pyzag.stochastic

Tools for converting deterministc models implemented in pytorch to stochastic models

class pyzag.stochastic.HierarchicalStatisticalModel(base, parameter_mapper, noise_prior, update_mask=False)

Converts a torch model over to being a Pyro-based hierarchical statistical model

Eventually the plan is to let the user provide a dictionary instead of a single parameter_mapper

Parameters:

• base (torch.nn.Module) – base torch module

• parameter_mapper (MapParameter) – mapper class describing how to convert from Parameter to Distribution

• noise_prior (float) – scale prior for white noise

Keyword Arguments:

update_mask (bool) – if True, update the mask to remove samples that are not valid

forward(*args, results=None, weights=None, **kwargs)

Class the base forward with the appropriate args

Parameters:

*args – whatever arguments the underlying model needs. But at least one must be a tensor so we can infer the correct batch shapes!

Keyword Arguments:

• results (torch.tensor or None) – results to condition on

• weights (torch.tensor or None) – weights on the results, default all ones

class pyzag.stochastic.MapNormal(cov, loc_suffix='_loc', scale_suffix='_scale')

A map between a deterministic torch parameter and a two-scale normal distribution

Parameters:

cov – coefficient of variation used to define the scale priors

Keyword Arguments:

• sep (str) – seperator character in names

• loc_suffix – suffix to add to parameter name to give the upper-level distribution for the scale

• scale_suffix – suffix to add to the parameter name to give the lower-level distribution for the scale