Create app.py

app.py

@@ -0,0 +1,126 @@
+from fastapi import FastAPI
+import requests
+from bs4 import BeautifulSoup
+from datetime import date
+from transformers import pipeline
+from sentence_transformers import SentenceTransformer, util
+from balldontlie import BalldontlieAPI
+import os
+import uvicorn
+from dotenv import load_dotenv
+
+load_dotenv()
+
+app = FastAPI(title="Sports Predictor", description="AI-powered sports predictions for soccer and NBA")
+
+sentiment_model = pipeline("zero-shot-classification", model="valhalla/distilbart-mnli-12-1")
+reasoning_model = pipeline("text2text-generation", model="google/flan-t5-base")
+similarity_model = SentenceTransformer("sentence-transformers/all-MiniLM-L6-v2")
+
+FOOTBALL_API_KEY = os.getenv("FOOTBALL_API_KEY", "YOUR_FOOTBALL_DATA_API_KEY")
+FOOTBALL_ENDPOINT = "https://api.football-data.org/v4/matches"
+NBA_API_KEY = os.getenv("NBA_API_KEY", "your-api-key")
+nba_api = BalldontlieAPI(api_key=NBA_API_KEY)
+
+SOCCER_LEAGUES = {
+    "EPL": 2021,
+    "LaLiga": 2014,
+    "Bundesliga": 2002
+}
+
+def get_team_news(team: str, sport: str = "football"):
+    url = f"https://news.google.com/search?q={team}+{sport}&hl=en"
+    r = requests.get(url, headers={"User-Agent": "Mozilla/5.0"})
+    soup = BeautifulSoup(r.text, "html.parser")
+    headlines = [h.text for h in soup.select("h3")[:5]]
+    return " ".join(headlines)
+
+@app.get("/soccer-predictions")
+def soccer_predictions():
+    today = date.today().isoformat()
+    predictions = []
+
+    for league, code in SOCCER_LEAGUES.items():
+        headers = {"X-Auth-Token": FOOTBALL_API_KEY}
+        resp = requests.get(
+            f"{FOOTBALL_ENDPOINT}?competitions={code}&dateFrom={today}&dateTo={today}",
+            headers=headers
+        )
+        matches = resp.json().get("matches", [])
+
+        for match in matches:
+            home, away = match["homeTeam"]["name"], match["awayTeam"]["name"]
+
+            home_news = get_team_news(home, "football")
+            away_news = get_team_news(away, "football")
+
+            labels = ["positive", "negative", "injury", "transfer", "motivation"]
+            home_sent = sentiment_model(home_news, labels)
+            away_sent = sentiment_model(away_news, labels)
+
+            emb_home = similarity_model.encode(home + " " + home_news, convert_to_tensor=True)
+            emb_away = similarity_model.encode(away + " " + away_news, convert_to_tensor=True)
+            similarity = util.pytorch_cos_sim(emb_home, emb_away).item()
+
+            context = f"""
+            Match: {home} vs {away}.
+            Home news: {home_news}.
+            Away news: {away_news}.
+            Semantic similarity: {similarity}.
+            """
+            reasoning = reasoning_model(
+                f"Predict 1X2 and Over/Under 2.5 goals for this soccer match. Context: {context}",
+                max_length=256
+            )[0]['generated_text']
+
+            predictions.append({
+                "league": league,
+                "match": f"{home} vs {away}",
+                "reason": reasoning,
+                "news_summary": {
+                    home: home_sent,
+                    away: away_sent
+                }
+            })
+
+    return {"date": today, "predictions": predictions}
+
+@app.get("/nba-predictions")
+def nba_predictions():
+    today = date.today().isoformat()
+    games = nba_api.nba.games.list(dates=[today])
+    predictions = []
+
+    for g in games:
+        home, away = g["home_team"]["full_name"], g["visitor_team"]["full_name"]
+
+        home_news = get_team_news(home, "NBA")
+        away_news = get_team_news(away, "NBA")
+
+        labels = ["positive", "negative", "injury", "motivation"]
+        home_sent = sentiment_model(home_news, labels)
+        away_sent = sentiment_model(away_news, labels)
+
+        context = f"""
+        Match: {home} vs {away}.
+        Home news: {home_news}.
+        Away news: {away_news}.
+        """
+        reasoning = reasoning_model(
+            f"Predict winner and Over/Under points for this NBA game. Context: {context}",
+            max_length=256
+        )[0]['generated_text']
+
+        predictions.append({
+            "match": f"{home} vs {away}",
+            "reason": reasoning,
+            "news_summary": {
+                home: home_sent,
+                away: away_sent
+            }
+        })
+
+    return {"date": today, "predictions": predictions}
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=7860)