Spaces:
Runtime error
Runtime error
Download app.py from Akshata-89/crudeoilpredictor: direct link, hf CLI and curl.
- Browser
- Download file 6.79 kB
-
https://huggingface.co/spaces/Akshata-89/crudeoilpredictor/resolve/main/app.py
- Command line
-
hf download hf://spaces/Akshata-89/crudeoilpredictor/app.py
-
curl -L -o app.py https://huggingface.co/spaces/Akshata-89/crudeoilpredictor/resolve/main/app.py
6.79 kB
| """ | |
| app.py β FastAPI Backend for Crude Oil Price Predictor | |
| Run: uvicorn app:app --reload --port 8000 | |
| """ | |
| from fastapi import FastAPI, HTTPException | |
| from fastapi.staticfiles import StaticFiles | |
| from fastapi.responses import FileResponse | |
| from fastapi.middleware.cors import CORSMiddleware | |
| from pydantic import BaseModel | |
| import traceback | |
| from predictor import ( | |
| load_model, | |
| load_data, | |
| run_walkforward, | |
| forecast_future, | |
| predict_date, | |
| get_historical, | |
| ) | |
| app = FastAPI(title="Crude Oil Price Predictor", version="1.0.0") | |
| # CORS β allow frontend to call backend | |
| app.add_middleware( | |
| CORSMiddleware, | |
| allow_origins=["*"], | |
| allow_methods=["*"], | |
| allow_headers=["*"], | |
| ) | |
| # Serve frontend | |
| app.mount("/static", StaticFiles(directory="frontend"), name="static") | |
| # ββ Startup: preload model + data ββ | |
| async def startup(): | |
| print("π Starting up...") | |
| load_data() | |
| load_model() | |
| print("β Ready!") | |
| # ββ Routes ββ | |
| def root(): | |
| return FileResponse("frontend/index.html") | |
| def health(): | |
| return {"status": "ok", "model": "TimesFM 2.5"} | |
| def historical(): | |
| """Return last 300 days of historical prices for chart.""" | |
| try: | |
| return get_historical() | |
| except Exception as e: | |
| raise HTTPException(status_code=500, detail=str(e)) | |
| def backtest(days: int = 60): | |
| """Run walk-forward validation on last N days.""" | |
| try: | |
| if days < 10 or days > 120: | |
| raise HTTPException(400, "days must be between 10 and 120") | |
| return run_walkforward(test_days=days) | |
| except HTTPException: | |
| raise | |
| except Exception as e: | |
| traceback.print_exc() | |
| raise HTTPException(status_code=500, detail=str(e)) | |
| def forecast(horizon: int = 60): | |
| """Return future forecast for next N business days.""" | |
| try: | |
| if horizon < 5 or horizon > 90: | |
| raise HTTPException(400, "horizon must be between 5 and 90") | |
| return forecast_future(horizon=horizon) | |
| except HTTPException: | |
| raise | |
| except Exception as e: | |
| traceback.print_exc() | |
| raise HTTPException(status_code=500, detail=str(e)) | |
| class PredictRequest(BaseModel): | |
| date: str # format: YYYY-MM-DD | |
| def predict(req: PredictRequest): | |
| """Predict price for a specific future date.""" | |
| try: | |
| result = predict_date(req.date) | |
| if "error" in result: | |
| raise HTTPException(status_code=400, detail=result["error"]) | |
| return result | |
| except HTTPException: | |
| raise | |
| except Exception as e: | |
| traceback.print_exc() | |
| raise HTTPException(status_code=500, detail=str(e)) | |
| class ChatRequest(BaseModel): | |
| message: str | |
| def chat(req: ChatRequest): | |
| """ | |
| Natural language query handler. | |
| Parses date from message and returns prediction. | |
| """ | |
| import re | |
