Update app.py

app.py

@@ -40,174 +40,101 @@ NBA_Data = 'https://docs.google.com/spreadsheets/d/1Yq0vGriWK-bS79e-bD6_u9pqrYE6
 def init_baselines():
     sh = gcservice_account.open_by_url(NBA_Data)
     
-    worksheet = sh.worksheet('Trending')
+    worksheet = sh.worksheet('Gamelog')
     raw_display = pd.DataFrame(worksheet.get_values())
     raw_display.columns = raw_display.iloc[0]
     raw_display = raw_display[1:]
     raw_display = raw_display.reset_index(drop=True)
-    trend_table = raw_display[raw_display['PLAYER_NAME'] != ""]
-    trend_table.replace('', np.nan, inplace=True)
-    trend_table = trend_table[['PLAYER_NAME', 'Team', 'Position', 'FD_Position', 'L10 MIN', 'L10 Fantasy', 'L10 Ceiling', 'L10 FD_Fantasy',
-                               'L10 FD_Ceiling', 'L5 MIN', 'L5 Fantasy', 'L5 Ceiling', 'L5 FD_Fantasy', 'L5 FD_Ceiling', 'L3 MIN', 'L3 Fantasy',
-                               'L3 Ceiling', 'L3 FD_Fantasy', 'L3 FD_Ceiling', 'Trend Min', 'Trend Median', 'DK_Proj', 'Adj Median', 'Adj Ceiling',
-                               'Trend FD_Median', 'FD_Proj', 'Adj FD_Median', 'Adj FD_Ceiling', 'DK_Salary', 'DK_Avg_Val', 'DK_Ceiling_Value',
-                               'FD_Salary', 'FD_Avg_Val', 'FD_Ceiling_Value']]
-    trend_table['DK_Salary'] = trend_table['DK_Salary'].str.replace(',', '').astype(float)
-    trend_table['FD_Salary'] = trend_table['FD_Salary'].str.replace(',', '').astype(float)
-    data_cols = trend_table.columns.drop(['PLAYER_NAME', 'Team', 'Position', 'FD_Position'])
-    trend_table[data_cols] = trend_table[data_cols].apply(pd.to_numeric, errors='coerce')
+    gamelog_table = raw_display[raw_display['PLAYER_NAME'] != ""]
+    gamelog_table.replace('', np.nan, inplace=True)
+    gamelog_table = gamelog_table[['PLAYER_NAME', 'TEAM_NAME', 'SEASON_ID', 'GAME_DATE', 'MATCHUP', 'MIN', 'touches', 'PTS', 'FGM', 'FGA', 'FG_PCT', 'FG3M', 'FG3A',
+                                   'FG3_PCT', 'FTM', 'FTA', 'FT_PCT', 'reboundChancesOffensive', 'OREB', 'reboundChancesDefensive', 'DREB', 'reboundChancesTotal', 'REB',
+                                   'passes', 'secondaryAssists', 'freeThrowAssists', 'assists', 'STL', 'BLK', 'TOV', 'PF', 'DD', 'TD', 'Fantasy', 'FD_Fantasy']]
+    gamelog_table['rebound%'] = gamelog_table['REB'] / gamelog_table['reboundChancesTotal']
+    gamelog_table['assists_per_pass'] = gamelog_table['assists'] / gamelog_table['passes']
+    gamelog_table['Fantasy_per_touch'] = gamelog_table['Fantasy'] / gamelog_table['touches']
+    gamelog_table['FD_Fantasy_per_touch'] = gamelog_table['FD_Fantasy'] / gamelog_table['touches']
     
