Add date filtering for pitcher dashboard

daily_weekly_leaderboard.py

@@ -10,7 +10,7 @@ import datetime
 df = (
     df
     # .join(game_df, on='game_pk')
-    .with_columns(pl.col('game_date').str.to_datetime())
+    # .with_columns(pl.col('game_date').str.to_datetime())
     .rename({
         'name': 'Name',
         'release_speed': 'Velocity',

data.py

@@ -8,6 +8,7 @@ from tqdm.auto import tqdm
 import os
 import re
 
+from seasons import SEASONS
 from translate import (
     translate_pa_outcome, translate_pitch_outcome,
     jp_pitch_to_en_pitch, jp_pitch_to_pitch_code,
@@ -31,7 +32,7 @@ def identify_bb_type(hit_type):
 
 
 DATA_DIR = 'data'
-SEASONS = sorted([folder for folder in os.listdir(DATA_DIR) if not folder.startswith('.')])
+SEASONS = [str(season) for season in SEASONS]
 
 game_df, pa_df, pitch_df, player_df, df = [], [], [], [], []
 
@@ -117,10 +118,10 @@ for season in SEASONS:
             pl.col('pitch_name').alias('jp_pitch_name')
         )
         .with_columns(
-            pl.col('jp_pitch_name').map_elements(lambda pitch_name: jp_pitch_to_en_pitch[pitch_name], return_dtype=str).alias('pitch_name'),
-            # pl.col('jp_pitch_name').replace_strict(jp_pitch_to_en_pitch).alias('pitch_name'),
-            pl.col('jp_pitch_name').map_elements(lambda pitch_name: jp_pitch_to_pitch_code[pitch_name], return_dtype=str).alias('pitch_type'),
-            # pl.col('jp_pitch_name').map_elements(jp_pitch_to_pitch_code).alias('pitch_type'),
+            # pl.col('jp_pitch_name').map_elements(lambda pitch_name: jp_pitch_to_en_pitch[pitch_name], return_dtype=str).alias('pitch_name'),
+            pl.col('jp_pitch_name').replace_strict(jp_pitch_to_en_pitch).alias('pitch_name'),
+            # pl.col('jp_pitch_name').map_elements(lambda pitch_name: jp_pitch_to_pitch_code[pitch_name], return_dtype=str).alias('pitch_type'),
+            pl.col('jp_pitch_name').replace_strict(jp_pitch_to_pitch_code).alias('pitch_type'),
             pl.col('description').str.split(' ').list.first().map_elements(translate_pitch_outcome, return_dtype=str),
             pl.when(
                 pl.col('release_speed') != '-'
@@ -167,7 +168,8 @@ for season in SEASONS:
             pl.col('description').is_in(['SS', 'K']).alias('whiff'),
             ~pl.col('description').is_in(['B', 'BB', 'LS', 'inv_K', 'bunt_K', 'HBP', 'SH', 'SH E', 'SH FC', 'obstruction', 'illegal_pitch', 'defensive_interference']).alias('swing'),
             pl.col('description').is_in(['SS', 'K', 'LS', 'inv_K']).alias('csw'),
-            ~pl.col('description').is_in(['obstruction', 'illegal_pitch', 'defensive_interference']).alias('normal_pitch') # guess
+            ~pl.col('description').is_in(['obstruction', 'illegal_pitch', 'defensive_interference']).alias('normal_pitch'), # guess
+            pl.col('game_date').str.to_datetime()
         )
     ).sort(['game_pk', 'pa_pk', 'pitch_id'])
 

gradio_function.py

@@ -26,20 +26,20 @@ INSUFFICIENT_PITCHES_MSG_MULTI_LINE = 'No visualization:<br>Not enough pitches t
 
 # GRADIO FUNCTIONS
 
-def clone_if_dataframe(item):
-  if isinstance(item, pl.DataFrame):
-    # print(type(item))
-    return item.clone()
-  else:
-    return item
-
-def clone_df(fn):
-  def _fn(*args, **kwargs):
-    args = [clone_if_dataframe(arg) for arg in args]
-    kwargs = {k: clone_if_dataframe(arg) for k, arg in kwargs.items()}
-    return fn(*args, **kwargs)
-  return _fn
-
+# def clone_if_dataframe(item):
+  # if isinstance(item, pl.DataFrame):
+    # # print(type(item))
+    # return item.clone()
+  # else:
+    # return item
+# 
+# def clone_df(fn):
+  # def _fn(*args, **kwargs):
+    # args = [clone_if_dataframe(arg) for arg in args]
+    # kwargs = {k: clone_if_dataframe(arg) for k, arg in kwargs.items()}
+    # return fn(*args, **kwargs)
+  # return _fn
+# 
 def copy_dataframe(df, num_copy_to):
   return [df.clone() for _ in range(num_copy_to)]
 
