Update pitch_summary_functions.py

pitch_summary_functions.py

@@ -22,7 +22,7 @@ colour_palette = ['#FFB000','#648FFF','#785EF0',
                   '#DC267F','#FE6100','#3D1EB2','#894D80','#16AA02','#B5592B','#A3C1ED']
 season_start = '2024-03-20'
 season_end = '2024-09-29'
-season_fg=2024 
+season_fg=2023
 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()
 cmap_sum = matplotlib.colors.LinearSegmentedColormap.from_list("", ['#648FFF','#FFFFFF','#FFB000',])
 
@@ -228,7 +228,6 @@ def percentile(n):
     return percentile_
 
 ### TJ STUFF+ DF CLEAN ###
-
 def df_clean(df):
     df_copy = df.copy()
     df_copy.loc[df_copy['pitcher_hand'] == 'L','hb'] *= -1
@@ -237,7 +236,7 @@ def df_clean(df):
 
     df_copy['pitch_l'] = [1 if x == 'L' else 0 for x in df_copy['pitcher_hand']]
     df_copy['bat_l'] = [1 if x == 'L' else 0 for x in df_copy['batter_hand']]
-    #df_copy = df_copy[~df_copy.pitch_type.isin(["EP", "PO", "KN", "CS", "SC", "FA"])].reset_index(drop=True)
+    df_copy = df_copy[~df_copy.pitch_type.isin(["PO"])].reset_index(drop=True)
     df_copy['pitch_type'] = df_copy['pitch_type'].replace({'FT':'SI'})
 
     # df_copy['des_new'] = df_copy['play_description'].map(des_dict)
@@ -315,7 +314,6 @@ def df_clean(df):
     df_copy = df_copy.dropna(subset=['pitch_type']).fillna(0)
     return df_copy
 
-
 ### PITCH COLOURS ###
 pitch_colours = {
         'Four-Seam Fastball':'#FF007D',#BC136F
@@ -337,7 +335,7 @@ pitch_colours = {
         'Slow Curve':'#274BFC',
         'Eephus':'#648FFF',
 
-        'Knuckle Ball':'#867A08',
+        'Knuckleball':'#867A08',
 
         'Pitch Out':'#472C30',
         'Other':'#9C8975',
@@ -659,10 +657,19 @@ def tj_stuff_roling(df,
         ax.xaxis.set_major_locator(MaxNLocator(integer=True))
 
 ### TJ STUFF+ ROLLING ###
-def tj_velocity_roling(df,
+def tj_stuff_roling_game(df,
                     window,
                     ax):
         ## Velocity Plot
+
+        # Create a dictionary to map each unique date to a unique number
+        date_to_number = {date: i+1 for i, date in enumerate(df['game_id'].unique())}
+
+        
+        df['start_number'] = df['game_id'].map(date_to_number)
+        
+        plot_game_roll = df.groupby(['start_number','game_id','game_date','pitch_type','pitch_description'])['tj_stuff_plus'].mean().reset_index()               
+        
         sorted_value_counts = df['pitch_type'].value_counts().sort_values(ascending=False)
 