-    dk_minutes_table = trend_table[['PLAYER_NAME', 'Team', 'L10 MIN', 'L5 MIN', 'L3 MIN', 'Trend Min']]
-    
-    fd_minutes_table = trend_table[['PLAYER_NAME', 'Team', 'L10 MIN', 'L5 MIN', 'L3 MIN', 'Trend Min']]
-    
-    dk_medians_table = trend_table[['PLAYER_NAME', 'Team', 'L10 Fantasy', 'L5 Fantasy', 'L3 Fantasy', 'Trend Median']]
-    
-    fd_medians_table = trend_table[['PLAYER_NAME', 'Team', 'L10 FD_Fantasy', 'L5 FD_Fantasy', 'L3 FD_Fantasy', 'Trend FD_Median']]
-    
-    dk_proj_medians_table = trend_table[['PLAYER_NAME', 'Team', 'Position', 'DK_Salary', 'DK_Proj', 'Adj Median', 'DK_Avg_Val', 'Adj Ceiling', 'DK_Ceiling_Value']]
-    
-    fd_proj_medians_table = trend_table[['PLAYER_NAME', 'Team', 'FD_Position', 'FD_Salary', 'FD_Proj', 'Adj FD_Median', 'FD_Avg_Val', 'Adj FD_Ceiling', 'FD_Ceiling_Value']]
+    season_long_table = gamelog_table[['PLAYER_NAME', 'TEAM_NAME']]
+    season_long_table['MIN'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['MIN'].transform('mean').astype(float)
+    season_long_table['touches'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['touches'].transform('mean').astype(float)
+    season_long_table['FGM'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FGM'].transform('mean').astype(float)
+    season_long_table['FGA'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FGA'].transform('mean').astype(float)
+    season_long_table['FG_PCT'] = (gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FGM'].transform('sum').astype(int) /
+                                   gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FGA'].transform('sum').astype(int))
+    season_long_table['FG3M'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FG3M'].transform('mean').astype(float)
+    season_long_table['FG3A'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FG3A'].transform('mean').astype(float)
+    season_long_table['FG3_PCT'] = (gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FG3M'].transform('sum').astype(int) /
+                                   gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FG3A'].transform('sum').astype(int))
+    season_long_table['FTM'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FTM'].transform('mean').astype(float)
+    season_long_table['FTA'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FTA'].transform('mean').astype(float)
+    season_long_table['FT_PCT'] = (gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FTM'].transform('sum').astype(int) /
+                                   gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FTA'].transform('sum').astype(int))
+    season_long_table['reboundChancesOffensive'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['reboundChancesOffensive'].transform('mean').astype(float)
+    season_long_table['OREB'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['OREB'].transform('mean').astype(float)
+    season_long_table['reboundChancesDefensive'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['reboundChancesDefensive'].transform('mean').astype(float)
+    season_long_table['DREB'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['DREB'].transform('mean').astype(float)
+    season_long_table['reboundChancesTotal'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['reboundChancesTotal'].transform('mean').astype(float)
+    season_long_table['REB'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['REB'].transform('mean').astype(float)
+    season_long_table['passes'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['passes'].transform('mean').astype(float)
+    season_long_table['secondaryAssists'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['secondaryAssists'].transform('mean').astype(float)
+    season_long_table['freeThrowAssists'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['freeThrowAssists'].transform('mean').astype(float)
+    season_long_table['assists'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['assists'].transform('mean').astype(float)
+    season_long_table['STL'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['STL'].transform('mean').astype(float)
+    season_long_table['BLK'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['BLK'].transform('mean').astype(float)
+    season_long_table['TOV'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['TOV'].transform('mean').astype(float)
+    season_long_table['PF'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['PF'].transform('mean').astype(float)
+    season_long_table['DD'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['DD'].transform('mean').astype(float)
+    season_long_table['TD'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['TD'].transform('mean').astype(float)
+    season_long_table['Fantasy'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['Fantasy'].transform('mean').astype(float)
+    season_long_table['FD_Fantasy'] = gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FD_Fantasy'].transform('mean').astype(float)
+    season_long_table['rebound%'] = (gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['REB'].transform('sum').astype(int) /
+                                     gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['reboundChancesTotal'].transform('sum').astype(int))
+    season_long_table['assists_per_pass'] = (gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['assists'].transform('sum').astype(int) /
+                                             gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['passes'].transform('sum').astype(int))
+    season_long_table['Fantasy_per_touch'] = (gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['Fantasy'].transform('sum').astype(int) /
+                                              gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['touches'].transform('sum').astype(int))
+    season_long_table['FD_Fantasy_per_touch'] = (gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['FD_Fantasy'].transform('sum').astype(int) /
+                                                 gamelog_table.groupby(['PLAYER_NAME', 'SEASON_ID'], sort=False)['touches'].transform('sum').astype(int))
+    season_long_table = season_long_table.drop_duplicates(subset='PLAYER_NAME')
+
+    return gamelog_table, season_long_table
 
