ermutation test for OrthPLS, running p permutations of Y.
Utilises the JTPpermutate() method written by Jake Pearce.
Compares the OrthPLS model against a randomised version to test the validity.
model = csm_orth_pls_permutate( spectra, Y, p );
model = csm_orth_pls( spectra, Y, p, 'num_pred_comp', num_pred_comp, 'num_Y_orth_comp', num_Y_orth_comp, 'num_cv_rounds',num_cv_rounds, 'scale_type', scale_type, 'model_type', model_type);
| Variable | Type | Default Value | Description |
|---|---|---|---|
| *spectra | csm_spectra | None | csm_spectra object containing spectral matrix. |
| *Y | m*1 | None | Matrix of predictors - Orthogonal components (For discriminant analysis this is a vector of 0/1's to define class) |
| *p | 1x1 | None | Number of permutations. |
| num_pred_comp | 1*1 | None | Number of predictive components. |
| num_Y_orth_comp | 1*1 | None | Number of Y-orthogonal components (OC in X). The number of components in A+oax should be kept to a minimum to prevent overfitting. |
| num_cv_rounds | 1*1 | None | Number of cross validation rounds (Default 7). |
| scale_type | str | 'mc' | Preprocessing; 'mc' for mean-centering, 'pa' for pareto, 'uv' for unit-variance scaling, or 'none' |
| model_type | str | 'da' | 'da' for discriminant analysis, 're' for regression. if 'da', sensitivity and specificity will be calculated |
| Variable | Type | Description |
|---|---|---|
| csm_orth_pls_permutate | csm_wrapper | csm_wrapper with some stored inputs, the outputs and metadata. |
| csm_orth_pls_permutate.output.pv | 1*1 | P-value (non-parametric). |
| csm_orth_pls_permutate.output.Q2p | 1*n | Q2 values for all permutations. |
| csm_orth_pls_permutate.output.R2p | 1*n | R2 values for all permutations. |
| csm_orth_pls_permutate.output.orth_pls_model | obj | Canonical OrthPLS model. |
Copyright Imperial College London 2019