         # Get the list of items ordered from most to least frequent
@@ -670,19 +677,39 @@ def tj_velocity_roling(df,
         
 
         for i in items_in_order:
-            if max(df[df['pitch_type']==i]['pitch_type_count_each']) >= window:
-                sns.lineplot(x=range(1,max(df[df['pitch_type']==i]['pitch_type_count_each'])+1),
-                        y=df[df['pitch_type']==i]['start_speed'].rolling(window).sum()/window,
-                    color=pitch_colours[df[df['pitch_type']==i]['pitch_description'].values[0]],
-                    ax=ax,linewidth=3)
+            df_item = plot_game_roll[plot_game_roll['pitch_type']==i].set_index(['start_number'])
+            df_item = df_item.reindex(range(df['start_number'].min(),df['start_number'].max()+1)).reset_index()
+            #df_item['start_number'] = df_item['start_number']
+            #df_item['tj_stuff_plus'] = df_item['tj_stuff_plus'].
+            
+            df_item['tj_stuff_plus'] = df_item['tj_stuff_plus'].ffill()
+            df_item['pitch_type'] = df_item['pitch_type'].ffill()
+            df_item['pitch_description'] = df_item['pitch_description'].ffill()
+            print(df_item)
+            sns.lineplot(x=range(1,max(df['start_number']+1)),
+                        y=df_item[df_item['pitch_type']==i]['tj_stuff_plus'].rolling(window).sum()/window,
+                        color=pitch_colours[df[df['pitch_type']==i]['pitch_description'].values[0]],
+                        ax=ax,linewidth=3)
+        
+            # for d in df_item[df_item['game_id'].isnull()]['start_number']:
+            #     df_item['tj_stuff_plus'] = df_item['tj_stuff_plus'].bfill()
+            #     df_item['pitch_type'] = df_item['pitch_type'].bfill()
+            #     df_item['pitch_description'] = df_item['pitch_description'].bfill()                
+            #     print(d)
+            #     sns.scatterplot(
+            #                     x=[d+1],
+            #                     y=(df_item[df_item['pitch_type']==i]['tj_stuff_plus'].rolling(window).sum()/window)[d],
+            #                     edgecolor=pitch_colours[df[df['pitch_type']==i]['pitch_description'].values[0]],color='white',
+            #                     s=50,zorder=30,
+            #                     ax=ax)
 
         # Adjust x-axis limits to start from 1
-        ax.set_xlim(window,max(df['pitch_type_count_each']))
-        ax.set_ylim(65,105)
+        ax.set_xlim(window,max(df['start_number']))
+        ax.set_ylim(70,130)
         #ax.get_legend().remove()
-        ax.set_xlabel('Pitches', fontdict=font_properties_axes)
-        ax.set_ylabel('Velocity (mph)', fontdict=font_properties_axes)
-        ax.set_title(f"{window} Pitch Rolling Velocity",fontdict=font_properties_titles)
+        ax.set_xlabel('Games', fontdict=font_properties_axes)
+        ax.set_ylabel('tjStuff+', fontdict=font_properties_axes)
+        ax.set_title(f"{window} Game Rolling tjStuff+",fontdict=font_properties_titles)
         # ax.axis('square')
         # ax.set_xlim(left=1)
         ax.xaxis.set_major_locator(MaxNLocator(integer=True))
@@ -1091,30 +1118,4 @@ def location_plot(df,ax,hand):
         if len(pitch_location_group['px'])>0:
             ax.get_legend().remove()
         ax.grid(False)
-        ax.set_title(f"Pitch Locations vs {hand}HH\n{pitch_location_group['pitches'].sum()} Pitches",fontdict=font_properties_titles)
-
-
-
-
-####LHH
-def location_pitch_plot(df,ax,hand):
-        #label_labels = df.sort_values(by=['prop','pitch_type'],ascending=[False,True]).pitch_description.unique()
-        ## Location Plot
-        df_pitch= df[df['batter_hand']==hand]
-        sns.scatterplot(ax=ax,x=df_pitch['px'],
-                                    y=df_pitch['pz'],
-                                    hue=df_pitch['pitch_description'],
-                                    palette=pitch_colours,ec='black',
-                                    s=175,
-                                    linewidth=2,
-                                    zorder=2)
-
-        ax.axis('square')
-        draw_line(ax,alpha_spot=0.75,catcher_p=False)
-        ax.axis('off')
-        ax.set_xlim((-2.75,2.75))
-        ax.set_ylim((-0.5,5))
-        if len(df_pitch['px'])>0:
-            ax.get_legend().remove()
-        ax.grid(False)
-        ax.set_title(f"Pitch Locations vs {hand}HH\n{df_pitch['pitches'].sum()} Pitches",fontdict=font_properties_titles)
+        ax.set_title(f"Pitch Locations vs {hand}HB\n{pitch_location_group['pitches'].sum()} Pitches",fontdict=font_properties_titles)