-    return trend_table, dk_minutes_table, fd_minutes_table, dk_medians_table, fd_medians_table, dk_proj_medians_table, fd_proj_medians_table
+@st.cache_data(show_spinner=False)
+def split_frame(input_df, rows):
+    df = [input_df.loc[i : i + rows - 1, :] for i in range(0, len(input_df), rows)]
+    return df
 
 def convert_df_to_csv(df):
     return df.to_csv().encode('utf-8')
 
-trend_table, dk_minutes_table, fd_minutes_table, dk_medians_table, fd_medians_table, dk_proj_medians_table, fd_proj_medians_table = init_baselines()
+gamelog_table, season_long_table = init_baselines()
 
 col1, col2 = st.columns([1, 9])
 with col1:
     if st.button("Reset Data", key='reset1'):
               st.cache_data.clear()
-              trend_table, dk_minutes_table, fd_minutes_table, dk_medians_table, fd_medians_table, dk_proj_medians_table, fd_proj_medians_table = init_baselines()
-    split_var1 = st.radio("What table would you like to view?", ('Minutes Trends', 'Fantasy Trends', 'Slate specific', 'Overall'), key='split_var1')
-    site_var1 = st.radio("What site would you like to view?", ('Draftkings', 'Fanduel'), key='site_var1')
-    if site_var1 == 'Draftkings':
-        trend_table = trend_table[['PLAYER_NAME', 'Team', 'Position', 'L10 MIN', 'L10 Fantasy', 'L10 Ceiling',
-                                   'L5 MIN', 'L5 Fantasy', 'L5 Ceiling', 'L3 MIN', 'L3 Fantasy',
-                                   'L3 Ceiling', 'Trend Min', 'Trend Median', 'DK_Proj', 'Adj Median', 'Adj Ceiling',
-                                   'DK_Salary', 'DK_Avg_Val', 'DK_Ceiling_Value']]
-        minutes_table = dk_minutes_table
-        medians_table = dk_medians_table
-        proj_medians_table = dk_proj_medians_table
-    elif site_var1 == 'Fanduel':
-        trend_table = trend_table[['PLAYER_NAME', 'Team', 'FD_Position', 'L10 MIN', 'L10 FD_Fantasy',
-                                   'L10 FD_Ceiling', 'L5 MIN', 'L5 FD_Fantasy', 'L5 FD_Ceiling', 'L3 MIN', 'L3 FD_Fantasy',
-                                   'L3 FD_Ceiling', 'Trend Min', 'Trend FD_Median', 'FD_Proj', 'Adj FD_Median', 'Adj FD_Ceiling',
-                                   'FD_Salary', 'FD_Avg_Val', 'FD_Ceiling_Value']]
-        minutes_table = fd_minutes_table
-        medians_table = fd_medians_table
-        proj_medians_table = fd_proj_medians_table
-    trend_table = trend_table.set_axis(['PLAYER_NAME', 'Team', 'Position', 'L10 MIN', 'L10 Fantasy', 'L10 Ceiling',
-                                        'L5 MIN', 'L5 Fantasy', 'L5 Ceiling', 'L3 MIN', 'L3 Fantasy',
-                                        'L3 Ceiling', 'Trend Min', 'Trend Median', 'Proj', 'Adj Median', 'Adj Ceiling',
-                                        'Salary', 'Avg_Val', 'Ceiling_Value'], axis=1)
-    minutes_table = minutes_table.set_axis(['PLAYER_NAME', 'Team', 'L10 MIN', 'L5 MIN', 'L3 MIN', 'Trend Min'], axis=1)
-    medians_table = medians_table.set_axis(['PLAYER_NAME', 'Team', 'L10 Fantasy','L5 Fantasy', 'L3 Fantasy', 'Trend Median'], axis=1)
-    proj_medians_table = proj_medians_table.set_axis(['PLAYER_NAME', 'Team', 'Position', 'Salary', 'Proj',
-                                                      'Adj Median', 'Avg_Val', 'Adj Ceiling', 'Ceiling_Value'], axis=1)
-    if split_var1 == 'Overall':
