Update app

data.py

@@ -153,6 +153,10 @@ df = (
     )
 ).sort(['game_pk', 'pa_pk', 'pitch_id'])
 
+# add players to pa_df
+# unfortunately we have pas that don't show up in the pitch data, so this would be useful for
+pa_df = pa_df.join(player_df.rename({'player_id': 'pitcher'}), on='pitcher', how='inner')
+
 pitch_stats, rhb_pitch_stats, lhb_pitch_stats = [
     (
         _df
@@ -164,6 +168,7 @@ pitch_stats, rhb_pitch_stats, lhb_pitch_stats = [
             pl.len().alias('Count')
         )
         .sort(['name', 'Count'], descending=[False, True])
+        .rename({'name': 'Player', 'pitch_name': 'Pitch'})
     )
     for _df
     in (

demo.py

@@ -39,86 +39,106 @@ with gr.Blocks(
   app_pitch_stats = gr.State(pitch_stats)
   app_league_pitch_stats = gr.State(league_pitch_stats)
 
-  with gr.Row():
-    player = gr.Dropdown(value=None, choices=sorted(player_df.filter(pl.col('name').is_not_null())['name'].to_list()), label='Player')
-    handedness = gr.Radio(value='Both', choices=['Both', 'Left', 'Right'], type='value', interactive=False, label='Batter Handedness')
+  with gr.Tab('Pitcher Dashboard'):
+    with gr.Row():
+      player = gr.Dropdown(value=None, choices=sorted(player_df.filter(pl.col('name').is_not_null())['name'].to_list()), label='Player')
+      handedness = gr.Radio(value='Both', choices=['Both', 'Left', 'Right'], type='value', interactive=False, label='Batter Handedness')
 
-  # preview = gr.DataFrame()
-  download_file = gr.DownloadButton(label='Download player data')
+    # preview = gr.DataFrame()
+    download_file = gr.DownloadButton(label='Download player data')
 
-  with gr.Group():
-    with gr.Row():
-      usage = gr.Plot(label='Pitch usage')
-      velo_summary = gr.Plot(label='Velocity summary', elem_classes='pitch-velo-summary')
-      loc_summary = gr.Plot(label='Overall location')
+    with gr.Group():
+      with gr.Row():
+        usage = gr.Plot(label='Pitch usage')
+        velo_summary = gr.Plot(label='Velocity summary', elem_classes='pitch-velo-summary')
+        loc_summary = gr.Plot(label='Overall location')
 
-  max_locs = len(jp_pitch_to_en_pitch)
-  locs_per_row = 4
-  max_rows = ceil(max_locs/locs_per_row)
+    max_locs = len(jp_pitch_to_en_pitch)
+    locs_per_row = 4
+    max_rows = ceil(max_locs/locs_per_row)
 
