default_convergence

caf.distribute.iterative_proportional_fitting.default_convergence(targets, achieved)

Calculate the default convergence used by ipfn.

Two lists of corresponding values are zipped together, differences taken (residuals) and the RMSE calculated.

Parameters:

• targets (Collection[ndarray]) – The targets that achieved should have reached. Must be the same length as achieved.

• achieved (Collection[ndarray]) – The achieved values. Must be the same length as targets

Returns:

A float value indicating the max convergence value achieved across residuals

Return type:

ipfn_convergence

Raises:

ValueError: – If targets and achieved are not the same length