README.md

metadata

license: mit
title: OCIconsumptionForecaster
sdk: gradio
emoji: 📈
colorFrom: yellow
colorTo: indigo
pinned: true
sdk_version: 5.46.1

OCI Consumption Forecasting (Hugging Face Space)

A Gradio app that ingests monthly OCI consumption and produces a 12‑month forecast with a 95% confidence interval plus a downloadable CSV.

Features

• Upload CSV/XLS/XLSX
• Validates continuous monthly data with ≥ 24 months
• Forecasts next 12 months using Holt (damped) and bootstrap CIs (or SARIMA CIs)
• Outputs:
  • Chart (actuals, forecast line, 95% CI)
  • forecast_12m_with_ci.csv for download (history + forecast + bounds)

Expected Input

Your file must contain either:

• Metered Service Date and Computed Amount CD (raw export), or
• Date and Amount (pre‑grouped)

The app normalizes to month‑end and sums duplicates per month. Data must be continuous—no missing months.

How to Run (Hugging Face Spaces)

1. Create a new Space → Python.
2. Upload app.py, requirements.txt, and this README.md.
3. The Space will build and start automatically.
4. In the UI, upload your file → Run Forecast → download the CSV.

Local Run (optional)

pip install -r requirements.txt
python app.py

Notes

• Default CI method is Bootstrap, robust for short series. Switch to SARIMA (intervals only) if you prefer state‑space intervals.
• Extra columns (e.g., Currency, Subscription Plan Number) are ignored.