@@ -79,7 +79,7 @@ colorscale = [
 ]
 
 
-@clone_df
+# @clone_df
 def plot_loc(df, handedness, league_df=None, min_pitches=3, max_pitches=5000):
 
   loc = df.select(['plate_x', 'plate_z'])
@@ -166,7 +166,7 @@ def plot_loc(df, handedness, league_df=None, min_pitches=3, max_pitches=5000):
 
 
 # velo distribution
-@clone_df
+# @clone_df
 def plot_velo(df=None, player=None, velos=None, pitch_type=None, pitch_name=None, min_pitches=2):
   assert not ((velos is None and player is None) or (velos is not None and player is not None)), 'exactly one of `player` or `velos` must be specified'
 
@@ -212,7 +212,7 @@ def plot_velo(df=None, player=None, velos=None, pitch_type=None, pitch_name=None
   )
   return fig
 
-@clone_df
+# @clone_df
 def plot_velo_summary(df, league_df, player):
 
   min_pitches = 2
@@ -227,7 +227,9 @@ def plot_velo_summary(df, league_df, player):
 
   fig = go.Figure()
 
-  velo_center = (player_df['release_speed'].min() + player_df['release_speed'].max()) / 2
+  min_velo = player_df['release_speed'].min() if len(player_df) else 130
+  max_velo = player_df['release_speed'].max() if len(player_df) else 160
+  velo_center = (min_velo + max_velo) / 2
   # for i, (pitch_name, count) in enumerate(pitch_counts.items()):
   for i, (pitch_name, count) in enumerate(pitch_counts.iter_rows()):
     # velos = player_df.loc[pitch_name, 'release_speed']
@@ -299,7 +301,7 @@ def plot_velo_summary(df, league_df, player):
   # ))
 
   # fig.update_xaxes(title='Velocity', range=[player_df['release_speed'].dropna().min() - 2, player_df['release_speed'].dropna().max() + 2])
-  fig.update_xaxes(title='Velocity', range=[player_df['release_speed'].min() - 2, player_df['release_speed'].max() + 2])
+  fig.update_xaxes(title='Velocity', range=[min_velo - 2, max_velo + 2])
   # fig.update_yaxes(range=[0, len(pitch_counts)+1-0.25], visible=False)
   fig.update_yaxes(range=[0, len(pitch_counts)-0.25], visible=False)
   fig.update_layout(
@@ -313,7 +315,8 @@ def plot_velo_summary(df, league_df, player):
   return fig
 
 
-def update_dfs(player, handedness, df):
+def update_dfs(player, handedness, start_date, end_date, df):
+  date_filter = (pl.col('game_date') >= start_date) & (pl.col('game_date') <= end_date)
   if handedness == 'Both':
     handedness_filter = pl.col('stand').is_in(['R', 'L'])
     # _pitch_stats = pitch_stats
@@ -327,9 +330,10 @@ def update_dfs(player, handedness, df):
     # _pitch_stats = lhb_pitch_stats
     # _league_pitch_stats = lhb_league_pitch_stats
   player_filter = pl.col('name') == player
-  final_filter = player_filter & handedness_filter
+  non_player_filter = handedness_filter & date_filter
+  final_filter = player_filter & non_player_filter
   _df = df.filter(final_filter)
-  _league_df = df.filter(handedness_filter)
+  _league_df = df.filter(non_player_filter)
 
   return (
     _df,
@@ -347,13 +351,13 @@ def create_set_download_file_fn(filepath):
 def preview_df(df):
   return df.head()
 
-@clone_df
+# @clone_df
 def plot_usage(df, player):
   fig = px.pie(df.select('pitch_name'), names='pitch_name')
   fig.update_traces(texttemplate='%{percent:.1%}', hovertemplate=f'<b>{player}</b><br>' + 'threw a <b>%{label}</b><br><b>%{percent:.1%}</b> of the time (<b>%{value}</b> pitches)')
   return fig
 
