Package index • lwpls

Model specification

Locally-weighted PLS models for parsnip and tidymodels.

lwpls(): Locally-weighted partial least squares
update(<lwpls>): Update a locally-weighted PLS specification

Fitting and prediction

The computational engine, usable on its own.

lwpls_fit(): Fit a locally-weighted partial least squares model
predict(<lwpls_fit>): Predict from a locally-weighted PLS model
multi_predict(<_lwpls_fit>): Predictions for several numbers of components

Tuning parameters

localization(): Localization parameter for LW-PLS
similarity() values_similarity: Similarity index for LW-PLS
sparsity(): Sparsity of the LW-PLS local models

Diagnostics and plots

The local models of the predictions and their plots.

lwpls_local(): Local models of LW-PLS predictions
plot_reliability() plot_coefficients() plot_robust_weights() plot_selection() autoplot(<lwpls_local>): Plots of LW-PLS local models

Metrics

Performance metrics for tuning robust models.

trimmed_rmse() trimmed_rmse_vec(): Trimmed root mean squared error