log_prob_pointwise()

log_prob_pointwise()#

liesel.model.log_prob_pointwise(vars_, samples, newdata=None)[source]#

Returns a dictionary of pointwise log probabilities for the supplied variables.

Parameters:

  • vars – Dictionary of variables for which to evaluate log probs.

  • samples (dict[str, Array | ndarray | bool | number | bool | int | float | complex | TypedNdArray]) – Dictionary of samples at which to evaluate log probs. If samples contains entries for weak variables or for nodes in model_nodes they are ignored.

  • newdata (dict[str, Array | ndarray | bool | number | bool | int | float | complex | TypedNdArray] | None, default: None) – Dictionary of new data at which to evaluate log probs. The keys should correspond to variable or node names in the model whose values should be set to the given values before evaluating predictions. If None (default), the current variable values are used.

Return type:

dict[str, Array]

Returns:

  • A dictionary with pointwise log probability evaluations as values and the

  • Dist node names of the supplied variables as keys.