VarSel {fuseMLR} | R Documentation |
Varsel Class
Description
This class implements a learner. A VarSel object can only exist as a component of a TrainLayer or a TrainMetaLayer object.
Methods
Public methods
Method new()
Variable selection parameter list.
Learner ID.
Usage
VarSel$new( id, package = NULL, varsel_fct, varsel_param, train_layer, na_action = "na.rm" )
Arguments
id
character
Package that implements the variable selection function. If NULL, the variable selection function is called from the current environment.package
character
Variable selection function name. Note: Variable selection functions, exceptBoruta
, must return a vector of selected variables.varsel_fct
character
Variable selection parameters.varsel_param
list
Layer on which the learner is stored.train_layer
TrainLayer
The training layer where to store the learner.na_action
character
Handling of missing values in meta-data. Set to "na.keep" to keep missing values, "na.rm" to remove individuals with missing values or "na.impute" (only applicable on meta-data) to impute missing values in meta-data. Only median and mode based imputations are actually handled. With the "na.keep" option, ensure that the provided learner can handle missing values. IfTRUE
, the individuals with missing predictor values will be removed from the training dataset.
Method print()
Printer
Usage
VarSel$print(...)
Arguments
...
any
Method summary()
Summary
Usage
VarSel$summary(...)
Arguments
...
any
Method interface()
Learner and prediction parameter interface. Use this function
to provide how the following parameters are named in the learning
function (lrn_fct
) you provided when creating the learner, or in the predicting function.
Usage
VarSel$interface( x = "x", y = "y", object = "object", data = "data", extract_var_fct = NULL )
Arguments
x
string
Name of the argument to pass the matrix of independent variables in the original learning function.y
string
Name of the argument to pass the response variable in the original learning function.object
string
Name of the argument to pass the model in the original predicting function.data
character
Name of the argument to pass new data in the original predicting function.extract_var_fct
character
orfunction
If the variable selection function that is called does not return a vector, then use this argument to specify a (or a name of a) function that can be used to extract vector of selected variables. Default value is NULL, if selected variables are in a vector.
Method varSelection()
Tains the current learner (from class Lrner) on the current training data (from class TrainData).
Usage
VarSel$varSelection(ind_subset = NULL)
Arguments
ind_subset
vector
Individual ID subset on which the training will be performed.
Returns
The resulting model, from class Model, is returned.
Method getTrainLayer()
The current layer is returned.
Usage
VarSel$getTrainLayer()
Returns
TrainLayer object.
Method getId()
Getter of the current learner ID.
Usage
VarSel$getId()
Returns
The current learner ID.
Method getPackage()
Getter of the variable selection package implementing the variable selection function.
Usage
VarSel$getPackage()
Returns
The name of the package implementing the variable selection function.
Method getVarSubSet()
Getter of the list of selected variables.
Usage
VarSel$getVarSubSet()
Returns
List of selected variables..
Method getParamInterface()
The current parameter interface is returned.
Usage
VarSel$getParamInterface()
Returns
A data.frame of interface.
Method getNaAction()
The current layer is returned.
Usage
VarSel$getNaAction()
Method getExtractVar()
The function to extract selected variables is returned.
Usage
VarSel$getExtractVar()
Returns
A data.frame of interface.