dados.py

import pandas as pd

df = pd.read_csv('pnad2012.csv', index_col=False)
df = df.fillna(df.mean())

#Retira indios e 
df = df[df.race != 0]
df = df[df.race != 8]

df.gender = df.gender.replace({1:'Homem', 2:'Mulher'})
df.race = df.race.replace({4:'N', 2:'N', 1:'B', 16:'B'})

df.education = df.education.replace([0],'Sem educação')
df.education = df.education.replace([1,2,3,4,5,6,7],'Ensino fundamental incompleto')
df.education = df.education.replace([8],'Ensino fundamental completo')
df.education = df.education.replace([9,10],'Ensino médio incompleto')
df.education = df.education.replace([11],'Ensino médio completo')
df.education = df.education.replace([12,13,14],'Ensino superior')
df.education = df.education.replace([15],'Pós-Graduação')

df.uf.replace({11: 'Rondônia', 12: 'Acre', 13: 'Amazonas', 14: 'Roraima', 15: 'Pará', 16: 'Amapá', 17: 'Tocantins', 21: 'Maranhão',
    22: 'Piauí', 23: 'Ceará', 24: 'Rio Grande do Norte', 25: 'Paraíba', 26: 'Pernambuco', 27: 'Alagoas', 28: 'Sergipe', 29: 'Bahia',
    31: 'Minas Gerais', 32: 'Espírito Santo', 33: 'Rio de Janeiro', 35: 'São Paulo', 41: 'Paraná', 42: 'Santa Catarina',
    43: 'Rio Grande do Sul', 50: 'Mato Grosso do Sul', 51: 'Mato Grosso', 52: 'Goiás', 53: 'Distrito Federal',
    }, inplace=True)

#Agrupa por genero, raça, escolaridade e estado.
def income(gender, race, uf, education):
    values = df[['gender', 'race', 'uf', 'education', 'income']].groupby(['gender','race','uf','education']).mean()
    renda = values['income'][gender, race, uf, education]
    return round(renda, 2)

def income_mean():
    values = df.income.mean()
    return values

def income_mean_uf(uf):
    values = df[['uf', 'income']].groupby(['uf']).mean()
    values = values['income'][uf]
    return values