=ax[0], title='Visits Per Capita: Females')
by_age.loc['m'].plot(ax=ax[1], title='Visits Per Capita: Males');

data['with_copay'] = (data['year'] < 2010)
data.head()

by_age_copay = data.pivot_table('perCapita', index=['sex', 'age'],
                                columns='with_copay', aggfunc='mean')
by_age_copay.head()

sns.set() # reset palette

fig, ax = plt.subplots(1, 2, figsize=(14, 5))
by_age_copay.loc['f'].plot(ax=ax[0], title='Visits Per Capita: Females')
by_age_copay.loc['m'].plot(ax=ax[1], title='Visits Per Capita: Males');

increase = 100 * (by_age_copay[False] / by_age_copay[True] - 1)

increase.loc['m'].plot(label='Males')
increase.loc['f'].plot(label='Females')
plt.legend();
plt.ylabel('increase (pct)')
plt.title('Percent Increase in per capita visits');

data['birth_year'] = data['year'] - data['age']
data.head()

pop_by_birthyear = data.pivot_table('population', index=['sex', 'birth_year'],
                                    columns='year', aggfunc='mean')

sns.set_palette('winter', 8)
fig, ax = plt.subplots(1, 2, figsize=(14, 5))
pop_by_birthyear.loc['m'].plot(title='Male Population', ax=ax[0]);
pop_by_birthyear.loc['f'].plot(title='Female Population', ax=ax[1]);

percap_by_birthyear = data.pivot_table('perCapita', index=['sex', 'birth_year'],
                                        columns='year', aggfunc='mean')
sns.set_palette('winter', 8)
fig, ax = plt.subplots(1, 2, figsize=(14, 5))
percap_by_birthyear.loc['m'].plot(title='Male Visits per Capita', ax=ax[0]);
percap_by_birthyear.loc['f'].plot(title='Female Visits per Capita', ax=ax[1]);

titanic = sns.load_dataset('titanic')
titanic.head()