API#

Dimensionality#

PyMC provides numerous methods, and syntactic sugar, to easily specify the dimensionality ofRandom Variables in modeling. Refer toDistribution Dimensionality notebook to see examplesdemonstrating the functionality.

API extensions#

Plots, stats and diagnostics#

Plots, stats and diagnostics are delegated to theArviZ.library, a general purpose library for“exploratory analysis of Bayesian models”.

  • Functions from thearviz.plots module are available throughpymc.<function> orpymc.plots.<function>,but for their API documentation please refer to theArviZ documentation.

  • Functions from thearviz.stats module are available throughpymc.<function> orpymc.stats.<function>,but for their API documentation please refer to theArviZ documentation.

ArviZ is a dependency of PyMC and so, in addition to the locations described above,importing ArviZ and usingarviz.<function> will also work without any extra installation.

Generalized Linear Models (GLMs)#

Generalized Linear Models are delegated to theBambi.library, a high-level Bayesian model-buildinginterface built on top of PyMC.

Bambi is not a dependency of PyMC and should be installed in addition to PyMCto use it to generate PyMC models via formula syntax.