Update app.py

app.py

@@ -8,37 +8,131 @@ matplotlib.use("Agg")
 
 import matplotlib.pyplot as plt
 import numpy as np
-# import plotly.express as px
+import plotly.express as px
 import pandas as pd
 
-df_pre = pd.read_csv('data')
-y = df_pre['audience']
+# Webページを取得して解析する
+load_url = "https://www.football-lab.jp/kyot/match/"
+html = requests.get(load_url)
+soup = bs(html.content, "html.parser")
+
+!pip install beautifulsoup4
+!pip install requests_html
+import requests
+from bs4 import BeautifulSoup as bs
+from requests_html import AsyncHTMLSession
+
+import pandas as pd
+from datetime import datetime
+from IPython.display import display
+
+url23 = 'https://www.football-lab.jp/ka-f/match/'
+dfs23 = pd.read_html(url23)
+url22 = 'https://www.football-lab.jp/ka-f/match/?year=2022'
+dfs22 = pd.read_html(url22)
+url21 = 'https://www.football-lab.jp/ka-f/match/?year=2021'
+dfs21 = pd.read_html(url21)
+url20 = 'https://www.football-lab.jp/ka-f/match/?year=2020'
+dfs20 = pd.read_html(url20)
+
+#シーズン毎に分類
+res23 = pd.DataFrame([['S2023']]*len(dfs23[0])).join(dfs23, lsuffix='0')
+res22 = pd.DataFrame([['S2022']]*len(dfs22[0])).join(dfs22, lsuffix='0')
+res21 = pd.DataFrame([['S2021']]*len(dfs21[0])).join(dfs21, lsuffix='0')
+res20 = pd.DataFrame([['S2020']]*len(dfs20[0])).join(dfs20, lsuffix='0')
+
+df_tmp = pd.concat([res23, res22, res21, res20]) 
+df = df_tmp
+
+df = df.rename(columns={'会場': 'stadium', 0: 'year', '開催日': 'date', '観客数': 'audience'})
+df = df.query('stadium=="等々力"').reset_index()
+df = df.query('audience.notna()', engine='python').reset_index()
+df = df[['audience', 'year', 'date']]
+#seasonカラムから年を抽出
+df["year"] = df["year"].apply(lambda x: str(x)[1:5])
+#開催日から月と日を分割
+df['month']  = df['date'].str.split(pat='.', expand=True)[0]
+df['day'] = df['date'].str.split(pat='.', expand=True)[1]
+#数値データを日付データに変換
+df['date'] = pd.to_datetime({'year': df['year'], 'month': df['month'], 'day': df['day']})
+#日付昇順に並び替える
+df = df.sort_values('date', ascending=True)
+df['date_ymd'] = pd.to_datetime(df['date']).dt.strftime('%Y%m%d')
+df['date_ym'] = pd.to_datetime(df['date']).dt.strftime('%Y%m')
+df["date_ymd"] = df["date_ymd"].astype(int)
+df['date_before'] = df['date_ymd'] - 1
+df["date_before"] = df["date_before"]
+df = df[['audience', 'date_ymd', 'date_before']] 
+df['last_audience'] = df['audience'].shift(1)
+
+df_aji = pd.read_csv('fish_price.csv')
+
+df_train = pd.merge(df, df_aji, left_on='date_before', right_on='date', how='left')
+
+df_train = df_train.query('date > 20201202')
+df_train = df_train.drop(['date_before', 'date_ymd'], axis=1)
+df_train["audience"] = df_train["audience"].str.replace(",", "").astype(int)
+df_train["last_audience"] = df_train["last_audience"].str.replace(",", "").astype(int)
+
+X = df_train.drop('audience', axis=1)
+y = df_train['audience']
+
+from sklearn.linear_model import LinearRegression
+from sklearn.metrics import log_loss
+from sklearn.preprocessing import StandardScaler
+linear_regression = LinearRegression()
+linear_regression.fit(X,y)
 
-filename = 'model.pkl'
-loaded_rf_model = pickle.load(open(filename, 'rb'))
-X_test = pd.DataFrame(
-    data={'saba': [300, 200, 3030, 400], 
-          'date_ymd': [20230328, 20230329, 20230330, 20230331]}
-)
-y_pred_ = loaded_rf_model.predict(X_test, num_iteration=gbm.best_iteration)
 
 def outbreak(date):
   if date:
-    # start_day = start_day
-    # final_day = date
-    # x = np.arange(start_day, final_day + 1)
+
+    
+
+    if __name__ == "__main__":
+      start_date = d_today
+      end_date = d_tom
+      df_aji_pre = get_fish_price_data(start_date=start_date, end_date=end_date)
+      # df_aji_pre.to_csv("fish_price_pre.csv", index=False)
+    
+    df_pre = df.tail(1).reset_index()
+    df_pre = df_pre.drop('index', axis=1)
+    df_aji_ft_pre = pd.concat([df_pre, df_aji_pre], axis=1)
+    df_aji_ft_pre = df_aji_ft_pre[['audience', 'date', 'low_price', 'center_price', 'high_price', 'quantity']]
+    df_aji_ft_pre = df_aji_ft_pre.rename(columns={'audience': 'last_audience', 0: 'year', '開催日': 'date', '観客数': 'audience'})
+
+    pred = linear_regression.predict(df_aji_ft_pre)
+    df_aji_ft_pre['audience_pred'] = pred
+    df_aji_ft_pre['date'] = df_aji_ft_pre['date'].astype(int)
+
     fig = plt.figure()
-    plt.plot(y_pred_)
-    plt.title("audience")
+    plt.plot(df_train['date'], df_train['audience'], label='original')
+    plt.plot(df_aji_ft_pre['date'], df_aji_ft_pre['audience_pred'], '*', label='predict')
+    plt.title("prediction of audince")
     plt.ylabel("audience")
     plt.xlabel("Days since Day 0")
     return fig
 
-demo = gr.Interface(
-    fn=outbreak,
-    inputs='checkbox',
-    outputs=gr.Plot()
-)
+with gr.Blocks() as demo:
+    gr.Markdown(
+                            """
+                            # 川崎フロンターレの観客動員数の予測
+                             川崎フロンターレの等々力陸上競技場での試合の観客数を「あじ」の価格をもとに予測する。
+                            ## 使用データ
+                             * 東京卸売市場日報
+                             * Football Lab 
+                            ## 予測ロジック
+                            観客動員数は雨天か否かで左右されると考えられる。そこで雨天の可能性をあじの価格を利用し表した。
+                            一般的に雨天の場合、低気圧の影響で海面が上昇し漁に出ることが難しくなる。
+                            そのため漁獲量が減少し、あじの価格が上昇すると考えられる。
+                             """
+                        )
+    with gr.Row():
+      with gr.Column():
+        date_input = gr.Checkbox(label='please input date')
+        prediction_btn = gr.Button(value="predict")
+      with gr.Column():
+        prediction = gr.Plot(label = "時系列プロット")
+    prediction_btn.click(outbreak, inputs=date_input, outputs=prediction)
 
-if __name__ == "__main__":
-    demo.launch()
+demo.launch()