Publication dateJul 10, 2025
Confidentiality public Public
Typing name : TASK.gws_design_of_experiments.PLSRegression Brick : gws_design_of_experiments PLS Regression
Partial Least Squares (PLS) Regression Task with automatic component selection.
This task performs PLS regression with cross-validation to determine the optimal
number of components. It supports both single and multi-output regression and
provides VIP (Variable Importance in Projection) scores for feature interpretation.
Inputs:
- data: Input table containing features and target variable(s)
Outputs:
- summary_table: Performance metrics (R², RMSE) per target for train and test sets
- vip_table: VIP scores ranked by importance
- plot_components: Cross-validation performance vs number of components
- vip_plot: Bar plot of VIP scores
- plot_train_set: Predicted vs true values for training set (faceted by target)
- plot_test_set: Predicted vs true values for test set (faceted by target)
Configuration:
- target: List of target column name(s) to predict
- columns_to_exclude: List of column names to exclude from analysis (optional)
- scale_data: Whether to scale the data before fitting (default: True)
- test_size: Proportion of data to use for testing (0.0 to 1.0)
login
Input
Input Data
2d excel like table
logout
Output
Summary Table
2d excel like table
Variable Importance Table
2d excel like table
Components Plot
Plotly resource
Variable Importance Plot
Plotly resource
Train Predictions Plot
Plotly resource
Test Predictions Plot
Plotly resource
settings
Configuration
columns_to_exclude
Optional
List of column names to exclude from RandomForest analysis
Type : listWhether to scale the data before fitting the PLS model.
Type : boolDefault value : trueProportion of the dataset to include in the test split (between 0.0 and 1.0).
Type : floatDefault value : 0.2Technical bricks to reuse or customize
Have you developed a brick?
Share it to accelerate projects for the entire community.
