Spaces:

Syntrex
/

2026_MLB_Model

Sleeping

App Files Files

Syntrex commited on Mar 7

Commit

cfd6eaf

verified ·

1 Parent(s): 7070f92

Update features/pitch_features.py

Browse files

Files changed (1) hide show

features/pitch_features.py +30 -130

features/pitch_features.py CHANGED Viewed

@@ -1,138 +1,38 @@
 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",
-        "estimated_woba_using_speedangle": "xwoba",
-        "events": "events",
-        "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]
-    out = df[keep_cols].copy()
-    out = out.rename(columns={col: rename_map[col] for col in keep_cols})
-    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",
-        "xwoba",
-        "inning",
-        "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
 import pandas as pd
+def add_pitch_features(df: pd.DataFrame) -> pd.DataFrame:
     if df.empty:
+        return df.copy()
+    out = df.copy()
+    if {"pfx_x", "pfx_z"}.issubset(out.columns):
+        out["movement_magnitude"] = (
+            out["pfx_x"].fillna(0).astype(float) ** 2
+            + out["pfx_z"].fillna(0).astype(float) ** 2
+        ) ** 0.5
+    if "release_spin_rate" in out.columns:
+        out["spin_efficiency_proxy"] = (
+            pd.to_numeric(out["release_spin_rate"], errors="coerce").fillna(0.0) / 3000.0
+        )
+    if {"release_pos_x", "release_pos_z"}.issubset(out.columns):
+        out["release_height_proxy"] = pd.to_numeric(
+            out["release_pos_z"], errors="coerce"
+        )
+        out["release_side_proxy"] = pd.to_numeric(
+            out["release_pos_x"], errors="coerce"
+        )
+    if {"balls", "strikes"}.issubset(out.columns):
+        out["count_string"] = (
+            out["balls"].fillna(0).astype(int).astype(str)
+            + "-"
+            + out["strikes"].fillna(0).astype(int).astype(str)
+        )
     return out