Simplified app with compatible versions

Dockerfile

@@ -6,13 +6,13 @@ WORKDIR /app
 COPY requirements.txt .
 RUN pip install --no-cache-dir -r requirements.txt
 
-# Copy model files and app
-COPY calibrated_*.joblib .
-COPY calibrated_model_metadata.json .
-COPY app.py .
+# Copy all model files and app
+COPY *.joblib .
+COPY *.json .
+COPY simple_app.py .
 
 # Expose port
 EXPOSE 7860
 
-# Run the application
-CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]
+# Run the simple application
+CMD ["uvicorn", "simple_app:app", "--host", "0.0.0.0", "--port", "7860"]

requirements.txt

@@ -1,7 +1,7 @@
-fastapi==0.104.1
-uvicorn[standard]==0.24.0
-joblib==1.3.2
-numpy==2.2.3
-pandas==2.0.3
-scikit-learn==1.7.2
-pydantic==2.4.2
+fastapi
+uvicorn[standard]
+joblib
+numpy
+pandas
+scikit-learn
+pydantic

simple_app.py

@@ -0,0 +1,85 @@
+from fastapi import FastAPI
+from pydantic import BaseModel
+import joblib
+import numpy as np
+import pandas as pd
+
+app = FastAPI(title="EPL Predictions")
+
+# Load model
+try:
+    model = joblib.load('simple_rf_model.joblib')
+    scaler = joblib.load('simple_scaler.joblib')
+    print("Models loaded successfully")
+except:
+    model = None
+    scaler = None
+    print("Failed to load models")
+
+class MatchRequest(BaseModel):
+    home_team: str
+    away_team: str
+    home_xg: float = 1.5
+    away_xg: float = 1.3
+
+class PredictionResponse(BaseModel):
+    home_team: str
+    away_team: str
+    home_win: float
+    draw: float
+    away_win: float
+    prediction: str
+    confidence: float
+
+@app.get("/")
+def root():
+    return {"status": "EPL Prediction API", "model_loaded": model is not None}
+
+@app.get("/health")
+def health():
+    return {"status": "healthy", "models_loaded": model is not None}
+
+@app.post("/predict")
+def predict(match: MatchRequest):
+    if model is None or scaler is None:
+        return {"error": "Models not loaded"}
+
+    # Prepare features
+    features = pd.DataFrame([{
+        'home_xg': match.home_xg * 0.82,  # Apply calibration
+        'away_xg': match.away_xg * 0.82
+    }])
+
+    # Scale and predict
+    X_scaled = scaler.transform(features)
+    probs = model.predict_proba(X_scaled)[0]
+
+    # Map probabilities (0=away, 1=draw, 2=home)
+    away_prob = probs[0]
+    draw_prob = probs[1] if len(probs) > 2 else 0.25
+    home_prob = probs[2] if len(probs) > 2 else probs[1]
+
+    # Get prediction
+    if home_prob > draw_prob and home_prob > away_prob:
+        prediction = "Home"
+        confidence = home_prob
+    elif away_prob > draw_prob:
+        prediction = "Away"
+        confidence = away_prob
+    else:
+        prediction = "Draw"
+        confidence = draw_prob
+
+    return PredictionResponse(
+        home_team=match.home_team,
+        away_team=match.away_team,
+        home_win=float(home_prob),
+        draw=float(draw_prob),
+        away_win=float(away_prob),
+        prediction=prediction,
+        confidence=float(confidence)
+    )
+
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=7860)

simple_metadata.json

@@ -0,0 +1,8 @@
+{
+  "features": [
+    "home_xg",
+    "away_xg"
+  ],
+  "training_samples": 1560,
+  "model_type": "RandomForestClassifier"
+}

simple_rf_model.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dab4165f1667e4e0a75f818ffeebac6a16cf8f42d930a9074ff5f6c6078bd3c3
+size 26257

simple_scaler.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3a4b9c7e698c810506da4385e6e36029c69cb1725f9048275b83dfc5927e2780
+size 935