pastas.solver.least_squares.LeastSquaresBase.fit_report#

pastas.solver.least_squares.LeastSquaresBase.fit_report(full_output: bool = False, corr: bool = False, stderr: bool = False, warnings: bool = True, obj_func: float = np.nan) → str#

Report on the fit after a model is optimized.

Parameters:

  • full_output (bool, optional) – If True, all options are shown in the fit report. This is a shortcut for corr=True, stderr=True, and warnings=True.

  • corr (bool, optional) – If True the parameter correlations are shown.

  • stderr (bool, optional) – If True the standard error of the parameter values are shown. Please be aware of the conditions for reliable uncertainty estimates, more information here: https://pastas.readthedocs.io/stable/examples/diagnostic_checking.html

  • warnings (bool, optional) – print warnings in case of optimization failure, parameters hitting bounds, or length of responses exceeding calibration period.

  • obj_func (float, optional) – Value of the found minimal loss function value from the optimization algorithm. Generally obtained from the result attribute which is not present when loading the solver, thus by default nan.

Returns:

report – String with the report.

Return type:

str

Examples

This method is called by the solve method if report=True, but can also be called on its own:

>>> print(ml.fit_report)

Notes

The reported values for the fit use the residuals time series where possible. If interpolation is used this means that the result may slightly differ compared to using ml.simulate() and ml.observations().