refactor

coronavis.py

@@ -6,77 +6,70 @@ import datetime
 import os
 import matplotlib.pyplot as pp
 import numpy as np
+import sys
+import importlib
+sys.path.append(".")
 
-
+#### config
-dataurl = "https://covid.ourworldindata.org/data/full_data.csv"
-date = datetime.date.today()
-
-datafile = f"{date}-full-data.csv"
-
-if not os.path.isfile(datafile):
-    r = requests.get(dataurl)
-    with open(datafile, "wb") as f:
-        f.write(r.content)
-else:
-    print(f"file found: {datafile}")
-
-# processing
-data = {}
-with open(datafile, "r") as f:
-    reader = csv.reader(f)
-    for row in reader:
-        date,location,new_cases,new_deaths,total_cases,total_deaths = row
-        if location=="location":
-            # table header
-            continue
-        if location not in data:
-            data[location] = []
-        year, month, day = date.split("-")
-        def toint(a):
-            try:
-                return int(a)
-            except:
-                return np.nan
-        data[location].append([datetime.date(int(year), int(month), int(day)), new_cases, new_deaths, total_cases, total_deaths])
-
-
-data2 = {}
-
-# reorganize data
-for loc in data:
-    time = []
-    new_cases, new_deaths, total_cases, total_deaths = [], [], [], []
-    for entry in data[loc]:
-        t_, new_cases_, new_deaths_, total_cases_, total_deaths_ = entry
-        time.append(t_)
-        new_cases.append(toint(new_cases_))
-        new_deaths.append(toint(new_deaths_))
-        total_cases.append(toint(total_cases_))
-        total_deaths.append(toint(total_deaths_))
-    data2[loc] = [time, new_cases, new_deaths, total_cases, total_deaths]
-
 # countries of interest
 countries = ["Germany", "Italy", "Spain", "Denmark", "China", "Japan", "South Korea", "Iran", "Belgium", "World"]
+# enabled plots
+plots = ["total_cases", "death_per_case", "normalized_to_first_death"]
+###
+def get_data():
+    """fetch data from remote, cache locally and reorganize internal data
+    not beautiful (at all), but effective!!"""
 
-# plotting
+    dataurl = "https://covid.ourworldindata.org/data/full_data.csv"
-for loc in data2:
+    date = datetime.date.today()
-    if loc not in countries:
-        continue
-    time, new_cases, new_deaths, total_cases, total_deaths = data2[loc]
 
-    # total cases
+    datafile = f"{date}-full-data.csv"
-    pp.figure("total_cases")
-    pp.plot(time, total_cases, label=f"{loc}")
 
-    # death/case
+    if not os.path.isfile(datafile):
-    pp.figure("death_per_case")
+        r = requests.get(dataurl)
-    pp.plot(time, np.array(total_deaths)/np.array(total_cases), label=f"{loc}")
+        with open(datafile, "wb") as f:
-# figure postprocessing
+            f.write(r.content)
-for fig in ["total_cases", "death_per_case"]:
+    else:
-    pp.figure(fig)
+        print(f"file found: {datafile}")
-    pp.yscale("log")
-    pp.xticks(rotation=90)
-    pp.legend(frameon=False)
-    pp.tight_layout()
 
-    pp.savefig(f"{fig}.png")
+    # processing
+    data = {}
+    with open(datafile, "r") as f:
+        reader = csv.reader(f)
+        for row in reader:
+            date,location,new_cases,new_deaths,total_cases,total_deaths = row
+            if location=="location":
+                # table header
+                continue
+            if location not in data:
+                data[location] = []
+            year, month, day = date.split("-")
+            def toint(a):
+                try:
+                    return int(a)
+                except:
+                    return np.nan
+            data[location].append([datetime.date(int(year), int(month), int(day)), new_cases, new_deaths, total_cases, total_deaths])
+    # reorganize data
+    data2 = {}
+    for loc in data:
+        time = []
+        new_cases, new_deaths, total_cases, total_deaths = [], [], [], []
+        for entry in data[loc]:
+            t_, new_cases_, new_deaths_, total_cases_, total_deaths_ = entry
+            time.append(t_)
+            new_cases.append(toint(new_cases_))
+            new_deaths.append(toint(new_deaths_))
+            total_cases.append(toint(total_cases_))
+            total_deaths.append(toint(total_deaths_))
+        data2[loc] = [time, new_cases, new_deaths, total_cases, total_deaths]
+    return data2
+
+data = get_data()
+
+for plot in plots:
+    #try:
+    i = importlib.import_module(plot)
+    i.plot(data, countries)
+    #except:
+    #    print(f"plotting {plot} failed")

death_per_case.py

@@ -0,0 +1,23 @@
+"""
+Plot total deaths per total cases of countries over time on log scale
+"""
+import matplotlib.pyplot as pp
+import numpy as np
+name="death_per_case"
+
+def plot(data, countries):
+    for loc in data:
+        if loc not in countries:
+            continue
+        time, new_cases, new_deaths, total_cases, total_deaths = data[loc]
+        
+        # death/case
+        pp.figure(name)
+        pp.plot(time, np.array(total_deaths)/np.array(total_cases), label=f"{loc}")
+
+    pp.yscale("log")
+    pp.xticks(rotation=90)
+    pp.legend(frameon=False)
+    pp.tight_layout()
+
+    pp.savefig(name+".png")

normalized_to_first_death.py

@@ -0,0 +1,24 @@
+"""
+Plot total cases timeshiftet to first death
+"""
+import matplotlib.pyplot as pp
+import numpy as np
+name="normalized_to_first_death"
+
+def plot(data, countries):
+    for loc in data:
+        if loc not in countries:
+            continue
+        time, new_cases, new_deaths, total_cases, total_deaths = data[loc]
+        
+        pp.figure(name)
+        day_of_first_death = np.argwhere(np.array(total_deaths) > 0)[0][0]
+        new_time_axis = np.arange(len(time)) - day_of_first_death
+        pp.plot(new_time_axis, np.array(total_cases), label=f"{loc}")
+
+    pp.yscale("log")
+    pp.xticks(rotation=90)
+    pp.legend(frameon=False)
+    pp.tight_layout()
+
+    pp.savefig(name+".png")

total_cases.py

@@ -0,0 +1,23 @@
+"""
+Plot total cases of countries over time on log scale
+"""
+import matplotlib.pyplot as pp
+name="total_cases"
+
+def plot(data, countries):
+
+    for loc in data:
+        if loc not in countries:
+            continue
+        time, new_cases, new_deaths, total_cases, total_deaths = data[loc]
+        
+        # total cases
+        pp.figure(name)
+        pp.plot(time, total_cases, label=f"{loc}")
+
+    pp.yscale("log")
+    pp.xticks(rotation=90)
+    pp.legend(frameon=False)
+    pp.tight_layout()
+
+    pp.savefig(name+".png")