Covid Stats analysis

covidstats_refactor.py

import csv
from datetime import datetime
from datetime import timedelta

__author__ = 'Jason Rahm'

# Assumes this script is in a <scripts> folder inside a cloned version of https://github.com/CSSEGISandData/COVID-19

# Total Kansas City Metro Area Stats
# 	Population: 2646933
# 	Cases: 5159 (per capita: 194.90482003133437 per 100,000 people)
# 	Deaths: 191 (per capita: 7.215898551266692 per 100,000 people)
# 	Death Rate by Known Cases (percentage): 3.702267881372359
#
#
# Total St Louis Metro Area Stats
# 	Population: 2803228
# 	Cases: 10017 (per capita: 357.3380402878396 per 100,000 people)
# 	Deaths: 752 (per capita: 26.82621606233956 per 100,000 people)
# 	Death Rate by Known Cases (percentage): 7.507237695916941
#
#
# Total Combined Kansas City & St Louis Metro Area Stats
# 	Population: 5450161
# 	Cases: 15176 (per capita: 278.4504898112184 per 100,000 people)
# 	Deaths: 943 (per capita: 17.30224116315096 per 100,000 people)
# 	Death Rate by Known Cases (percentage): 6.21375856615709

# St Louis Metro Statistical Area Counties
# UID, Name
# Illinois Counties
# 84017005, Bond
# 84017013, Calhoun
# 84017027, Clinton
# 84017083, Jersey
# 84017117, Macoupin
# 84017119, Madison
# 84017133, Monroe
# 84017163, St. Clair
# Missouri Counties
# 84029071, Franklin
# 84029099, Jefferson
# 84029113, Lincoln
# 84029183, St. Charles
# 84029189, St. Louis
# 84029510, St. Louis City
# 84029219, Warren
# Kansas City Metro Statistical Area Counties
# FIPS, Name
# Kansas Counties
# 84020091, Johnson
# 84020103, Leavenworth
# 84020107, Linn
# 84020121, Miami
# 84020209, Wyandotte
# Missouri Counties
# 84029013, Bates
# 84029025, Caldwell
# 84029037, Cass
# 84029047, Clay
# 84029049, Clinton
# 84029107, Lafayette
# 84029095, Jackson
# 84070003, Kansas City
# 84029165, Platte
# 84029177, Ray

stl_msa = [84017005,
           84017013,
           84017027,
           84017083,
           84017117,
           84017119,
           84017133,
           84017163,
           84029071,
           84029099,
           84029113,
           84029183,
           84029189,
           84029510,
           84029219]
kc_msa = [84020091,
          84020103,
          84020107,
          84020121,
          84020209,
          84029013,
          84029025,
          84029037,
          84029047,
          84029049,
          84029107,
          84029095,
          84029165,
          84029177,
          84070003]

s_stl = set(stl_msa)
s_kc = set(kc_msa)
all_counties = s_stl.union(s_kc)

# Get keys for today/yesterday
today = datetime.now()
yesterday = today - timedelta(days = 1)
today = today.strftime('%-m/%-d/%y')
yesterday = yesterday.strftime('%-m/%-d/%y')

# TESTING, Need to excuse non-date keys from the sort
# for x in combined_data:
#     print(sorted(x.items(), key=lambda x:datetime.strptime(x[0], '%m/%d/%y') ))

def merge(d1, d2):
  for k in set({**d1, **d2}):
    if k in d1:
      v1 = d1[k]
      if k in d2:
        v2 = d2[k]
        if k in ['FIPS', 'Combined_Key', 'Lat', 'Long_', 'Population']:
            yield k, (v1 if v1 == v2 else [v1, v2])
        else:
            yield k, ([v1, v2])
      else:
        yield k, v1
    else:
      yield k, d2[k]


def parse_covid_data(csv_file):
    with open(csv_file) as f:
        rdr = csv.reader(f, delimiter=',')
        colnames = next(rdr)
        colnames = [colnames[0]] + colnames[8:]
        ldata = []
        for row in rdr:
            if int(row[0]) in all_counties:
                row_data = [row[0]] + row[8:]
                data_dict = {colnames[i]: row_data[i] for i in range(len(colnames))}
                ldata.append(data_dict)
                if row[0] == str(max(all_counties)):
                    break
    return ldata


def print_aggregate_stats(population, cases, deaths):
    print(f'\tPopulation: {population}')
    print(f'\tCases: {cases} (per capita: {100000 * cases / population} per 100,000 people)')
    print(f'\tDeaths: {deaths} (per capita: {100000 * deaths / population} per 100,000 people)')
    print(f'\tDeath Rate by Known Cases (percentage): {100 * deaths / cases}')


cases_data = parse_covid_data('../csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_confirmed_US.csv')
deaths_data = parse_covid_data('../csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_deaths_US.csv')
print(deaths_data)
combined_data = [dict(merge(d1, d2)) for d1, d2 in zip(cases_data, deaths_data)]

total_cases = 0
total_kc_cases = 0
total_stl_cases = 0
total_deaths = 0
total_kc_deaths = 0
total_stl_deaths = 0
total_population = 0
total_kc_population = 0
total_stl_population = 0

print('\n\n#################')
print('# Today\'s Stats #')
print('#################')

for county in combined_data:
    # todo Trend Data for All Dates

    # Aggregate Data for Current Date
    population = int(county["Population"])
    if county.get(today) != None:
        stats = county[today]
    else:
        stats = county[yesterday]
    cases = int(stats[0])
    deaths = int(stats[1])
    total_cases += cases
    total_deaths += deaths
    total_population += population
    if int(county["UID"][0]) in stl_msa:
        print(f'\n\nStats for {county["Combined_Key"]} (STL Metro)')
        total_stl_cases += cases
        total_stl_deaths += deaths
        total_stl_population += population
    elif int(county["UID"][0]) in kc_msa:
        print(f'\n\nStats for {county["Combined_Key"]} (KC Metro)')
        total_kc_cases += cases
        total_kc_deaths += deaths
        total_kc_population += population

    population = int(county["Population"])
    print_aggregate_stats(population, cases, deaths)


print('\n\nTotal Kansas City Metro Area Stats')
print_aggregate_stats(total_kc_population, total_kc_cases, total_kc_deaths)

print('\n\nTotal St Louis Metro Area Stats')
print_aggregate_stats(total_stl_population, total_stl_cases, total_stl_deaths)

print('\n\nTotal Combined Kansas City & St Louis Metro Area Stats')
print_aggregate_stats(total_population, total_cases, total_deaths)