Why choose TreePPL?
TreePPL expresses models as computer programs that generate simulations based on input data. Specialized inference machinery then estimate posterior probability distributions for the program.
This approach lets users focus on model description while getting inference automatically.
The modeling language is designed to feel familiar to R and Python users and employs a functional programming style that works well with generic inference algorithms. Users can conveniently compile and run model programs from Python or R environments, which handle pre-processing, data input, inference control, and output processing.
Funding
