def main():
# get the latests database files as a CSV
try:
pathToCSV_owid = CovidCasesOWID.download_CSV_file()
pathToCSV_who = CovidCasesWHO.download_CSV_file()
except FileNotFoundError:
# print an error message
print('Unable to download the database. Try again later.')
return
# some benchmarking
start = time.time()
# create instances
covidCases_owid = CovidCasesOWID(pathToCSV_owid)
covidCases_who = CovidCasesWHO(pathToCSV_who)
# create tuples of instances and country codes
objList = [covidCases_owid, covidCases_who]
# just in case we want to use some optionals
numCasesSince = 1000
lastN = 90
# the list of comma separated geoIDs
countryList = 'DE, GB, FR, ES, IT, CH, AT, EL, NA'
# get the combined dataframe
#df = CovidCases.create_combined_dataframe_by_geoid_string_list(objList, countryList)
#df = CovidCases.create_combined_dataframe_by_geoid_string_list(objList, countryList, lastNdays=lastN)
df = CovidCases.create_combined_dataframe_by_geoid_string_list(
objList, countryList, sinceNcases=numCasesSince)
# the width for a filter
width = 7
for obj in objList:
# add lowpass filtered DailyCases
df = obj.add_lowpass_filter_for_attribute(df, 'DailyCases', width)
# add r0
df = obj.add_r0(df)
# add lowpass filtered R
df = obj.add_lowpass_filter_for_attribute(df, "R", 7)
# benchmarking
end = time.time()
print(str((end - start)) + 's')
# plot it
plot_the_data(df)
# show the plot
plt.show()
# plot a map
#plot_map(covidCases_who)
return
def get_pygal_oceania_geoid_list():
"""Returns a list of GeoIDs of Oceanian countries that are available in PayGal and
the WHO data.
Be aware:
Not all countries of the WHO are available in PayGal and some names are different
(GB in PyGal = UK in WHO, GR in PyGal = EL in WHO). PyGal uses lower case and WHO
upper case.
Returns:
list: List of strings of GeoID's
"""
# just the main countries for a map
geoIDs = re.split(
r',\s*',
CovidCases.get_pygal_oceania_geoid_string_list().upper())
return geoIDs
def generate_plot(self,
geo_ids,
wanted_attrib,
log=False,
last_n=-1,
since_n=-1,
bar=False):
"""
Generates a plot for given GeoIds and returns it in form of a byteIO stream
Parameters:
geo_ids: [String] -> countries that should be plotted
wanted_attrib: String -> the field you want to plot, e.g. Cases
log: bool -> should the plot be logarithmic
last_n: int -> plot the last n days, if not further specified all available data is plotted
since_n: int -> plot since the nth case, if not further specified all available data is plotted
"""
# load the cases
csv_file = CovidCases.download_CSV_file()
self.covid_cases = CovidCases(csv_file)
# try to collect the data for given geoIds, if a wrong geoId is passed, the operation will abort with a 400
# bad request error
try:
df = self.covid_cases.get_country_data_by_geoid_list(
geo_ids, lastNdays=last_n, sinceNcases=since_n)
except IndexError:
raise HTTPException(status_code=400, detail="Couldn't load data")
# if the wanted attribute is one that need to be calculated, calculate it
if wanted_attrib == Attributes.R or Attributes.R7:
df = self.covid_cases.add_r0(df)
if wanted_attrib == Attributes.R7:
df = self.covid_cases.add_lowpass_filter_for_attribute(df, 'R', 7)
if wanted_attrib == Attributes.DailyCases7:
df = self.covid_cases.add_lowpass_filter_for_attribute(
df, 'DailyCases', 7)
if wanted_attrib == Attributes.DailyDeaths7:
df = self.covid_cases.add_lowpass_filter_for_attribute(
df, 'DailyDeaths', 7)
if wanted_attrib == Attributes.DoublingTime7:
df = self.covid_cases.add_lowpass_filter_for_attribute(
df, 'DoublingTime', 7)
if wanted_attrib == Attributes.Incidence7DayPer100Kpopulation:
df = self.covid_cases.add_incidence_7day_per_100Kpopulation(df)
# concat to one DataFrame and set date
df[['Date']] = pd.to_datetime(df['Date'], format='%d/%m/%Y')
# create pivot table with all needed values, if the x-axis shows a timedelta with days since the nth case the index
# has to change
pldf = df.pivot_table(values=wanted_attrib.name, index='Date', columns='Country') if since_n == -1 \
else df.pivot_table(values=wanted_attrib.name, index=df.index, columns='Country')
# use the PlotterBuilder to set up the plot
builder = (
PlotterBuilder(wanted_attrib)
#.set_title(re.sub(r"([a-z])([A-Z])", r"\g<1> \g<2>", wanted_attrib.name))
.set_title(AttributeTitles[wanted_attrib.value].value).set_grid())
if log:
builder.set_log()
if since_n != -1:
# if the plot has a timedelta on the x-axis, the label has to be reset
builder.set_axis_labels(xlabel="Days since case " + str(since_n))
builder.set_xaxis_index()
# generate plot
fig, ax = builder.build()
if bar:
pldf.plot(ax=ax, kind='bar')
else:
pldf.plot(ax=ax)
ax.grid()
# write image to io stream
byte_io = io.BytesIO()
plt.savefig(byte_io, dpi=fig.dpi)
byte_io.seek(0)
return byte_io
class Rest_API:
def generate_plot(self,
geo_ids,
wanted_attrib,
log=False,
last_n=-1,
since_n=-1,
bar=False):
"""
Generates a plot for given GeoIds and returns it in form of a byteIO stream