-@clone_df
+# @clone_df
 def plot_pitch_cards(df, league_df, pitch_stats, handedness):
   pitch_counts = df['pitch_name'].value_counts().sort('count', descending=True)
 
@@ -400,7 +404,7 @@ def plot_pitch_cards(df, league_df, pitch_stats, handedness):
 
   return pitch_rows + pitch_groups + pitch_names + pitch_infos + pitch_velos + pitch_locs
 
-@clone_df
+# @clone_df
 def update_velo_stats(pitch_stats, league_pitch_stats):
   return (
       pitch_stats

pitch_leaderboard.py

@@ -2,11 +2,12 @@ import gradio as gr
 import polars as pl
 
 
-from data import df, game_df, compute_pitch_stats, SEASONS
+from data import df, game_df, compute_pitch_stats
+from seasons import SEASONS, LATEST_SEASON
 from gradio_function import *
 from css import css
 
-SEASONS = [int(season) for season in SEASONS]
+# SEASONS = [int(season) for season in SEASONS]
 
 def filter_pitch_leaderboard(season, min_pitches):
   return (
@@ -21,7 +22,7 @@ def create_pitch_leaderboard():
         css=css
     ) as demo:
         init_min_pitches = 100
-        init_season = max(SEASONS)
+        init_season = LATEST_SEASON
         init_pitch_stats = filter_pitch_leaderboard(init_season, init_min_pitches)
         init_pitch_stats.write_csv('pitch_leaderboard.csv')
         pitch_leaderboard_df = gr.State(init_pitch_stats)

pitcher_dashboard.py

@@ -1,13 +1,16 @@
 
 import gradio as gr
 # import pandas as pd
+from gradio_calendar import Calendar
 import polars as pl
 
 from math import ceil
+import datetime
 import os
 
 from data import df, pitch_stats, league_pitch_stats, player_df
 from gradio_function import *
+from seasons import LATEST_SEASON
 from translate import jp_pitch_to_en_pitch, max_pitch_types
 from css import css
 
@@ -30,7 +33,10 @@ def create_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')
+      start_date = Calendar(value=f'{LATEST_SEASON}-03-01', type='datetime', label='Start Date')
+      end_date = Calendar(value=f'{LATEST_SEASON}-11-30', type='datetime', label='End Date')
       handedness = gr.Radio(value='Both', choices=['Both', 'Left', 'Right'], type='value', interactive=False, label='Batter Handedness')
+    gr.Markdown('Note: We do not have spring training data, or 2024 postseason data')
 
     # preview = gr.DataFrame()
     download_file = gr.DownloadButton(label='Download player data')
@@ -91,12 +97,21 @@ def create_pitcher_dashboard():
       fn_configs[k]['df'] = gr.State(df)
       fn_configs[k]['inputs'] = [fn_configs[k]['df']] + fn_configs[k]['inputs']
 
+    update_dfs_kwargs = dict(
+        fn=update_dfs,
+        inputs=[player, handedness, start_date, end_date, source_df],
+        outputs=[app_df, app_league_df, app_pitch_stats, app_league_pitch_stats]
+    )
+    non_player_search_inputs = [handedness, start_date, end_date]
     (
         player
-        .input(update_dfs, inputs=[player, handedness, source_df], outputs=[app_df, app_league_df, app_pitch_stats, app_league_pitch_stats])
+        .input(**update_dfs_kwargs)
         .then(lambda : gr.update(value='Both', interactive=True), outputs=handedness)
+        # .then(lambda: [gr.update(interactive=True) for _ in range(len(non_player_search_inputs))], outputs=non_player_search_inputs) # breaks Calendar for some reason
     )
-    handedness.input(update_dfs, inputs=[player, handedness, source_df], outputs=[app_df, app_league_df, app_pitch_stats, app_league_pitch_stats])
+    for component in non_player_search_inputs:
+        component.input(**update_dfs_kwargs)
+    # start_date.input(**update_dfs_kwargs)
 
     # app_df.change(preview_df, inputs=app_df, outputs=preview)
     # app_df.change(set_download_file, inputs=app_df, outputs=download_file)

seasons.py

@@ -0,0 +1,5 @@
+import os
+
+DATA_DIR = 'data'
+SEASONS = sorted([int(folder) for folder in os.listdir(DATA_DIR) if not folder.startswith('.')])
+LATEST_SEASON = max(SEASONS)