bayesplain.priors.resolve_table¶
- bayesplain.priors.resolve_table(prior)[source]¶
Turn whatever the user passed as
prior=into aConcentrationPrior.Accepts a preset name, a positive number, an existing
ConcentrationPrior, orNonefor the default. ABetaPrioris also accepted when its two shapes are equal, so that the sameprior=argument can be handed to a proportion analysis and a table analysis and mean the same thing.- Parameters:¶
- prior : str, float, ConcentrationPrior, BetaPrior, or None¶
The prior specification.
Noneresolves to"uninformed".
- Returns:¶
The resolved prior.
- Return type:¶
- Raises:¶
ValueError – If a name is not a known preset, or a
BetaPriorwith unequal shapes is given, since that has no symmetric table equivalent.