facebook/prophet ? reverse-engineered prompt

Build me a time series forecasting library that feels like Prophet.

I want something I can point at a date and value table, then fit a model, make future predictions, and get a forecast plot without a lot of setup. It should handle strong seasonality like daily, weekly, and yearly patterns, support holiday effects, and work well even when there are missing values, outliers, or trend changes. Please make the main experience simple in Python, and if it makes sense, include an R interface too so both sides behave the same way.

I’d also like a few ready to run examples and notebooks that show the basic workflow, plus a clean README that explains installation and the quickest way to use it. If you need to check current docs or best practices for the forecasting backend, go ahead and look them up online. Make it installable and easy to test locally.