-  gr.Markdown('''
-  ## Pitch Locations
-  Pitcher's persective
-  <br>
-  `NPB` refers to the top 10% of pitches thrown across the league with the current search constraints e.g. handedness
-  <br>
-  Note: To speed up the KDE, we restrict the league-wide pitches to 5,000 pitches
-  ''')
-  pitch_rows = []
-  pitch_groups = []
-  pitch_names = []
-  pitch_infos = []
-  pitch_velos = []
-  pitch_locs = []
-  for row in range(max_rows):
-    visible = row==0
-    pitch_row = gr.Row(visible=visible)
-    pitch_rows.append(pitch_row)
-    with pitch_row:
-      _locs_per_row = locs_per_row if row < max_rows-1 else max_locs - locs_per_row * (max_rows - 1)
-      for col in range(_locs_per_row):
-        with gr.Column(min_width=256):
-          pitch_group = gr.Group(visible=visible)
-          pitch_groups.append(pitch_group)
-          with pitch_group:
-            pitch_names.append(gr.Markdown(f'### Pitch {col+1}', visible=visible))
-            pitch_infos.append(gr.DataFrame(pl.DataFrame([{'Whiff%': None, 'CSW%': None}]), interactive=False, visible=visible))
-            pitch_velos.append(gr.Plot(show_label=False, elem_classes='pitch-velo', visible=visible))
-            pitch_locs.append(gr.Plot(label='Pitch Location', elem_classes='pitch-loc', visible=visible))
-
-  gr.Markdown('## Pitch Velocity')
-  velo_stats = gr.DataFrame(pl.DataFrame([{'Avg. Velo': None, 'League Avg. Velo': None}]), interactive=False, label='Pitch Velocity')
-
-  (
-      player
-      .input(update_dfs, inputs=[player, handedness, source_df], outputs=[app_df, app_league_df, app_pitch_stats, app_league_pitch_stats])
-      .then(lambda : gr.update(value='Both', interactive=True), outputs=handedness)
-  )
-  handedness.input(update_dfs, inputs=[player, handedness, source_df], outputs=[app_df, app_league_df, app_pitch_stats, app_league_pitch_stats])
-
-  # app_df.change(preview_df, inputs=app_df, outputs=preview)
-  # app_df.change(set_download_file, inputs=app_df, outputs=download_file)
-  # app_df.change(plot_usage, inputs=[app_df, player], outputs=usage)
-  # app_df.change(plot_velo_summary, inputs=[app_df, app_league_df, player], outputs=velo_summary)
-  # app_df.change(lambda df: plot_loc(df), inputs=app_df, outputs=loc_summary)
-  # app_df.change(plot_pitch_cards, inputs=[app_df, app_pitch_stats], outputs=pitch_rows+pitch_groups+pitch_names+pitch_infos+pitch_velos+pitch_locs)
-  app_pitch_stats.change(update_velo_stats, inputs=[app_pitch_stats, app_league_pitch_stats], outputs=velo_stats)
-
-  (
-      app_df
-      .change(set_download_file, inputs=app_df, outputs=download_file)
-      .then(plot_usage, inputs=[app_df, player], outputs=usage)
-      .then(plot_velo_summary, inputs=[app_df, app_league_df, player], outputs=velo_summary)
-      .then(lambda df: plot_loc(df), inputs=app_df, outputs=loc_summary)
-      .then(plot_pitch_cards, inputs=[app_df, app_league_df, app_pitch_stats], outputs=pitch_rows+pitch_groups+pitch_names+pitch_infos+pitch_velos+pitch_locs)
-  )
-
-  gr.Markdown('## Bugs and other notes')
-  with gr.Accordion('Click to open', open=False):
     gr.Markdown('''
-    - Y axis ticks messy when no velocity distribution is plotted
-    - DataFrame precision inconsistent
-    '''
+    ## Pitch Locations
+    Pitcher's persective
+    <br>
+    `NPB` refers to the top 10% of pitches thrown across the league with the current search constraints e.g. handedness
+    <br>
+    Note: To speed up the KDE, we restrict the league-wide pitches to 5,000 pitches
+    ''')
+    pitch_rows = []
+    pitch_groups = []
+    pitch_names = []
+    pitch_infos = []
+    pitch_velos = []
+    pitch_locs = []
+    for row in range(max_rows):
+      visible = row==0
+      pitch_row = gr.Row(visible=visible)
+      pitch_rows.append(pitch_row)
+      with pitch_row:
+        _locs_per_row = locs_per_row if row < max_rows-1 else max_locs - locs_per_row * (max_rows - 1)
+        for col in range(_locs_per_row):
+          with gr.Column(min_width=256):
+            pitch_group = gr.Group(visible=visible)
+            pitch_groups.append(pitch_group)
