Spaces:

Syntrex
/

2026_MLB_Model

Running

App Files Files

Syntrex commited on Mar 7

Commit

d144d29

verified ·

1 Parent(s): 38cb69d

Update data/statcast.py

Browse files

Files changed (1) hide show

data/statcast.py +88 -65

data/statcast.py CHANGED Viewed

@@ -1,96 +1,107 @@
 from __future__ import annotations
 from io import StringIO
-from typing import Any
 import pandas as pd
 import requests
 from config.settings import STATCAST_SEARCH_URL
 HEADERS = {
     "User-Agent": "Mozilla/5.0",
     "Accept-Language": "en-US,en;q=0.9",
 }
-def fetch_wbc_statcast_range(start_date: str, end_date: str, season: str = "2026") -> pd.DataFrame:
     """
-    Pull WBC pitch/event-level Statcast-style CSV from Baseball Savant.
-    Baseball Savant has a dedicated WBC search surface, but CSV exports are still served
-    from the same csv backend path pattern. The key difference is using tournament filters.
     """
-    params = {
-        "all": "true",
-        "hfPT": "",
-        "hfAB": "",
-        "hfBBT": "",
-        "hfPR": "",
-        "hfZ": "",
-        "stadium": "",
-        "hfBBL": "",
-        "hfNewZones": "",
-        "hfGT": "F|D|L|W|",  # game types commonly used in savant filters
-        "hfC": "",
-        "hfSea": f"{season}|",
-        "hfSit": "",
-        "player_type": "batter",
-        "hfOuts": "",
-        "opponent": "",
-        "pitcher_throws": "",
-        "batter_stands": "",
-        "hfSA": "",
-        "game_date_gt": start_date,
-        "game_date_lt": end_date,
-        "team": "",
-        "position": "",
-        "hfRO": "",
-        "home_road": "",
-        "hfFlag": "",
-        "metric_1": "",
-        "hfInn": "",
-        "min_pitches": "0",
-        "min_results": "0",
-        "group_by": "name",
-        "sort_col": "pitches",
-        "player_event_sort": "h_launch_speed",
-        "sort_order": "desc",
-        "min_abs": "0",
-        "type": "details",
-    }
-    response = requests.get(STATCAST_SEARCH_URL, params=params, headers=HEADERS, timeout=60)
-    response.raise_for_status()
-    text = response.text.strip()
-    if not text or text.startswith("<!DOCTYPE html"):
-        return pd.DataFrame()
-    try:
-        df = pd.read_csv(StringIO(text))
-    except Exception:
-        return pd.DataFrame()
-    return df
-def normalize_wbc_statcast(df: pd.DataFrame) -> pd.DataFrame:
     if df.empty:
         return df
     rename_map = {
         "player_name": "player_name",
         "pitch_type": "pitch_type",
         "release_speed": "release_speed",
         "release_spin_rate": "release_spin_rate",
-        "pfx_x": "pfx_x",
-        "pfx_z": "pfx_z",
         "release_pos_x": "release_pos_x",
         "release_pos_z": "release_pos_z",
         "plate_x": "plate_x",
         "plate_z": "plate_z",
         "launch_speed": "launch_speed",
         "launch_angle": "launch_angle",
         "estimated_ba_using_speedangle": "xba",
@@ -99,13 +110,18 @@ def normalize_wbc_statcast(df: pd.DataFrame) -> pd.DataFrame:
         "description": "description",
         "stand": "batter_stand",
         "p_throws": "pitcher_hand",
-        "game_date": "game_date",
         "home_team": "home_team",
         "away_team": "away_team",
         "inning": "inning",
         "outs_when_up": "outs_when_up",
         "balls": "balls",
         "strikes": "strikes",
     }
     keep_cols = [col for col in rename_map if col in df.columns]
@@ -115,12 +131,12 @@ def normalize_wbc_statcast(df: pd.DataFrame) -> pd.DataFrame:
     numeric_cols = [
         "release_speed",
         "release_spin_rate",
-        "pfx_x",
-        "pfx_z",
         "release_pos_x",
         "release_pos_z",
         "plate_x",
         "plate_z",
         "launch_speed",
         "launch_angle",
         "xba",
@@ -129,10 +145,17 @@ def normalize_wbc_statcast(df: pd.DataFrame) -> pd.DataFrame:
         "outs_when_up",
         "balls",
         "strikes",
     ]
     for col in numeric_cols:
         if col in out.columns:
             out[col] = pd.to_numeric(out[col], errors="coerce")
     return out

