pymc.model.core.Model.check_start_vals#

Model.check_start_vals(start)[source]#

Check that the starting values for MCMC do not cause the relevant log probability to evaluate to something invalid (e.g. Inf or NaN)

Parameters:

startdict, or array of dict: Starting point in parameter space (or partial point) Defaults to trace.point(-1)) if there is a trace provided and model.initial_point if not (defaults to empty dict). Initialization methods for NUTS (see init keyword) can overwrite the default.

Returns:

None

Raises:

KeyError if the parameters provided by start do not agree with the
parameters contained within the model.
pymc.exceptions.SamplingError if the evaluation of the parameters
in start leads to an invalid (i.e. non-finite) state