Upload app.py

app.py

@@ -17,7 +17,7 @@ colour_palette = ['#FFB000','#648FFF','#785EF0',
 import datasets
 from datasets import load_dataset
 ### Import Datasets
-dataset = load_dataset('nesticot/mlb_data', data_files=['mlb_pitch_data_2024.csv' ])
+dataset = load_dataset('nesticot/mlb_data', data_files=['aaa_pitch_data_2024.csv' ])
 dataset_train = dataset['train']
 df_2024 = dataset_train.to_pandas().set_index(list(dataset_train.features.keys())[0]).reset_index(drop=True).drop_duplicates(subset=['play_id'],keep='last')
 
@@ -59,17 +59,17 @@ spring_teams = df_2024.groupby(['pitcher_id']).tail(1)[['pitcher_id','pitcher_te
 season_start = '2024-03-20'
 season_end = '2024-09-29'
 season_fg=2024 
-chad_fg = requests.get(f'https://www.fangraphs.com/api/leaders/major-league/data?age=&pos=all&stats=pit&lg=all&qual=0&season={season_fg}&season={season_fg}&month=1000&season1={season_fg}&ind=0&pageitems=2000000000&pagenum=1&ind=0&rost=0&players=&type=36&postseason=&sortdir=default&sortstat=sp_pitching').json()
+#chad_fg = requests.get(f'https://www.fangraphs.com/api/leaders/major-league/data?age=&pos=all&stats=pit&lg=all&qual=0&season={season_fg}&season={season_fg}&month=1000&season1={season_fg}&ind=0&pageitems=2000000000&pagenum=1&ind=0&rost=0&players=&type=36&postseason=&sortdir=default&sortstat=sp_pitching').json()
 
 
-chadwick_df_small = pd.DataFrame(data={
-'key_mlbam':[x['xMLBAMID'] for x in chad_fg['data']],
-'key_fangraphs':[x['playerid'] for x in chad_fg['data']],
-'Name':[x['PlayerName'] for x in chad_fg['data']],
-})    
+# chadwick_df_small = pd.DataFrame(data={
+# 'key_mlbam':[x['xMLBAMID'] for x in chad_fg['data']],
+# 'key_fangraphs':[x['playerid'] for x in chad_fg['data']],
+# 'Name':[x['PlayerName'] for x in chad_fg['data']],
+# })    
 
-pitcher_dicts = chadwick_df_small.set_index('key_mlbam')['Name'].sort_values().to_dict()
-mlb_fg_dicts = chadwick_df_small.set_index('key_mlbam')['key_fangraphs'].sort_values().to_dict()
+pitcher_dicts = df_2024.set_index('pitcher_id')['pitcher_name'].sort_values().to_dict()
+# mlb_fg_dicts = chadwick_df_small.set_index('key_mlbam')['key_fangraphs'].sort_values().to_dict()
 
 
 statcast_pitch_summary = pd.read_csv('statcast_pitch_summary.csv')
@@ -403,16 +403,50 @@ def server(input, output, session):
         font_properties_titles = {'family': 'calibi', 'size': 20}
         font_properties_axes = {'family': 'calibi', 'size': 16}
 
