extract.features {OmicsMarkeR} | R Documentation |
Extracts features from models that have been previously fit.
extract.features(x, dat = NULL, grp = NULL, method, model.features = FALSE, bestTune = NULL, f, comp.catch = NULL)
x |
Previously fitted model |
dat |
Numeric variable data used for fitted models (In appropriate format) |
grp |
Vector of training classes |
method |
String indicating the INDIVIDUAL model being extracted from |
model.features |
Logical argument dictating if features selected determined by models instead of user determined number of features. |
bestTune |
If |
f |
Number of features to subset |
comp.catch |
An internal check for plsda models. If the optimal model contains only 1 component, the ncomp paramter must be set to 2 for the model. However, features are still extracted only from the first component. |
Returns list of the features selected from the fitted model.