-        view_var1 = trend_table.Team.values.tolist()
-        split_var2 = st.radio("Would you like to view all teams or specific ones?", ('All', 'Specific Teams'), key='split_var2')
-        
-        if split_var2 == 'Specific Teams':
-            team_var1 = st.multiselect('Which teams would you like to include in the tables?', options = view_var1, key='team_var1')
-        elif split_var2 == 'All':
-            team_var1 = view_var1
-        
-        split_var3 = st.radio("Would you like to view all positions or specific ones?", ('All', 'Specific Positions'), key='split_var3')
-        if split_var3 == 'Specific Positions':
-            pos_var1 = st.multiselect('Which positions would you like to include in the tables?', options = ['PG', 'SG', 'SF', 'PF', 'C'], key='pos_var1')
-        elif split_var3 == 'All':
-            pos_var1 = ['PG', 'SG', 'SF', 'PF', 'C']
-        
-        proj_var1 = st.slider("Is there a certain projection range you want to view?", 0, 100, (10, 100), key='proj_var1')
-            
-    elif split_var1 == 'Minutes Trends':
-        view_var1 = trend_table.Team.values.tolist()
-        split_var2 = st.radio("Would you like to view all teams or specific ones?", ('All', 'Specific Teams'), key='split_var2')
-        
-        if split_var2 == 'Specific Teams':
-            team_var1 = st.multiselect('Which teams would you like to include in the tables?', options = view_var1, key='team_var1')
-        elif split_var2 == 'All':
-            team_var1 = view_var1
-        
-    elif split_var1 == 'Fantasy Trends':
-        view_var1 = trend_table.Team.values.tolist()
-        split_var2 = st.radio("Would you like to view all teams or specific ones?", ('All', 'Specific Teams'), key='split_var2')
-        
-        if split_var2 == 'Specific Teams':
-            team_var1 = st.multiselect('Which teams would you like to include in the tables?', options = view_var1, key='team_var1')
-        elif split_var2 == 'All':
-            team_var1 = view_var1
-        
-    elif split_var1 == 'Slate specific':
-        view_var1 = trend_table.Team.values.tolist()
-        split_var2 = st.radio("Would you like to view all teams or specific ones?", ('All', 'Specific Teams'), key='split_var2')
-        
-        if split_var2 == 'Specific Teams':
-            team_var1 = st.multiselect('Which teams would you like to include in the tables?', options = view_var1, key='team_var1')
-        elif split_var2 == 'All':
-            team_var1 = view_var1
-        
-        split_var3 = st.radio("Would you like to view all positions or specific ones?", ('All', 'Specific Positions'), key='split_var3')
-        if split_var3 == 'Specific Positions':
-            pos_var1 = st.multiselect('Which positions would you like to include in the tables?', options = ['PG', 'SG', 'SF', 'PF', 'C'], key='pos_var1')
-        elif split_var3 == 'All':
-            pos_var1 = ['PG', 'SG', 'SF', 'PF', 'C']
-        
-        proj_var1 = st.slider("Is there a certain projection range you want to view?", 0, 100, (10, 100), key='proj_var1')
+              gamelog_table, season_long_table = init_baselines()
+    split_var1 = st.radio("What table would you like to view?", ('Season Long', 'Gamelogs'), key='split_var1')
 
