Gloria documentation#
Version: 0.1.1
Gloria is a flexible time series forecasting tool inspired by Prophet. It enhances Prophet’s GLM-based structure with advanced probabilistic modeling using tailored distributions (e.g., binomial, Poisson, beta) – ideal for real-world forecasting under uncertainty.
Get started#
Learn by doing. Tutorials and workflows for building, training, and evaluating models.
Explore the internals. Detailed reference for functions, models, and utility classes.
Built for Professionals#
Go beyond the normal distribution and model count data (Poisson, Binomial, Negative Binomial, Beta-Binomial), bounded rates (Beta), or non-negative floats (Gamma) natively
Gloria handles arbitrary sampling intervals (not just daily)
Parametric and extensible event library to handle holidays, campaigns, or maintenance windows - any event, any shape, for realistic impacts and reduced overfitting.
Gloria’s models are explicit, fully documented, and always inspectable.
Gloria’S modular design lets you build custom forecasting workflows by combining and extending components like events, distributions, and regressors.
Type hints, pydantic for validation, and a clean API design reminiscent of Prophet. but with a much more maintainable and extensible codebase.
Get involved#
Gloria is Open Source and thrives through your ideas, usage, and feedback. Try it, contribute, or just explore: