import pandas as pd
from regression import Regression
x = Regression()
df = pd.read_csv("trips_summary_covid_pub.csv")
df['date'] = pd.to_datetime(df['date'])
df['pre_covid'] = (df.date < "12/19/2020").astype("int")
df = df.loc[df.pre_covid == 0]
# TOTAL REVENUE MODEL
x.select_cols = [
'pickup_community_area', 'hour', 'week_day', 'cases', 'total_revenue'
]
x.dummy_cols = ['pickup_community_area', 'hour', 'week_day']
x.y_col = 'total_revenue'
res_revenue = x.time_split(df)
# TOTAL COUNT MODEL
x.select_cols = ['pickup_community_area', 'hour', 'week_day', 'cases', 'count']
x.y_col = 'count'
res_count = x.time_split(df)
# TOTAL SECONDS MODEL
x.select_cols = [
'pickup_community_area', 'hour', 'week_day', 'cases', 'trip_seconds_tot'
]
x.y_col = 'trip_seconds_tot'
# READ DATASET
df = pd.read_csv("trips_summary_covid_pub.csv")
df['date'] = pd.to_datetime(df['date'])
# FILTER PRE AND POST PANDEMIC
df['pre_covid'] = (df.date < "12/19/2020").astype("int")
df_pre_covid = df.loc[df.pre_covid == 1]
df = df.loc[df.pre_covid == 0]
df = df.loc[df['total_revenue'] > 0]
df_pre_covid = df_pre_covid.loc[df_pre_covid['total_revenue'] > 0]
# TOTAL REVENUE MODEL
x.select_cols = ['pickup_community_area', 'hour', 'total_revenue']
x.dummy_cols = ['pickup_community_area', 'hour']
x.y_col = 'total_revenue'
res_revenue = x.poisson_regression(df, 0.7)
# TOTAL REVENUE PRE COVID
x.select_cols = ['pickup_community_area', 'hour', 'total_revenue']
x.y_col = 'total_revenue'
res_revenue_pre = x.poisson_regression(df_pre_covid, 0.7)
# TOTAL COUNT MODEL
x.select_cols = ['pickup_community_area', 'hour', 'count']
x.y_col = 'count'
res_count = x.poisson_regression(df, 0.7)