import pandas as pd
import statsmodels.api as sm

db = pd.read_csv('https://raw.github.com/darribas/buzz_adam/master/buzz_data.csv')

db

db.head().T

db.tail().T

db.describe().T

db['div_i'].plot(kind='kde')

scatter(db['div_i'], db['industrie_pct_area'])

from pandas.tools.plotting import scatter_matrix
scatter_matrix(db.ix[:, :5], figsize=(12, 12))
show()

ols = sm.OLS(db['checkins_all'], db[['total_units', 'div_i']]).fit()

smry = ols.summary()
smry

inlatex = smry.as_latex()
print inlatex