Tutorials

Executable, end-to-end tutorials covering every public feature of normix. Each page runs against the current source at build time, so the code and outputs you see are always in sync with the library.

The tutorials are grouped into five tracks. Core builds the conceptual foundation; Distributions tours every class; EM covers fitting in practice; Statistics adds model comparison and goodness of fit; and Finance applies everything to real market data.

Core concepts

• The exponential family
• A tour of the GH family
• Bessel functions and log_kv
• Random sampling

Distribution tour

• Univariate positive distributions
• The Generalized Inverse Gaussian
• The multivariate normal
• Normal variance-mean mixtures
• Factor mixtures for high dimensions

EM in practice

• Batch EM in practice
• Incremental (mini-batch) EM
• Initialization and multi-start
• EM vs MCECM Algorithm Comparison

Statistical analysis

• Divergences between models
• Goodness of fit

Finance

• A heavy-tailed index series
• A multivariate stock basket
• Factor mixtures for a Dow Jones 30 portfolio
• Portfolio CVaR and its derivatives