 from __future__ import annotations
 from io import StringIO
 import pandas as pd
 import requests
 from config.settings import STATCAST_SEARCH_URL
 HEADERS = {
     "User-Agent": "Mozilla/5.0",
     "Accept-Language": "en-US,en;q=0.9",
 }
+def fetch_statcast_range(start_date: str, end_date: str) -> pd.DataFrame:
     """
+    WBC-first Statcast pull.
+    This uses Savant's CSV backend with tournament-style filters and recent dates.
+    If 2026 returns no rows, it falls back to 2023 WBC season coverage.
     """
+    season_candidates = ["2026", "2023"]
+    for season in season_candidates:
+        params = {
+            "all": "true",
+            "hfPT": "",
+            "hfAB": "",
+            "hfBBT": "",
+            "hfPR": "",
+            "hfZ": "",
+            "stadium": "",
+            "hfBBL": "",
+            "hfNewZones": "",
+            "hfGT": "F|D|L|W|",
+            "hfC": "",
+            "hfSea": f"{season}|",
+            "hfSit": "",
+            "player_type": "batter",
+            "hfOuts": "",
+            "opponent": "",
+            "pitcher_throws": "",
+            "batter_stands": "",
+            "hfSA": "",
+            "game_date_gt": start_date,
+            "game_date_lt": end_date,
+            "team": "",
+            "position": "",
+            "hfRO": "",
+            "home_road": "",
+            "hfFlag": "",
+            "metric_1": "",
+            "hfInn": "",
+            "min_pitches": "0",
+            "min_results": "0",
+            "group_by": "name",
+            "sort_col": "pitches",
+            "player_event_sort": "api_h_launch_speed",
+            "sort_order": "desc",
+            "min_abs": "0",
+            "type": "details",
+        }
+        response = requests.get(
+            STATCAST_SEARCH_URL,
+            params=params,
+            headers=HEADERS,
+            timeout=60,
+        )
+        response.raise_for_status()
+        text = response.text.strip()
+        if not text or text.startswith("<!DOCTYPE html"):
+            continue
+        try:
+            df = pd.read_csv(StringIO(text))
+        except Exception:
+            continue
+        if not df.empty:
+            return df
+    return pd.DataFrame()
+def normalize_statcast(df: pd.DataFrame) -> pd.DataFrame:
     if df.empty:
         return df
     rename_map = {
         "player_name": "player_name",
         "pitch_type": "pitch_type",
+        "pitch_name": "pitch_name",
         "release_speed": "release_speed",
         "release_spin_rate": "release_spin_rate",
         "release_pos_x": "release_pos_x",
         "release_pos_z": "release_pos_z",
         "plate_x": "plate_x",
         "plate_z": "plate_z",
+        "pfx_x": "pfx_x",
+        "pfx_z": "pfx_z",
         "launch_speed": "launch_speed",
         "launch_angle": "launch_angle",
         "estimated_ba_using_speedangle": "xba",
         "description": "description",
         "stand": "batter_stand",
         "p_throws": "pitcher_hand",
         "home_team": "home_team",
         "away_team": "away_team",
+        "game_date": "game_date",
+        "game_pk": "game_pk",
         "inning": "inning",
         "outs_when_up": "outs_when_up",
         "balls": "balls",
         "strikes": "strikes",
+        "bat_score": "bat_score",
+        "fld_score": "fld_score",
+        "post_bat_score": "post_bat_score",
+        "post_fld_score": "post_fld_score",
     }
     keep_cols = [col for col in rename_map if col in df.columns]
     numeric_cols = [
         "release_speed",
         "release_spin_rate",
         "release_pos_x",
         "release_pos_z",
         "plate_x",
         "plate_z",
+        "pfx_x",
+        "pfx_z",
         "launch_speed",
         "launch_angle",
         "xba",
         "outs_when_up",
         "balls",
         "strikes",
+        "bat_score",
+        "fld_score",
+        "post_bat_score",
+        "post_fld_score",
     ]
     for col in numeric_cols:
         if col in out.columns:
             out[col] = pd.to_numeric(out[col], errors="coerce")
+    if "game_date" in out.columns:
+        out["game_date"] = pd.to_datetime(out["game_date"], errors="coerce")
     return out