Inference Functions

birdman.inference.concatenate_inferences(inf_list: List[InferenceData], coords: dict, concatenation_name: str = 'feature') → InferenceData

Concatenates multiple single feature fits into one object.

Parameters

  • inf_list (List[az.InferenceData]) – List of InferenceData objects for each feature

  • coords (dict) – Coordinates containing concatenation name labels

  • concatenation_name (str) – Name of feature dimension used when concatenating, defaults to “feature”

Returns

Combined InferenceData object

Return type

az.InferenceData

birdman.inference.fit_to_inference(fit: Union[CmdStanMCMC, CmdStanVB], chains: int, draws: int, params: Sequence[str], coords: dict, dims: dict, posterior_predictive: Optional[str] = None, log_likelihood: Optional[str] = None) → InferenceData

Convert a fitted model to an arviz InferenceData object.

Parameters

  • fit (Either CmdStanMCMC or CmdStanVB) – Fitted CmdStan model

  • chains (int) – Number of chains

  • draws (int) – Number of draws

  • params (Sequence[str]) – Parameters to include in inference

  • coords (dict) – Coordinates for InferenceData

  • dims (dict) – Dimensions for InferenceData

  • posterior_predictive (str) – Name of posterior predictive var in model

  • log_likelihood (str) – Name of log likelihood var in model

Returns

Model converted to InferenceData

Return type

az.InferenceData

birdman.inference.stan_var_to_da(data: ndarray, coords: dict, dims: dict, chains: int, draws: int) → DataArray

Convert results of stan_var to DataArray.

Params data

Result of stan_var

Params coords

Coordinates of variables

Params dims

Dimensions of variables

Params chains

Number of chains

Params draws

Number of draws

Returns

DataArray representation of stan variables

Return type

xr.DataArray