PLS Quick Specs - Options Tab

Select the Options tab of the PLS Quick Specs dialog box to access the options described here. Computational details for most options are available in the General Linear Model (GLM) Introductory Overview.

Element Name Description
Sigma-restricted Select the Sigma-restricted check box to compute the design matrix for categorical predictors in the model based on sigma-restricted coding; if this check box is not selected, the overparameterized model will be used. The sigma-restricted model is the default parameterization, except for models that involve nested effects; see The Sigma-Restricted and Overparameterized Model for details.
No intercept Select the No intercept check box to exclude the intercept from the model. The intercept in PLS models is handled in a different manner than in general linear models. If No intercept is selected, all columns of the (predictor) design matrix X and matrix of dependent or response variables Y are centered at their means, and the intercept term is excluded from the model. If this check box is not selected, these matrices are not centered and the intercept term is included in the model.
Auto-scaling Select the Auto-scaling check box to divide each column in the predictor design matrix X and matrix of dependent (response) variables Y by its respective standard deviation, and all computations will be performed on these scaled matrices X and Y. For more information about auto-scaling, refer to Geladi and Kowalski(1986).
SIMPLS Select the SIMPLS check box to use the SIMPLS algorithm (de Jong, 1993) to compute the PLS components; if this check box is not selected, the standard PLS analysis using the NIPALS algorithm (Rannar, Lindgren, Geladi, and Wold, 1994) will be performed.
Cross-validation Click the Cross-validation button to display the Cross-Validation dialog box for specifying a categorical variable and a (code) value to identify observations that should be included in the computations for fitting the model (the analysis or training sample); all other observations with valid data for all predictor variables and dependent variables will automatically be classified as belonging to the validation (verification) sample; note that all observations with valid data for all predictor variables but missing data for the dependent variables will automatically be classified as belonging to the prediction sample.
Eigen delta Enter the negative exponent for a base-10 constant delta (delta = 10-Edelta) in the Eigen delta field; delta is used for checking the convergence of the iterative computation of eigenvectors for each PLS component. See the description of the EDELTA keyword for additional details.
R2 delta Enter the negative exponent for a base-10 constant delta (delta = 10-RDelta) in the R2 delta field; delta is used by STATISTICA as a criterion for determining whether to stop extracting additional PLS components. Extraction of components will continue until the increase in the average R-square value for the Y (dependent or response) variables becomes less than delta, or until the average R-square value for the X variables (in the design matrix) becomes larger than 1-delta, or until the number of components exceeds the number specified in the Max. components field (see below), or via the NCOMPO keyword.
Max. iterations. Specify the maximum number of iterations for the iterative computation of eigenvectors for each PLS component in the Max. iterations field. The default value is 200. PLS uses an iterative power method (see Golub and van Loan, 1996) to compute the eigenvector of Y'XX'Y for each component (where X is the design matrix, and Y is the matrix of dependent or response variables). The number of iterations will also be affected by the value specified in the Eigen delta field (see above), or via the EDELTA keyword.
Max. components Specify the maximum number of components to be extracted in the Max. components field; extraction of components will continue until the number of components exceeds the number specified in this field, or until the increase in the average R-square value for the Y (dependent or response) variables becomes less than 10-RDelta, or until the average R-square value for the X variables (in the design matrix) becomes larger than 1-10-RDelta; RDelta can be specified via the R2 delta field (see above).

For alternative ways of specifying designs in PLS, see Methods for Specifying Designs. For additional details about these options, see the Introductory Overview or the Partial Least Squares (PLS) Index.