| from datetime import datetime, timedelta | |
| import pandas as pd | |
| msg = req.message.lower().strip() | |
| # Try to extract a date (YYYY-MM-DD or common phrases) | |
| date_pattern = re.search(r"\d{4}-\d{2}-\d{2}", req.message) | |
| today = datetime.today() | |
| target_date = None | |
| if date_pattern: | |
| target_date = date_pattern.group() | |
| elif "tomorrow" in msg: | |
| target_date = str((today + timedelta(days=1)).date()) | |
| elif "next week" in msg: | |
| target_date = str((today + timedelta(days=7)).date()) | |
| elif "next month" in msg or "1 month" in msg or "one month" in msg: | |
| target_date = str((today + timedelta(days=30)).date()) | |
| elif "2 week" in msg or "two week" in msg: | |
| target_date = str((today + timedelta(days=14)).date()) | |
| elif "3 month" in msg or "three month" in msg: | |
| target_date = str((today + timedelta(days=90)).date()) | |
| elif "6 month" in msg or "six month" in msg: | |
| target_date = str((today + timedelta(days=180)).date()) | |
| elif any(k in msg for k in ["today", "now", "current", "latest"]): | |
| _, dates, data = load_data() | |
| from predictor import _series | |
| return { | |
| "type" : "current", | |
| "message" : f"The latest Brent Crude Oil price is " | |
| f"${float(data['DCOILBRENTEU'].iloc[-1]):.2f}/barrel " | |
| f"as of {str(dates[-1])[:10]}.", | |
| } | |
| # Month name detection (e.g. "May 2026", "June 2026") | |
| if not target_date: | |
| months = { | |
| "january":1,"february":2,"march":3,"april":4, | |
| "may":5,"june":6,"july":7,"august":8, | |
| "september":9,"october":10,"november":11,"december":12, | |
| } | |
| for month_name, month_num in months.items(): | |
| if month_name in msg: | |
| year_match = re.search(r"\b(202\d)\b", msg) | |
| year = int(year_match.group()) if year_match else today.year | |
| try: | |
| target_date = f"{year}-{month_num:02d}-15" | |
| except: | |
| pass | |
| break | |
| if not target_date: | |
| return { | |
| "type" : "error", | |
| "message": ( | |
| "I couldn't understand that date. Try asking like:\n" | |
| "β’ 'What will oil cost on 2026-06-01?'\n" | |
| "β’ 'Predict price for next month'\n" | |
| "β’ 'What is the price for May 2026?'" | |
| ), | |
| } | |
| try: | |
| result = predict_date(target_date) | |
| if "error" in result: | |
| return {"type": "error", "message": result["error"]} | |
| direction = "π UP" if result["change_usd"] >= 0 else "π DOWN" | |
| conf_map = {"HIGH": "π’", "MEDIUM": "π‘", "LOW": "π ", "VERY LOW": "π΄"} | |
| conf_icon = conf_map.get(result["confidence"], "βͺ") | |
| message = ( | |
| f"π’οΈ **Brent Crude Oil Forecast for {target_date}**\n\n" | |
| f"π **Predicted Price:** ${result['predicted']:.2f}/barrel\n" | |
| f"π Low (10%): ${result['lower_10pct']:.2f}\n" | |
| f"π High (90%): ${result['upper_90pct']:.2f}\n" | |
| f"π Uncertainty: Β±${result['uncertainty']/2:.2f}\n\n" | |
| f"{direction} ${abs(result['change_usd']):.2f} ({result['change_pct']:+.1f}%) " | |
| f"vs last known ${result['last_known']['price']:.2f}\n\n" | |
| f"{conf_icon} Confidence: **{result['confidence']}** " | |
| f"({result['days_ahead']} business days ahead)\n\n" | |
| f"_Based on TimesFM 2.5 | Walk-Forward MAE ~$3.66_" | |
| ) | |
| return { | |
| "type" : "prediction", | |
| "message": message, | |
| "data" : result, | |
| } | |
| except Exception as e: | |
| return {"type": "error", "message": str(e)} | |