Add middle-middle location% and team filtering

app.py

@@ -4,6 +4,8 @@ from data import data_df
 from pitcher_overview import create_pitcher_overview
 from pitch_leaderboard import create_pitch_leaderboard
 
+updated = '2025-07-21'
+
 if __name__ == '__main__':
   with gr.Blocks() as app:
     with gr.Tab('Pitcher Overview'):
@@ -11,5 +13,5 @@ if __name__ == '__main__':
     with gr.Tab('Pitch Leaderboard'):
       create_pitch_leaderboard()
 
-    gr.Markdown('Last updated: 2025-07-19')
+    gr.Markdown(f'Last updated: {updated}')
   app.launch()

data.py

@@ -230,6 +230,8 @@ data_df = (
         .then(pl.lit('Postseason'))
         .otherwise('GameKindName')
         .alias('coarse_game_kind'),
+
+        pl.when(pl.col('half_inning').str.ends_with(1)).then('HomeTeamNameES').otherwise('VisitorTeamNameES').alias('pitcher_team')
     )
     .with_columns(
         pl.col('presult_id').replace_strict(presult).alias('presult')

pitch_leaderboard.py

@@ -8,16 +8,21 @@ from data import data_df
 from stats import compute_pitch_stats, filter_data_by_date_and_game_kind
 from convert import ball_kind
 
-STATS = ['Count', 'Usage', 'Swing%', 'Z-Swing%', 'Chase%', 'Contact%', 'Z-Contact%', 'O-Contact%', 'SwStr%', 'Whiff%', 'CSW%', 'GB%', 'FB%', 'LD%', 'Zone%', 'Arm%', 'Glove%', 'High%', 'Low%']
-PCT_STATS = ['Usage', 'Swing%', 'Z-Swing%', 'Chase%', 'Contact%', 'Z-Contact%', 'O-Contact%', 'SwStr%', 'Whiff%', 'CSW%', 'GB%', 'FB%', 'LD%', 'Zone%', 'Arm%', 'Glove%', 'High%', 'Low%']
+STATS = ['Count', 'Usage', 'Swing%', 'Z-Swing%', 'Chase%', 'Contact%', 'Z-Contact%', 'O-Contact%', 'SwStr%', 'Whiff%', 'CSW%', 'GB%', 'FB%', 'LD%', 'Zone%', 'Arm%', 'Glove%', 'High%', 'Low%', 'MM%']
+PCT_STATS = ['Usage', 'Swing%', 'Z-Swing%', 'Chase%', 'Contact%', 'Z-Contact%', 'O-Contact%', 'SwStr%', 'Whiff%', 'CSW%', 'GB%', 'FB%', 'LD%', 'Zone%', 'Arm%', 'Glove%', 'High%', 'Low%', 'MM%']
 STATS_WITH_PCTLS = ['Swing%', 'Z-Swing%', 'Chase%', 'Contact%', 'Z-Contact%', 'O-Contact%', 'SwStr%', 'Whiff%', 'CSW%', 'GB%', 'FB%', 'LD%', 'Zone%']
 
-todo = '''
+PITCH_TYPES = [pitch_type for pitch_type in ball_kind.values() if pitch_type != '-']
+TEAMS  = ['G', 'S', 'DB', 'D', 'T', 'C', 'F', 'E', 'L', 'M', 'B', 'H']
+notes = '''**Limitations**
+- Foreign players names are in Hebpurn romanization.
+
 **To-do**
 - Color cells according to percentiles
 '''
 
-def gr_create_pitch_leaderboard(start_date, end_date, min_pitches, pitcher_lr, include_pitches):
+
+def gr_create_pitch_leaderboard(start_date, end_date, min_pitches, pitcher_lr='Both', include_pitches=PITCH_TYPES, include_teams=None):
   assert pitcher_lr in ['Both', 'Left', 'Right']
 
   data = data_df.filter(pl.col('ballKind_code') != '-')
@@ -25,6 +30,9 @@ def gr_create_pitch_leaderboard(start_date, end_date, min_pitches, pitcher_lr, i
   data = filter_data_by_date_and_game_kind(data, start_date=start_date, end_date=end_date, game_kind='Regular Season')
   if pitcher_lr != 'Both':
     data = data.filter(pl.col('batLR') == pitcher_lr[0].lower())
+
+  if include_teams is not None:
+    data = data.filter(pl.col('pitcher_team').is_in(include_teams))
     
   # both, left, right = [
     # (
@@ -51,12 +59,12 @@ def gr_create_pitch_leaderboard(start_date, end_date, min_pitches, pitcher_lr, i
         compute_pitch_stats(data, player_type='pitcher', min_pitches=min_pitches, pitch_class_type='specific')
         .filter(pl.col('qualified') & (pl.col('ballKind').is_in(include_pitches)))
         .drop('pitId', 'ballKind_code', 'qualified')
-        .rename({'pitcher_name': 'Pitcher', 'count': 'Count', 'usage': 'Usage', 'ballKind': 'Pitch', 'general_ballKind': 'Pitch (General)'} | {f'{stat}_pctl': f'{stat} (Pctl)' for stat in STATS_WITH_PCTLS})
+        .rename({'pitcher_name': 'Pitcher', 'pitcher_team': 'Team', 'count': 'Count', 'usage': 'Usage', 'ballKind': 'Pitch', 'general_ballKind': 'Pitch (General)'} | {f'{stat}_pctl': f'{stat} (Pctl)' for stat in STATS_WITH_PCTLS})
         .with_columns(
           pl.col(stat).mul(100).round(1)
           for stat in PCT_STATS + [f'{stat} (Pctl)' for stat in STATS_WITH_PCTLS]
         )
-        [['Pitcher', 'Pitch', 'Pitch (General)'] + STATS]
+        [['Pitcher', 'Team', 'Pitch', 'Pitch (General)'] + STATS]
   )
   return pitch_stats
 
