routes/ML_RandomForestClassifier_CreditApproval.py

from fastapi import APIRouter
from pydantic import BaseModel
import joblib
import pandas as pd
from typing import Optional, Any

from .config_huggingface import build_model_url, download_artifact_if_needed

router = APIRouter(tags=["Machine Learning"])


class RandomForestClassifierRequest(BaseModel):
    income_k: float = 65.0
    debt_k: float = 15.0
    employment_years: float = 5.0
    credit_score: float = 710.0


MODEL_STATE: dict[str, Optional[Any]] = {
    "model": None,
    "error": None,
}

MODEL_URL = build_model_url("ML_RandomForestClassifier_CreditApproval.joblib")


def _ensure_model_loaded() -> None:
    if MODEL_STATE["model"] is not None:
        return
    try:
        model_path = download_artifact_if_needed(MODEL_URL)
        MODEL_STATE["model"] = joblib.load(model_path)
        MODEL_STATE["error"] = None
    except Exception as e:
        MODEL_STATE["error"] = str(e)
        raise


@router.post("/models/random_forest_classifier", summary="Predict loan approval with Random Forest")
def predict_random_forest_classifier(data: RandomForestClassifierRequest):
    import traceback
    try:
        _ensure_model_loaded()
    except Exception:
        detail = "Model not loaded."
        if MODEL_STATE["error"]:
            detail = f"Model not loaded: {MODEL_STATE['error']}"
        return {"error": detail, "traceback": traceback.format_exc(), "status": 500}

    model = MODEL_STATE["model"]
    if model is None:
        return {"error": f"Model is None after loading. Error: {MODEL_STATE['error']}", "status": 500}

    input_df = pd.DataFrame(
        [[data.income_k, data.debt_k, data.employment_years, data.credit_score]],
        columns=["income_k", "debt_k", "employment_years", "credit_score"],
    )

    try:
        pred = int(model.predict(input_df)[0])
        proba = model.predict_proba(input_df)[0]
    except Exception as e:
        return {"error": f"Prediction failed: {str(e)}", "traceback": traceback.format_exc(), "status": 500}

    return {
        "prediction": "Approved" if pred == 1 else "Denied",
        "confidence": f"{max(proba):.0%}",
    }