-        ## FANGRAPHS TABLE ###
-        data_pull = psf.fangraphs_scrape(pitcher_id=pitcher_id_select,
-                            split=input.split_id(),
-                            start_date=input.date_range_id()[0],
-                            end_date=input.date_range_id()[1])
+        # ## FANGRAPHS TABLE ###
+        # data_pull = psf.fangraphs_scrape(pitcher_id=pitcher_id_select,
+        #                     split=input.split_id(),
+        #                     start_date=input.date_range_id()[0],
+        #                     end_date=input.date_range_id()[1])
+
+
+        # psf.fangraphs_table(data=data_pull,
+        #                     stats=['IP','WHIP','ERA','FIP','TBF','K%','BB%','K-BB%'],
+        #                     ax=ax1_table)
+
+        start_date = str(pd.to_datetime(input.date_range_id()[0]).strftime('%m/%d/%Y'))
+        end_date = str(pd.to_datetime(input.date_range_id()[1]).strftime('%m/%d/%Y'))
+
+
+        pitcher_stats_call = requests.get(f'https://statsapi.mlb.com/api/v1/people/{pitcher_id_select}?appContext=minorLeague&hydrate=stats(group=[pitching],type=[byDateRange],sportId=11,startDate={start_date},endDate={end_date})').json()
+
+        pitcher_stats_call_header = [x for x in pitcher_stats_call['people'][0]['stats'][0]['splits'][0]['stat']]
+        pitcher_stats_call_values = [pitcher_stats_call['people'][0]['stats'][0]['splits'][0]['stat'][x] for x in pitcher_stats_call['people'][0]['stats'][0]['splits'][0]['stat']]
+        pitcher_stats_call_df = pd.DataFrame(data=dict(zip(pitcher_stats_call_header,pitcher_stats_call_values)),index=[0])
+        pitcher_stats_call_df['k_percent'] = pitcher_stats_call_df['strikeOuts']/pitcher_stats_call_df['battersFaced']
+        pitcher_stats_call_df['bb_percent'] = pitcher_stats_call_df['baseOnBalls']/pitcher_stats_call_df['battersFaced']
+        pitcher_stats_call_df['k_bb_percent'] = pitcher_stats_call_df['k_percent']-pitcher_stats_call_df['bb_percent']
+        pitcher_stats_call_df_small = pitcher_stats_call_df[['inningsPitched','battersFaced','era','whip','k_percent','bb_percent','k_bb_percent']]
+
+        pitcher_stats_call_df_small['k_percent'] = pitcher_stats_call_df_small['k_percent'].astype(float).apply(lambda x: '{:.1%}'.format(x))  
+        pitcher_stats_call_df_small['bb_percent'] = pitcher_stats_call_df_small['bb_percent'].astype(float).apply(lambda x: '{:.1%}'.format(x))  
+        pitcher_stats_call_df_small['k_bb_percent'] = pitcher_stats_call_df_small['k_bb_percent'].astype(float).apply(lambda x: '{:.1%}'.format(x)) 
+
+        table_fg = ax1_table.table(cellText=pitcher_stats_call_df_small.values, colLabels=pitcher_stats_call_df_small.columns, cellLoc='center',
+                        bbox=[0.04, 0.2, 0.92, 0.8])
+
+        min_font_size = 20
+        table_fg.set_fontsize(min_font_size)
+
+
+        new_column_names = ['$\\bf{IP}$','$\\bf{PA}$','$\\bf{ERA}$','$\\bf{WHIP}$','$\\bf{K\%}$','$\\bf{BB\%}$','$\\bf{K-BB\%}$']
+        # #new_column_names = ['Pitch Name', 'Pitch%', 'Velocity', 'Spin Rate','Exit Velocity', 'Whiff%', 'CSW%']
+        for i, col_name in enumerate(new_column_names):
+            table_fg.get_celld()[(0, i)].get_text().set_text(col_name)
+
+        ax1_table.axis('off')
 
 
-        psf.fangraphs_table(data=data_pull,
-                            stats=['IP','WHIP','ERA','FIP','TBF','K%','BB%','K-BB%'],
-                            ax=ax1_table)
 
         for x,y,z in zip([input.plot_id_1(),input.plot_id_2(),input.plot_id_3()],[ax2_left,ax2_middle,ax2_right],[1,2,3]):
             if x == 'velocity_kde':
@@ -481,17 +515,21 @@ def server(input, output, session):
 
 
         ax0.text(x=0.5,y=0.8,s=title_spot,fontname='Calibri',ha='center',fontsize=56,va='top')
-        ax0.text(x=0.5,y=0.5,s='Season Pitching Summary',fontname='Calibri',ha='center',fontsize=40,va='top',fontstyle='italic')
+        ax0.text(x=0.5,y=0.35,s='AAA Season Pitching Summary',fontname='Calibri',ha='center',fontsize=40,va='top',fontstyle='italic')
 
