ace.validation.validate_smoothers module

A few validation problems to make sure the smoothers are working as expected.

These depend on the mace module, which was created using f2py from Friedman’s FORTRAN source (supsmu.f and friends).

ace.validation.validate_smoothers.validate_basic_smoother()

Run Friedman’s test from Figure 2b.

ace.validation.validate_smoothers.validate_basic_smoother_resid()

Compare residuals.

ace.validation.validate_smoothers.validate_supersmoother()

Validate the supersmoother.

ace.validation.validate_smoothers.validate_supersmoother_bass()

Validate the supersmoother with extra bass.

ace.validation.validate_smoothers.validate_average_best_span()

Figure 2d? from Friedman.

ace.validation.validate_smoothers.validate_known_curve()

Validate on a sin function.

ace.validation.validate_smoothers.finish_plot()

Help with plotting.

ace.validation.validate_smoothers.run_freidman_supsmu(x, y, bass_enhancement=0.0)

Run the FORTRAN supersmoother.

ace.validation.validate_smoothers.run_friedman_smooth(x, y, span)

Run the FORTRAN smoother.

ace.validation.validate_smoothers.run_mace_smothr(x, y, bass_enhancement=0.0)

Run the FORTRAN SMOTHR.

class ace.validation.validate_smoothers.BasicFixedSpanSmootherBreiman

Bases: Smoother

Runs FORTRAN Smooth.

compute()

Run smoother.

class ace.validation.validate_smoothers.SuperSmootherBreiman

Bases: Smoother

Run FORTRAN Supersmoother.

compute()

Run SuperSmoother.

ace.validation.validate_smoothers.sort_data(x, y)

Sort the data.