Add line plot

wsgi.py

@@ -1,16 +1,17 @@
-import datetime
-from pathlib import Path
-import io
 import base64
+import datetime
+import io
+from itertools import chain
+from pathlib import Path
 
-
+import matplotlib.dates as mdates
+import matplotlib.pyplot as plt
+import matplotlib.ticker as mtick
+import numpy as np
 import pandas as pd
 from flask import Flask, render_template, request
 from flask_caching import Cache
-
 from matplotlib.backends.backend_agg import FigureCanvasAgg as FigureCanvas
-import matplotlib.pyplot as plt
-import matplotlib.ticker as mtick
 
 from download_digital import construct_dataframe, get_bez_data
 
@@ -42,7 +43,14 @@ def get_tables(url: str) -> tuple[pd.DataFrame, pd.DataFrame]:
     return df, df_state
 
 
-def plot(current_df: pd.DataFrame | None = None, data_folder: str = "data", sheet_name: str = "digital", total_target: int = 1500, plot_all: bool = False) -> str:
+def plot(
+    current_df: pd.DataFrame | None = None,
+    data_folder: str = "data",
+    sheet_name: str = "digital",
+    total_target: int = 1500,
+    plot_all: bool = False,
+    alpha: float | None = None,
+) -> str:
     data_dict = {}
     for f in sorted(Path(data_folder).iterdir()):
         with f.open("rb") as ff:
@@ -50,32 +58,76 @@ def plot(current_df: pd.DataFrame | None = None, data_folder: str = "data", shee
         df = df.astype({"Digitale Befragung": "Int32"})
         sum_val = df[["Digitale Befragung"]].sum().iloc[0]
         key = f.name[:10]
-        if plot_all or not (int(key[-2:]) - 15) % 7:
-            data_dict[key] = sum_val
+        data_dict[key] = sum_val
 
     data_dict["2023-08-15"] = 275
-    if not plot_all and current_df is not None:
+
+    series = pd.Series(data_dict.values(), index=data_dict)
+    series.index = series.index.astype("datetime64[ns]") + pd.DateOffset(hours=10)
+
+    df = series.to_frame("Digitale Befragung")
+    df = df.reindex(
+        pd.date_range(start="2023-08-15", end=max(data_dict.keys()))
+        + pd.DateOffset(hours=10)
+    )
+
+    if current_df is not None:
         current_df = current_df.astype({"Digitale Befragung": "Int32"})
         sum_val = current_df[["Digitale Befragung"]].sum().iloc[0]
-        data_dict["jetzt"] = sum_val
+        df.loc[datetime.datetime.now()] = sum_val
 
-    if plot_all:
-        for day in range(16, 19):
-            data_dict[f"2023-08-{day}"] = 0
-        for day in range(25, 27):
-            data_dict[f"2023-08-{day}"] = 0
+    plt.figure(dpi=300)
 
-    sorted_data_dict = {
-        k: v
-        for k, v in sorted(data_dict.items(), key=lambda x: x[0])
-    }
+    # fill weekends
+    max_date = max(data_dict.keys())
+    max_date = datetime.datetime.strptime(max_date, "%Y-%m-%d") + datetime.timedelta(
+        days=1
+    )
+    days = pd.date_range(start="2023-08-14", end=max_date)
+    for idx, day in enumerate(days[:-1]):
+        if day.weekday() >= 5:
+            plt.gca().axvspan(days[idx], days[idx + 1], alpha=0.2, color="gray")
 
-    plt.figure(figsize=(6, 5), dpi=300)
+    if alpha is not None:
+        plt.fill_between(
+            df.dropna().index,
+            df.dropna()["Digitale Befragung"],
+            color="#e4004e",
+            alpha=alpha,
+        )
 
-    plt.bar(sorted_data_dict.keys(), sorted_data_dict.values(), color="#e4004e")
-    plt.title("Entwicklung Digitale Beschäftigtenbefragung")
+    plt.plot(
+        df.dropna().index,
+        df.dropna()["Digitale Befragung"],
+        ls="--",
+        marker="o",
+        lw=1,
+        color="#e4004e",
+        markersize=4,
+    )
 
+    if current_df is not None:
+        plt.annotate(
+            "Jetzt",
+            (df.dropna().index[-1], df.dropna()["Digitale Befragung"][-1] * 1.03),
+            fontsize=8,
+            ha="center",
+        )
+
+    plt.plot(df.index, df["Digitale Befragung"], lw=1.5, color="#e4004e")
+
+    plt.title("Teilnahme an Digitaler Beschäftigtenbefragung")
     plt.ylabel("# Teilnahmen")
+    plt.ylim(0, total_target + 100)
+
+    # plt.gcf().autofmt_xdate()
+
+    plt.gca().xaxis.set_major_locator(mdates.WeekdayLocator([mdates.TU]))
+    plt.gca().xaxis.set_minor_locator(mdates.DayLocator())
+    plt.gca().xaxis.set_major_formatter(mdates.DateFormatter("%a %d.%m."))
+
+    plt.grid(True, which="major", axis="both")
+    plt.grid(True, which="minor", axis="x")
 
     def val_to_perc(val):
         return 100 * val / total_target
@@ -88,10 +140,6 @@ def plot(current_df: pd.DataFrame | None = None, data_folder: str = "data", shee
     sec_ax.yaxis.set_major_formatter(mtick.PercentFormatter())
 
     plt.axhline(y=total_target, color="#48a9be", linestyle="--")
-    # plt.axhline(y=federal_target_sheets[1], color="#48a9be", linestyle="--")
-    # plt.axhline(y=federal_target_sheets[2], color="#48a9be", linestyle="--")
-
-    # tikzplotlib.save("plot.tikz", axis_width="\\linewidth")
     plt.tight_layout()
 
     # Convert plot to PNG image
@@ -100,7 +148,7 @@ def plot(current_df: pd.DataFrame | None = None, data_folder: str = "data", shee
 
     # Encode PNG image to base64 string
     pngImageB64String = "data:image/png;base64,"
-    pngImageB64String += base64.b64encode(pngImage.getvalue()).decode('utf8')
+    pngImageB64String += base64.b64encode(pngImage.getvalue()).decode("utf8")
 
     return pngImageB64String
 
@@ -152,9 +200,7 @@ def tables(
     _print_as_html(df_state)
     _print_as_html(df)
 
-
-    image = plot(df_state)
-
+    image = plot(df_state, plot_all=True)
 
     return render_template(
         "base.html",