+        player_bio = requests.get(url=f"https://statsapi.mlb.com/api/v1/people?personIds={pitcher_id_select}&hydrate=currentTeam").json()
+        #ax0.text(x=0.5,y=0.05,s=f'{ball_title}{strike_title}{split_title}',fontname='Calibri',ha='center',fontsize=20,va='top')
+        ax0.axis('off')
+        ax0.text(x=0.5,y=0.5,s=f"{ player_bio['people'][0]['pitchHand']['code']}HP, Age: {player_bio['people'][0]['currentAge']}, {player_bio['people'][0]['height']}/{player_bio['people'][0]['weight']}",fontname='Calibri',ha='center',fontsize=24,va='top')
 
         #ax0.text(x=0.5,y=0.25,s=f'2024 Spring Training',fontname='Calibri',ha='center',fontsize=30,va='top')
         # ax0.text(x=0.5,y=0.25,s=f'{season_fg} MLB Season',fontname='Calibri',ha='center',fontsize=30,va='top')
         # ax0.axis('off')
 
 
-        ax0.text(x=0.5,y=0.25,s=f'{input.date_range_id()[0]} to {input.date_range_id()[1]}',fontname='Calibri',ha='center',fontsize=30,va='top',fontstyle='italic')
+        ax0.text(x=0.5,y=0.15,s=f'{input.date_range_id()[0]} to {input.date_range_id()[1]}',fontname='Calibri',ha='center',fontsize=30,va='top',fontstyle='italic')
         
-        ax0.text(x=0.5,y=0.05,s=f'{ball_title}{strike_title}{split_title}',fontname='Calibri',ha='center',fontsize=20,va='top')
+        ax0.text(x=0.5,y=0.0,s=f'{ball_title}{strike_title}{split_title}',fontname='Calibri',ha='center',fontsize=20,va='top')
         ax0.axis('off')
 
         from matplotlib.offsetbox import (OffsetImage, AnnotationBbox)
@@ -501,15 +539,16 @@ def server(input, output, session):
         from urllib.error import HTTPError
 
         try: 
-            url = f'https://img.mlbstatic.com/mlb-photos/image/upload/d_people:generic:headshot:67:current.png/w_213,q_auto:best/v1/people/{pitcher_id_select}/headshot/67/current.png'
+            url = f'https://img.mlbstatic.com/mlb-photos/image/upload/c_fill,g_auto/w_180/v1/people/{pitcher_id_select}/headshot/milb/current.png'
             test_mage = plt.imread(url)
         except urllib.error.HTTPError as err:
             url = f'https://img.mlbstatic.com/mlb-photos/image/upload/d_people:generic:headshot:67:current.png/w_213,q_auto:best/v1/people/1/headshot/67/current.png'                
+            test_mage = plt.imread(url)    
         imagebox = OffsetImage(test_mage, zoom = 0.5)
         ab = AnnotationBbox(imagebox, (0.125, 0.4), frameon = False)
         ax0.add_artist(ab)
 
-        player_bio = requests.get(url=f"https://statsapi.mlb.com/api/v1/people?personIds={pitcher_id_select}&hydrate=currentTeam").json()
+        #player_bio = requests.get(url=f"https://statsapi.mlb.com/api/v1/people?personIds={pitcher_id_select}&hydrate=currentTeam").json()
 
 
 
@@ -667,16 +706,49 @@ def server(input, output, session):
         font_properties_axes = {'family': 'calibi', 'size': 16}
 
         print(df_2024_update['game_date'].values[0])
-        ## FANGRAPHS TABLE ###
-        data_pull = psf.fangraphs_scrape(pitcher_id=pitcher_id_select,
-                            split=input.split_id(),
-                            start_date=df_plot['game_date'].values[0],
-                            end_date=df_plot['game_date'].values[0])
+        # ## FANGRAPHS TABLE ###
+        # data_pull = psf.fangraphs_scrape(pitcher_id=pitcher_id_select,
+        #                     split=input.split_id(),
+        #                     start_date=df_plot['game_date'].values[0],
+        #                     end_date=df_plot['game_date'].values[0])
+
+        start_date = str(pd.to_datetime(df_plot['game_date'].values[0]).strftime('%m/%d/%Y'))
+        end_date = str(pd.to_datetime(df_plot['game_date'].values[0]).strftime('%m/%d/%Y'))
+
 