+            with pitch_group:
+              pitch_names.append(gr.Markdown(f'### Pitch {col+1}', visible=visible))
+              pitch_infos.append(gr.DataFrame(pl.DataFrame([{'Whiff%': None, 'CSW%': None}]), interactive=False, visible=visible))
+              pitch_velos.append(gr.Plot(show_label=False, elem_classes='pitch-velo', visible=visible))
+              pitch_locs.append(gr.Plot(label='Pitch Location', elem_classes='pitch-loc', visible=visible))
+
+    gr.Markdown('## Pitch Velocity')
+    velo_stats = gr.DataFrame(pl.DataFrame([{'Avg. Velo': None, 'League Avg. Velo': None}]), interactive=False, label='Pitch Velocity')
+
+    (
+        player
+        .input(update_dfs, inputs=[player, handedness, source_df], outputs=[app_df, app_league_df, app_pitch_stats, app_league_pitch_stats])
+        .then(lambda : gr.update(value='Both', interactive=True), outputs=handedness)
+    )
+    handedness.input(update_dfs, inputs=[player, handedness, source_df], outputs=[app_df, app_league_df, app_pitch_stats, app_league_pitch_stats])
+
+    # app_df.change(preview_df, inputs=app_df, outputs=preview)
+    # app_df.change(set_download_file, inputs=app_df, outputs=download_file)
+    # app_df.change(plot_usage, inputs=[app_df, player], outputs=usage)
+    # app_df.change(plot_velo_summary, inputs=[app_df, app_league_df, player], outputs=velo_summary)
+    # app_df.change(lambda df: plot_loc(df), inputs=app_df, outputs=loc_summary)
+    # app_df.change(plot_pitch_cards, inputs=[app_df, app_pitch_stats], outputs=pitch_rows+pitch_groups+pitch_names+pitch_infos+pitch_velos+pitch_locs)
+    app_pitch_stats.change(update_velo_stats, inputs=[app_pitch_stats, app_league_pitch_stats], outputs=velo_stats)
+
+    (
+        app_df
+        .change(create_set_download_file_fn('files/player.csv'), inputs=app_df, outputs=download_file)
+        .then(plot_usage, inputs=[app_df, player], outputs=usage)
+        .then(plot_velo_summary, inputs=[app_df, app_league_df, player], outputs=velo_summary)
+        .then(lambda df: plot_loc(df), inputs=app_df, outputs=loc_summary)
+        .then(plot_pitch_cards, inputs=[app_df, app_league_df, app_pitch_stats], outputs=pitch_rows+pitch_groups+pitch_names+pitch_infos+pitch_velos+pitch_locs)
+    )
+
+    gr.Markdown('## Bugs and other notes')
+    with gr.Accordion('Click to open', open=False):
+      gr.Markdown('''
+      - Y axis ticks messy when no velocity distribution is plotted
+      - DataFrame precision inconsistent
+      '''
+      )
+  
+  with gr.Tab('Pitch Leaderboard'):
+    def filter_pitch_leaderboard(min_pitches):
+      return pitch_stats.filter(pl.col('Count') >= min_pitches).sort('CSW%', descending=True)
+
+    init_min_pitches = 100
+    pitch_stats.write_csv('pitch_leaderboard.csv')
+    pitch_leaderboard_df = gr.State(filter_pitch_leaderboard(init_min_pitches))
+
+    min_pitches = gr.Number(init_min_pitches, precision=0, label='Min. Pitches')
+    pitch_leaderboard_download_file = gr.DownloadButton(value='pitch_leaderboard.csv', label='Download leaderboard')
+    pitch_leaderboard = gr.Dataframe(value=pitch_leaderboard_df.value)
+    
+    min_pitches.change(filter_pitch_leaderboard, inputs=min_pitches, outputs=pitch_leaderboard_df)
+    (
+        pitch_leaderboard_df
+        .change(create_set_download_file_fn('files/pitch_leaderboard.csv'), inputs=pitch_leaderboard_df, outputs=pitch_leaderboard_download_file)
+        .then(lambda df: df, inputs=pitch_leaderboard_df, outputs=pitch_leaderboard)
     )
 
 demo.launch(

gradio_function.py

@@ -325,10 +325,11 @@ def update_dfs(player, handedness, df):
   final_filter = player_filter & handedness_filter
   return df.filter(final_filter), df.filter(handedness_filter), _pitch_stats.filter(player_filter), _league_pitch_stats,
 
-def set_download_file(df):
-  file_path = 'files/npb.csv'
-  df.write_csv(file_path)
-  return file_path
+def create_set_download_file_fn(filepath):
+  def set_download_file(df):
+    df.write_csv(filepath)
+    return filepath
+  return set_download_file
 
 def preview_df(df):
   return df.head()