 with col2:
-    if split_var1 == 'Overall':
-        table_display = trend_table[trend_table['Proj'] >= proj_var1[0]]
-        table_display = table_display[table_display['Proj'] <= proj_var1[1]]
-        table_display = table_display[table_display['Team'].isin(team_var1)]
-        table_display = table_display[table_display['Position'].str.contains('|'.join(pos_var1))]
-        table_display = table_display.sort_values(by='Adj Ceiling', ascending=False)
-        table_display = table_display.set_index('PLAYER_NAME')
-        st.dataframe(table_display.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').format(precision=2), use_container_width = True)
-        st.download_button(
-            label="Export Trending Numbers",
-            data=convert_df_to_csv(table_display),
-            file_name='Trending_export.csv',
-            mime='text/csv',
-        )
-            
-    elif split_var1 == 'Minutes Trends':
-        table_display = minutes_table[minutes_table['Team'].isin(team_var1)]
-        table_display = table_display.set_index('PLAYER_NAME')
-        st.dataframe(table_display.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').format(precision=2), use_container_width = True)
-        st.download_button(
-            label="Export Trending Numbers",
-            data=convert_df_to_csv(table_display),
-            file_name='Trending_export.csv',
-            mime='text/csv',
-        )
-        
-    elif split_var1 == 'Fantasy Trends':
-        table_display = medians_table[medians_table['Team'].isin(team_var1)]
-        table_display = table_display.set_index('PLAYER_NAME')
-        st.dataframe(table_display.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').format(precision=2), use_container_width = True)
-        st.download_button(
-            label="Export Trending Numbers",
-            data=convert_df_to_csv(table_display),
-            file_name='Trending_export.csv',
-            mime='text/csv',
-        )
-        
-    elif split_var1 == 'Slate specific':
-        table_display = proj_medians_table[proj_medians_table['Proj'] >= proj_var1[0]]
-        table_display = table_display[table_display['Proj'] <= proj_var1[1]]
-        table_display = table_display[table_display['Team'].isin(team_var1)]
-        table_display = table_display[table_display['Position'].str.contains('|'.join(pos_var1))]
-        table_display = table_display.sort_values(by='Adj Ceiling', ascending=False)
-        table_display = table_display.set_index('PLAYER_NAME')
-        st.dataframe(table_display.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').format(precision=2), use_container_width = True)
-        st.download_button(
-            label="Export Trending Numbers",
-            data=convert_df_to_csv(table_display),
-            file_name='NBA_Trending_export.csv',
-            mime='text/csv',
-        )
+    if split_var1 == 'Season Logs':
+        st.dataframe(season_long_table.style.background_gradient(axis=0).background_gradient(cmap='RdYlGn').format(precision=2), use_container_width = True)  
+    elif split_var1 == 'Gamelogs':
+        pagination = st.container()
+    
+        bottom_menu = st.columns((4, 1, 1))
+        with bottom_menu[2]:
+            batch_size = st.selectbox("Page Size", options=[25, 50, 100])
+        with bottom_menu[1]:
+            total_pages = (
+                int(len(gamelog_table) / batch_size) if int(len(gamelog_table) / batch_size) > 0 else 1
+            )
+            current_page = st.number_input(
+                "Page", min_value=1, max_value=total_pages, step=1
+            )
+        with bottom_menu[0]:
+            st.markdown(f"Page **{current_page}** of **{total_pages}** ")
+    
+        pages = split_frame(gamelog_table, batch_size)
+        pagination.dataframe(data=pages[current_page - 1], use_container_width=True)