import pandas as pd
pd.set_option("display.max_rows", None)
from oscovida import get_incidence_rates_germany

cases_incidence, deaths_incidence = get_incidence_rates_germany()

cases_incidence

cases_incidence.sort_values(by="14-day-incidence-rate", ascending=False)

from oscovida import fetch_data_germany, germany_get_population
import datetime

period = 14 # Set the period we compute the incidence rate over

germany = fetch_data_germany() # Downloads population data for Germany
germany = germany.rename(
    columns={"AnzahlFall": "cases", "AnzahlTodesfall": "deaths"}
) # Rename the columns to English
germany.tail() # Print example of the table format and data

yesterday = datetime.datetime.now() - datetime.timedelta(days=1)
fortnight_ago = yesterday - datetime.timedelta(days=period)
periods = (fortnight_ago < germany.index) & (germany.index < yesterday) # Create a period mask
germany = germany.iloc[periods]

cases_germany = germany[["Landkreis", "cases"]]
cases_germany.tail() # The data for germany has the date as the index, and the location data on each column

cases_sum = (
    cases_germany.groupby("Landkreis")
    .sum()
    .rename(columns={"cases": f"{period}-day-sum"})
) # Here we sum the daily data to get total data
cases_sum.tail()

population = (
    germany_get_population()
    .population
    .to_frame()
) # This gives us population per county, with county as the index
population.tail()

cases_incidence = cases_sum.join(population)
cases_incidence[f"{period}-day-incidence-rate"] = (
    cases_incidence[f"{period}-day-sum"] / cases_incidence["population"] * 100_000
) # By convention the incedence rate is the total cases in the past x days / population * 100_000

cases_incidence.tail()

cases_incidence.loc['SK Hamburg']