ds-time-series

What problem does it solve?

Guides users through the full forecasting lifecycle — from understanding temporal patterns to deploying production-grade forecast models.

Core Features & Use Cases

• Time Series EDA: initial plots, seasonal decomposition, stationarity checks, and handling missing timestamps.
• Model selection & building: ETS, ARIMA/SARIMA, Prophet, regression with time features, and ML approaches with lag features.
• Evaluation & validation: time-aware splits, MAE, RMSE, MAPE, MASE, and predictive interval assessment.
• Production readiness: forecast pipelines, deployment considerations, and robust diagnostic checks.

Quick Start

Load your time-stamped data, perform initial EDA and stationarity checks, fit a chosen model, and generate forecasts with prediction intervals.