Parameters:
geo_ids: [String] -> countries that should be plotted
wanted_attrib: String -> the field you want to plot, e.g. Cases
log: bool -> should the plot be logarithmic
last_n: int -> plot the last n days, if not further specified all available data is plotted
since_n: int -> plot since the nth case, if not further specified all available data is plotted
"""
# load the cases
csv_file = CovidCases.download_CSV_file()
self.covid_cases = CovidCases(csv_file)
# try to collect the data for given geoIds, if a wrong geoId is passed, the operation will abort with a 400
# bad request error
try:
df = self.covid_cases.get_country_data_by_geoid_list(
geo_ids, lastNdays=last_n, sinceNcases=since_n)
except IndexError:
raise HTTPException(status_code=400, detail="Couldn't load data")
# if the wanted attribute is one that need to be calculated, calculate it
if wanted_attrib == Attributes.R or Attributes.R7:
df = self.covid_cases.add_r0(df)
if wanted_attrib == Attributes.R7:
df = self.covid_cases.add_lowpass_filter_for_attribute(df, 'R', 7)
if wanted_attrib == Attributes.DailyCases7:
df = self.covid_cases.add_lowpass_filter_for_attribute(
df, 'DailyCases', 7)
if wanted_attrib == Attributes.DailyDeaths7:
df = self.covid_cases.add_lowpass_filter_for_attribute(
df, 'DailyDeaths', 7)
if wanted_attrib == Attributes.DoublingTime7:
df = self.covid_cases.add_lowpass_filter_for_attribute(
df, 'DoublingTime', 7)
if wanted_attrib == Attributes.Incidence7DayPer100Kpopulation:
df = self.covid_cases.add_incidence_7day_per_100Kpopulation(df)
# concat to one DataFrame and set date
df[['Date']] = pd.to_datetime(df['Date'], format='%d/%m/%Y')
# create pivot table with all needed values, if the x-axis shows a timedelta with days since the nth case the index
# has to change
pldf = df.pivot_table(values=wanted_attrib.name, index='Date', columns='Country') if since_n == -1 \
else df.pivot_table(values=wanted_attrib.name, index=df.index, columns='Country')
# use the PlotterBuilder to set up the plot
builder = (
PlotterBuilder(wanted_attrib)
#.set_title(re.sub(r"([a-z])([A-Z])", r"\g<1> \g<2>", wanted_attrib.name))
.set_title(AttributeTitles[wanted_attrib.value].value).set_grid())
if log:
builder.set_log()
if since_n != -1:
# if the plot has a timedelta on the x-axis, the label has to be reset
builder.set_axis_labels(xlabel="Days since case " + str(since_n))
builder.set_xaxis_index()
# generate plot
fig, ax = builder.build()
if bar:
pldf.plot(ax=ax, kind='bar')
else:
pldf.plot(ax=ax)
ax.grid()
# write image to io stream
byte_io = io.BytesIO()
plt.savefig(byte_io, dpi=fig.dpi)
byte_io.seek(0)
return byte_io
def setup_routes(self, app: FastAPI):
"""
Setup of the route. The url has to be in the form:
/api/data/<country codes comma separated>/<attribute to be plotted>
?log=(True or False)[&lastN=X if you want to plot lastNdays][&sinceN=X if you want to plot since the Nth case]
"""
# setting up routes and implement methods
@app.get('/api/data/{countries}/{wanted_attrib}')
def get_data(countries: str,
wanted_attrib: Attributes,
sinceN: Optional[int] = None,
lastN: Optional[int] = None,
log: Optional[bool] = None,
bar: Optional[bool] = None):
"""
Returns a png image of the plotted Attribute (see Attributes) for a list of Countries(comma seperated). The URL needs to be in the following form:
**/api/data/<country codes comma separated>/<attribute to be plotted>
?log=(True or False)[&lastN=X if you want to plot lastNdays][&sinceN=X if you want to plot since the Nth case]**
SinceN and lastN plots the data starting from the given case or just the lastN days. Log is a boolean value that converts the y-scale to the logarithmic unit.
"""
# read the geoIds and plot the file
geo_ids = re.split(r",\s*", countries.upper())
file = self.generate_plot(geo_ids,
wanted_attrib,
last_n=lastN if lastN != None else -1,
log=log,
since_n=sinceN if sinceN != None else -1,
bar=bar)
# return the created stream as png image
return StreamingResponse(file, media_type="image/png")
import datetime
from datetime import date
import math
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import pandas as pd
from CovidCases import CovidCases
from PlotterBuilder import PlotterBuilder
# get the latests database file as a CSV
try:
pathToCSV = CovidCases.download_CSV_file()
except FileNotFoundError:
pathToCSV = "data/db.csv"
print('Unable to download the database. Try again later.')
except IOError:
pathToCSV = "data/db.csv"
print('Error writing file.')
# the list of comma separated geoIDs
countryList = 'DE, IT, ES, UK, FR, AT, NL'
# just in case we want to use some optionals
numCasesStart = 500
lastN = 40
# create an instance
covidCases = CovidCases(pathToCSV)
# get the dataframe for our countries
#df = covidCases.get_country_data_by_geoid_string_list(countryList, sinceNcases=numCasesStart)
df = covidCases.get_country_data_by_geoid_string_list(countryList,