cumulative.py

from datetime import datetime
vaccs = vaccine_df.copy()
daily = daily_df.copy()
daily.head()
# standardise the dates 
vaccs.date =vaccs.date.apply(lambda x: datetime.strptime(x, "%Y-%m-%d"))
daily.date =daily.date.apply(lambda x: datetime.strptime(x, "%Y-%m-%d"))

# use only common countries and dates 
countries = vaccs.dropna(subset=['daily_vaccinations'])['country'].unique()
dates = vaccs.dropna(subset=['daily_vaccinations'])['date'].unique()
country_mask = daily.country.apply(lambda x: x in countries)
date_mask = daily.date.apply(lambda x: x in dates)

# generate the visualization data 
columns_to_sum = ['daily_new_cases', 'cumulative_total_cases', 'cumulative_total_deaths', 'active_cases']
daily_cases = daily[country_mask & date_mask].groupby('date')[columns_to_sum].sum()
daily_vaccs = vaccs.groupby('date')[[ 'daily_vaccinations']].sum()

# make it a dataframe for convenience  
data = pd.DataFrame(daily_cases).join(pd.DataFrame(daily_vaccs))

# bring back the vaccine data we prepared in the previous section 
cumulative_vaccines = pd.DataFrame(vaccines.groupby('date')['total_vaccinations'].sum())
data = data.join(cumulative_vaccines).reset_index()