Analysis

Specification of analyses that can be used to evaluate model fit.

Statistic definitions are used to define a set of statistical analyses that may be used to fit a model. Definitions of individual analyses are stored in Analysis objects, which specify how an analysis should be run and the expected output variables.

Definitions for a set of analyses are stored in a Statistics object. Statistics objects may be used to evaluate statistics on a given dataset, and can be used with fit_indiv() to fit parameters to specific statistics, rather than overall likelihood.

Statistic definitions may be saved to a JSON file to document exactly what statistics were fitted, in a form that can be loaded and reused.

Analysis definitions

Analysis(function, independent, dependent, level)	Manage analysis specification.
Analysis.eval(data)	Evaluate an analysis on a dataset.

Statistic definitions

Statistics([error_stat, weighting])	Manage statistics specifications.
Statistics.set_stat(stat, *args, **kwargs)	Configure an analysis to generate a statistic.
Statistics.eval_stats(data)	Evaluate all statistics.
Statistics.compare_stats(stats1, stats2)	Compare statistics.

Statistic definition files

Statistics.to_json(json_file)	Write statistics definitions to a JSON file.
read_json(json_file)	Read statistics definition from a JSON file.