plot-betten-sn.py

#!/usr/bin/env python3

# %% Imports
import datetime
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
import urllib.request as request
import json
import locale

# %% Fetch remote data
hospital_url = (
    "https://www.coronavirus.sachsen.de/corona-statistics/rest/hospitalDevelopment.jsp"
)
incidence_url = (
    "https://www.coronavirus.sachsen.de/corona-statistics/rest/incidenceDevelopment.jsp"
)

hospital_data = json.loads(request.urlopen(hospital_url).read())
incidence_data = json.loads(request.urlopen(incidence_url).read())

# %% Post-processing remote data
incidence_saxony = None
for incidence_set in incidence_data:
    if incidence_set["boundingArea"] == "Sachsen gesamt":
        incidence_saxony = incidence_set["incidenceDevelopment"]
        break

if incidence_saxony is None:
    raise SystemExit("Oh noes, saxony is gone...")
else:
    incidence_start = datetime.datetime.fromisoformat(incidence_saxony["startDate"])
    incidence_x = np.array(
        [
            incidence_start + datetime.timedelta(days=x)
            for x in range(len(incidence_saxony["values"]))
        ]
    )
    incidence_saxony["values"] = np.array(incidence_saxony["values"]).astype(np.double)
    incidence_mask = np.isfinite(incidence_saxony["values"])

hospital_x, occupied_beds = zip(*(hospital_data["numberOfOccupiedBeds"]))
_, occupied_beds_its = zip(*(hospital_data["numberOfOccupiedItsBeds"]))

# %% Create plot
locale.setlocale(locale.LC_ALL, "de_DE.UTF_8")
px = 1 / plt.rcParams["figure.dpi"]
fig, ax_beds = plt.subplots(figsize=(800 * px, 450 * px))
ax_incidence = ax_beds.twinx()

# ax_beds.set_xlim(datetime.datetime(2020, 12, 1, 0, 0), incidence_x[-1])
ax_beds.set_xlim(incidence_x[0], incidence_x[-1])

(p1,) = ax_beds.plot_date(
    mdates.epoch2num(np.divide(hospital_x, 1000)),
    np.subtract(occupied_beds, occupied_beds_its),
    "-",
    color="tab:brown",
    label="Belegt auf C19-Normalstation",
)
(p3,) = ax_beds.plot_date(
    mdates.epoch2num(np.divide(hospital_x, 1000)),
    occupied_beds,
    ":",
    color=p1.get_color(),
    label="Belegt mit C19 gesamt",
)
bed_threshold = 1300
ax_beds.axhline(y=bed_threshold, color="red", linestyle="--")
ax_beds.text(
    ax_beds.get_xlim()[0] + 1, bed_threshold * 1.05, "Neue Notbremse", color="red"
)
ax_beds.text(
    0.99,
    0.02,
    "Stand: {}\nDaten von coronavirus.sachsen.de".format(
        incidence_x[-1].strftime("%x")
    ),
    ha="right",
    transform=ax_beds.transAxes,
)

(p2,) = ax_incidence.plot_date(
    incidence_x[incidence_mask],
    incidence_saxony["values"][incidence_mask],
    "-",
    color="green",
    label="Inzidenz Sachsen",
)

ax_beds.set_ylabel("Betten")
ax_incidence.set_ylabel("7-Tages-Inzidenz")

ax_beds.grid(True)
# ax_incidence.grid(True)

ax_beds.yaxis.label.set_color(p1.get_color())
ax_incidence.yaxis.label.set_color(p2.get_color())

tkw = dict(size=4, width=1.5)
ax_beds.tick_params(axis="y", colors=p1.get_color(), **tkw)
ax_incidence.tick_params(axis="y", colors=p2.get_color(), **tkw)

ax_beds.legend(handles=[p1, p3, p2])  # loc='upper left')

fmt_major = mdates.MonthLocator(interval=1)
# fmt_major = mdates.WeekdayLocator()
ax_beds.xaxis.set_major_locator(fmt_major)
fmt_minor = mdates.DayLocator(interval=1)
ax_beds.xaxis.set_minor_locator(fmt_minor)

fig.autofmt_xdate()

# %% Save plot
fig.savefig("inzidenz-betten.pdf")