@@ -65,33 +73,38 @@ def create_pitch_leaderboard():
   now = datetime.now()
   start_datetime_init = datetime(now.year, 1, 1)
   end_datetime_init = now
-  pitch_types = [pitch_type for pitch_type in ball_kind.values() if pitch_type != '-']
   with gr.Blocks() as app:
     gr.Markdown('# Pitch Leaderboard')
     with gr.Row():
       start_date = gr.DateTime(start_datetime_init, include_time=False, type='datetime', label='Start')
       end_date = gr.DateTime(end_datetime_init, include_time=False, type='datetime', label='End')
     with gr.Row():
-      include_pitches = gr.CheckboxGroup(pitch_types, value=pitch_types, label='Pitches', scale=3)
+      include_pitches = gr.CheckboxGroup(PITCH_TYPES, value=PITCH_TYPES, label='Pitches', scale=3)
       with gr.Column(scale=1):
-        all_pitches = gr.Button('Select/Deselect all pitches', scale=1)
+        all_pitches = gr.Button('Select/Deselect all pitches')
         min_pitches = gr.Number(100, label='Min. Pitches', precision=0, minimum=0)
         pitcher_lr = gr.Radio(['Both', 'Left', 'Right'], value='Both', label='Batter handedness')
+    with gr.Row():
+      include_teams = gr.CheckboxGroup(TEAMS, value=TEAMS, label='Teams', scale=3)
+      all_teams = gr.Button('Select/Deselect all teams')
         
     search = gr.Button('Search')
+    # pin_columns = gr.Checkbox(True, 'Pin columns')
     leaderboard = gr.DataFrame(
       pl.DataFrame({'Pitcher': [], 'Pitch': []}),
-      # gr_create_pitch_leaderboard(start_date=start_date_init, end_date=end_date_init, min_pitches=100),
-      column_widths=[200]*3 + [100]*len(STATS),
+      column_widths=[200, 60, 200, 200] + [100]*len(STATS),
       show_copy_button=True,
       show_search=True,
-      pinned_columns=2
+      pinned_columns=3
     )
 
+    gr.Markdown(notes)
 
-    search.click(gr_create_pitch_leaderboard, inputs=[start_date, end_date, min_pitches, pitcher_lr, include_pitches], outputs=leaderboard)
-    all_pitches.click(lambda _pitch_types : [] if _pitch_types == pitch_types else pitch_types, inputs=include_pitches, outputs=include_pitches)
-    gr.Markdown(todo)
+    search.click(gr_create_pitch_leaderboard, inputs=[start_date, end_date, min_pitches, pitcher_lr, include_pitches, include_teams], outputs=leaderboard)
+    all_pitches.click(lambda _pitch_types : [] if _pitch_types == PITCH_TYPES else PITCH_TYPES, inputs=include_pitches, outputs=include_pitches)
+    all_teams.click(lambda _teams : [] if _teams == TEAMS else TEAMS, inputs=include_teams, outputs=include_teams)
+    # pin_columns.input(lambda _pin_columns : (gr.update(pinned_columns=None if _pin_columns else 3), not _pin_columns), inputs=pin_columns, outputs=[leaderboard, pin_columns])
+    
   return app
 
 if __name__ == '__main__':

stats.py

@@ -60,7 +60,7 @@ def compute_pitch_stats(data, player_type, pitch_class_type, min_pitches=1):
   pitch_name_col = 'ballKind' if pitch_class_type == 'specific' else 'general_ballKind'
   pitch_stats = (
       data
-      .group_by(id_col, pitch_col)
+      .group_by(id_col, pitch_col, 'pitcher_team')
       .agg(
           pl.first('pitcher_name'),
           *([pl.first('general_ballKind')] if pitch_class_type == 'specific' else []),
@@ -80,7 +80,8 @@ def compute_pitch_stats(data, player_type, pitch_class_type, min_pitches=1):
           (pl.when(pl.col('pitLR') == 'r').then(pl.col('x') < 0).otherwise(pl.col('x') > 0)).mean().alias('Glove%'),
           (pl.when(pl.col('pitLR') == 'r').then(pl.col('x') >= 0).otherwise(pl.col('x') <= 0)).mean().alias('Arm%'),
           (pl.col('y') > 125).mean().alias('High%'),
-          (pl.col('y') <= 125).mean().alias('Low%')
+          (pl.col('y') <= 125).mean().alias('Low%'),
+          (pl.col('x').is_between(-20, 20) & pl.col('y').is_between(100, 100+50)).mean().alias('MM%')
       )
       .with_columns(
           (pl.col('count')/pl.sum('count').over('pitId')).alias('usage'),