meta/prophet ? reverse-engineered prompt

Build me a Python library for forecasting time series data that can handle strong seasonality, trend changes, missing dates, and outliers.

I want it to feel like an easy automatic forecasting tool, where someone can give it a table with dates and values, fit a model, and get future predictions with uncertainty ranges and a chart. It should support common seasonal patterns like weekly, yearly, and daily, plus holiday effects and extra regressors when needed.

Please include clear quick start examples, a couple of notebook or demo scripts, and basic documentation so it’s easy to try on a real dataset. If it makes sense, add an R interface too so the same core forecasting behavior works in both languages.

Use current best practices for the underlying model and look up any current docs online if you need to. Focus on making it reliable, easy to install, and useful out of the box.