Skip to contents

PLS

Source: R/multivariate.R
pls.Rd

Simple wrappers for fitting a PLS model using mixOmics package. The result can then be passed to many of the mixOmics functions for prediction, performance evaluation etc.

Usage

mixomics_pls(
  object,
  y,
  ncomp,
  plot_scores = TRUE,
  all_features = FALSE,
  covariates = NULL,
  assay.type = NULL,
  ...
)

mixomics_pls_optimize(
  object,
  y,
  ncomp,
  folds = 5,
  nrepeat = 50,
  plot_scores = TRUE,
  all_features = FALSE,
  covariates = NULL,
  assay.type = NULL,
  ...
)

mixomics_spls_optimize(
  object,
  y,
  ncomp,
  n_features = c(seq_len(10), seq(20, 300, 10)),
  folds = 5,
  nrepeat = 50,
  plot_scores = TRUE,
  all_features = FALSE,
  covariates = NULL,
  assay.type = NULL,
  ...
)

Arguments

object

a SummarizedExperiment or MetaboSet object

y

character vector, column names of the grouping variable to predict

ncomp

number of X components

plot_scores

logical, if TRUE, a scatter plot with the first two PLS- components as x and y-axis will be drawn, colored by the y-variable. Only really makes sense if y is a single variable

all_features

logical, should all features be included in the model? if FALSE, flagged features are left out

covariates

character, column names of pheno datato use as covariates in the model, in addition to molecular features

assay.type

character, assay to be used in case of multiple assays

...

any parameters passed to pls or spls

folds

the number of folds to use in k-fold cross validation

nrepeat

the number of times to repeat the cross validation. Lower this for faster testing.

n_features

the number of features to try for each component

Value

An object of class "mixo_pls" or "mixo_spls".

Details

  • mixomics_pls A simple PLS model with set number of components and all features

  • mixomics_pls_optimize Test different numbers of components

  • mixomics_spls_optimize sPLS model: Test different numbers of components and features

See also

pls, perf, spls, tune.spls

Examples

data(example_set)
pls_res <- mixomics_pls(example_set, y = "Injection_order", ncomp = 3)
#> INFO [2025-06-23 22:37:54] Fitting PLS
# Cross-validation repeated only 5 times for quick run time 
pls_opt <- mixomics_pls_optimize(example_set, 
  y = "Injection_order", ncomp = 3, nrepeat = 5)
#> INFO [2025-06-23 22:37:55] Fitting PLS
#> INFO [2025-06-23 22:37:55] Evaluating PLS performance
#> INFO [2025-06-23 22:37:55] Choosing a PLS model with 1 component(s) based on the minimal MSE
#> Take a look at the plot and make sure this is the correct number of components
#> INFO [2025-06-23 22:37:55] Fitting PLS
spls_opt <- mixomics_spls_optimize(example_set,
  y = "Injection_order", ncomp = 3,
  n_features <- c(1:10, 12, 15, 20), nrepeat = 5
)
#> INFO [2025-06-23 22:37:55] Tuning sPLS
#> INFO [2025-06-23 22:37:59] Final model has 1 components with the numbers of features: 20