+        pitcher_stats_call = requests.get(f'https://statsapi.mlb.com/api/v1/people/{pitcher_id_select}?appContext=minorLeague&hydrate=stats(group=[pitching],type=[byDateRange],sportId=11,startDate={start_date},endDate={end_date})').json()
+        
+        pitcher_stats_call_header = [x for x in pitcher_stats_call['people'][0]['stats'][0]['splits'][0]['stat']]
+        pitcher_stats_call_values = [pitcher_stats_call['people'][0]['stats'][0]['splits'][0]['stat'][x] for x in pitcher_stats_call['people'][0]['stats'][0]['splits'][0]['stat']]
+        pitcher_stats_call_df = pd.DataFrame(data=dict(zip(pitcher_stats_call_header,pitcher_stats_call_values)),index=[0])
+        # pitcher_stats_call_df['k_percent'] = pitcher_stats_call_df['strikeOuts']/pitcher_stats_call_df['battersFaced']
+        # pitcher_stats_call_df['bb_percent'] = pitcher_stats_call_df['baseOnBalls']/pitcher_stats_call_df['battersFaced']
+        # pitcher_stats_call_df['k_bb_percent'] = pitcher_stats_call_df['k_percent']-pitcher_stats_call_df['bb_percent']
+        pitcher_stats_call_df_small = pitcher_stats_call_df[['inningsPitched','battersFaced','earnedRuns','hits','strikeOuts','baseOnBalls','hitByPitch','homeRuns']]
+        pitcher_stats_call_df_small['whiffs'] = int(df_plot['is_whiff'].sum())
+        # pitcher_stats_call_df_small['k_percent'] = pitcher_stats_call_df_small['k_percent'].astype(float).apply(lambda x: '{:.1%}'.format(x))  
+        # pitcher_stats_call_df_small['bb_percent'] = pitcher_stats_call_df_small['bb_percent'].astype(float).apply(lambda x: '{:.1%}'.format(x))  
+        # pitcher_stats_call_df_small['k_bb_percent'] = pitcher_stats_call_df_small['k_bb_percent'].astype(float).apply(lambda x: '{:.1%}'.format(x)) 
+
+        table_fg = ax1_table.table(cellText=pitcher_stats_call_df_small.values, colLabels=pitcher_stats_call_df_small.columns, cellLoc='center',
+                        bbox=[0.04, 0.2, 0.92, 0.8])
+
+        min_font_size = 20
+        table_fg.set_fontsize(min_font_size)
+
+
+        new_column_names = ['$\\bf{IP}$','$\\bf{PA}$','$\\bf{ER}$','$\\bf{H}$','$\\bf{K}$','$\\bf{BB}$','$\\bf{HBP}$','$\\bf{HR}$','$\\bf{Whiffs}$']
+        # #new_column_names = ['Pitch Name', 'Pitch%', 'Velocity', 'Spin Rate','Exit Velocity', 'Whiff%', 'CSW%']
+        for i, col_name in enumerate(new_column_names):
+            table_fg.get_celld()[(0, i)].get_text().set_text(col_name)
+
+        ax1_table.axis('off')
+
+
+        # psf.fangraphs_table(data=data_pull,
+        #                     stats=['IP','WHIP','ERA','FIP','TBF','K%','BB%','K-BB%'],
+        #                     ax=ax1_table)
 
-        psf.fangraphs_table(data=data_pull,
-                            stats=['IP','WHIP','ERA','FIP','TBF','K%','BB%','K-BB%'],
-                            ax=ax1_table)
 
