You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
124 lines
5.1 KiB
124 lines
5.1 KiB
#!/usr/bin/env python3
|
|
#
|
|
# Usage: ./update_plots.py
|
|
# Updates plots from the Plotly section so they show the latest data.
|
|
|
|
from pathlib import Path
|
|
import datetime
|
|
import pandas as pd
|
|
from plotly.express import line
|
|
import plotly.graph_objects as go
|
|
import re
|
|
|
|
|
|
def main():
|
|
print('Updating covid deaths...')
|
|
update_covid_deaths()
|
|
print('Updating covid cases...')
|
|
update_confirmed_cases()
|
|
|
|
|
|
def update_covid_deaths():
|
|
covid = pd.read_csv('https://covid.ourworldindata.org/data/owid-covid-data.csv',
|
|
usecols=['iso_code', 'date', 'total_deaths', 'population'])
|
|
continents = pd.read_csv('https://gist.githubusercontent.com/stevewithington/20a69c0b6d2ff'
|
|
'846ea5d35e5fc47f26c/raw/country-and-continent-codes-list-csv.csv',
|
|
usecols=['Three_Letter_Country_Code', 'Continent_Name'])
|
|
df = pd.merge(covid, continents, left_on='iso_code', right_on='Three_Letter_Country_Code')
|
|
df = df.groupby(['Continent_Name', 'date']).sum().reset_index()
|
|
df['Total Deaths per Million'] = round(df.total_deaths * 1e6 / df.population)
|
|
today = str(datetime.date.today())
|
|
df = df[('2020-02-22' < df.date) & (df.date < today)]
|
|
df = df.rename({'date': 'Date', 'Continent_Name': 'Continent'}, axis='columns')
|
|
gb = df.groupby('Continent')
|
|
df['Max Total Deaths'] = gb[['Total Deaths per Million']].transform('max')
|
|
df = df.sort_values(['Max Total Deaths', 'Date'], ascending=[False, True])
|
|
f = line(df, x='Date', y='Total Deaths per Million', color='Continent')
|
|
f.update_layout(margin=dict(t=24, b=0), paper_bgcolor='rgba(0, 0, 0, 0)')
|
|
update_file('covid_deaths.js', f)
|
|
f.layout.paper_bgcolor = 'rgb(255, 255, 255)'
|
|
write_to_png_file('covid_deaths.png', f, width=960, height=340)
|
|
|
|
|
|
def update_confirmed_cases():
|
|
def main():
|
|
df = wrangle_data(*scrape_data())
|
|
f = get_figure(df)
|
|
update_file('covid_cases.js', f)
|
|
f.layout.paper_bgcolor = 'rgb(255, 255, 255)'
|
|
write_to_png_file('covid_cases.png', f, width=960, height=315)
|
|
|
|
def scrape_data():
|
|
def scrape_covid():
|
|
url = 'https://covid.ourworldindata.org/data/owid-covid-data.csv'
|
|
df = pd.read_csv(url, usecols=['location', 'date', 'total_cases'])
|
|
return df[df.location == 'World'].set_index('date').total_cases
|
|
def scrape_yahoo(slug):
|
|
url = f'https://query1.finance.yahoo.com/v7/finance/download/{slug}' + \
|
|
'?period1=1579651200&period2=9999999999&interval=1d&events=history'
|
|
df = pd.read_csv(url, usecols=['Date', 'Close'])
|
|
return df.set_index('Date').Close
|
|
out = [scrape_covid(), scrape_yahoo('BTC-USD'), scrape_yahoo('GC=F'),
|
|
scrape_yahoo('^DJI')]
|
|
return map(pd.Series.rename, out, ['Total Cases', 'Bitcoin', 'Gold', 'Dow Jones'])
|
|
|
|
def wrangle_data(covid, bitcoin, gold, dow):
|
|
df = pd.concat([dow, gold, bitcoin], axis=1) # Joins columns on dates.
|
|
df = df.sort_index().interpolate() # Sorts by date and interpolates NaN-s.
|
|
yesterday = str(datetime.date.today() - datetime.timedelta(1))
|
|
df = df.loc['2020-02-23':yesterday] # Discards rows before '2020-02-23'.
|
|
df = round((df / df.iloc[0]) * 100, 2) # Calculates percentages relative to day 1
|
|
df = df.join(covid) # Adds column with covid cases.
|
|
return df.sort_values(df.index[-1], axis=1) # Sorts columns by last day's value.
|
|
|
|
def get_figure(df):
|
|
figure = go.Figure()
|
|
for col_name in reversed(df.columns):
|
|
yaxis = 'y1' if col_name == 'Total Cases' else 'y2'
|
|
colors = {'Total Cases': '#EF553B', 'Bitcoin': '#636efa', 'Gold': '#FFA15A',
|
|
'Dow Jones': '#00cc96'}
|
|
trace = go.Scatter(x=df.index, y=df[col_name], name=col_name, yaxis=yaxis,
|
|
line=dict(color=colors[col_name]))
|
|
figure.add_trace(trace)
|
|
figure.update_layout(
|
|
yaxis1=dict(title='Total Cases', rangemode='tozero'),
|
|
yaxis2=dict(title='%', rangemode='tozero', overlaying='y', side='right'),
|
|
legend=dict(x=1.1),
|
|
margin=dict(t=24, b=0),
|
|
paper_bgcolor='rgba(0, 0, 0, 0)'
|
|
)
|
|
return figure
|
|
|
|
main()
|
|
|
|
|
|
###
|
|
## UTIL
|
|
#
|
|
|
|
def update_file(filename, figure):
|
|
lines = read_file(filename)
|
|
f_json = figure.to_json(pretty=True).replace('\n', '\n ')
|
|
out = lines[:6] + [f' {f_json}\n', ' )\n', '};\n']
|
|
write_to_file(filename, out)
|
|
|
|
|
|
def read_file(filename):
|
|
p = Path(__file__).resolve().parent / filename
|
|
with open(p, encoding='utf-8') as file:
|
|
return file.readlines()
|
|
|
|
|
|
def write_to_file(filename, lines):
|
|
p = Path(__file__).resolve().parent / filename
|
|
with open(p, 'w', encoding='utf-8') as file:
|
|
file.writelines(lines)
|
|
|
|
|
|
def write_to_png_file(filename, figure, width, height):
|
|
p = Path(__file__).resolve().parent / filename
|
|
figure.write_image(str(p), width=width, height=height)
|
|
|
|
|
|
if __name__ == '__main__':
|
|
main()
|