         # psf.velocity_kdes(df=df_plot,
         #                 ax=ax2_loc,
@@ -746,7 +818,7 @@ def server(input, output, session):
 
 
         ax0.text(x=0.5,y=0.8,s=title_spot,fontname='Calibri',ha='center',fontsize=56,va='top')
-        ax0.text(x=0.5,y=0.5,s='Game Pitching Summary',fontname='Calibri',ha='center',fontsize=40,va='top',fontstyle='italic')
+        ax0.text(x=0.5,y=0.35,s='AAA Game Pitching Summary',fontname='Calibri',ha='center',fontsize=40,va='top',fontstyle='italic')
 
 
         #ax0.text(x=0.5,y=0.25,s=f'2024 Spring Training',fontname='Calibri',ha='center',fontsize=30,va='top')
@@ -754,9 +826,9 @@ def server(input, output, session):
         #ax0.text(x=0.5,y=0.25,s=f'2024 Spring Training',fontname='Calibri',ha='center',fontsize=30,va='top')
         # ax0.text(x=0.5,y=0.25,s=f'{season_fg} MLB Season',fontname='Calibri',ha='center',fontsize=30,va='top')
 
-        ax0.text(x=0.5,y=0.25,s= df_plot['game_opp'].values[0],fontname='Calibri',ha='center',fontstyle='italic',fontsize=30,va='top')
+        ax0.text(x=0.5,y=0.15,s= df_plot['game_opp'].values[0],fontname='Calibri',ha='center',fontstyle='italic',fontsize=30,va='top')
         
-        ax0.text(x=0.5,y=0.05,s=f'{ball_title}{strike_title}{split_title}',fontname='Calibri',ha='center',fontsize=20,va='top')
+        ax0.text(x=0.5,y=0.00,s=f'{ball_title}{strike_title}{split_title}',fontname='Calibri',ha='center',fontsize=20,va='top')
         ax0.axis('off')
 
 
@@ -767,16 +839,21 @@ def server(input, output, session):
         from urllib.error import HTTPError
 
         try: 
-            url = f'https://img.mlbstatic.com/mlb-photos/image/upload/d_people:generic:headshot:67:current.png/w_213,q_auto:best/v1/people/{pitcher_id_select}/headshot/67/current.png'
+            url = f'https://img.mlbstatic.com/mlb-photos/image/upload/c_fill,g_auto/w_180/v1/people/{pitcher_id_select}/headshot/milb/current.png'
             test_mage = plt.imread(url)
         except urllib.error.HTTPError as err:
             url = f'https://img.mlbstatic.com/mlb-photos/image/upload/d_people:generic:headshot:67:current.png/w_213,q_auto:best/v1/people/1/headshot/67/current.png'                
+            test_mage = plt.imread(url)
         imagebox = OffsetImage(test_mage, zoom = 0.5)
         ab = AnnotationBbox(imagebox, (0.125, 0.4), frameon = False)
         ax0.add_artist(ab)
 
-        player_bio = requests.get(url=f"https://statsapi.mlb.com/api/v1/people?personIds={pitcher_id_select}&hydrate=currentTeam").json()
+        #player_bio = requests.get(url=f"https://statsapi.mlb.com/api/v1/people?personIds={pitcher_id_select}&hydrate=currentTeam").json()
 
+        player_bio = requests.get(url=f"https://statsapi.mlb.com/api/v1/people?personIds={pitcher_id_select}&hydrate=currentTeam").json()
+        #ax0.text(x=0.5,y=0.05,s=f'{ball_title}{strike_title}{split_title}',fontname='Calibri',ha='center',fontsize=20,va='top')
+        ax0.axis('off')
+        ax0.text(x=0.5,y=0.5,s=f"{ player_bio['people'][0]['pitchHand']['code']}HP, Age: {player_bio['people'][0]['currentAge']}, {player_bio['people'][0]['height']}/{player_bio['people'][0]['weight']}",fontname='Calibri',ha='center',fontsize=24,va='top')