screen.FSelector | R Documentation |
A SuperLearner-compatible interface to functions in the
FSelector
package.
screen.FSelector(
Y,
X,
family,
obsWeights,
id,
filter = c("cfs", "chi.squared", "consistency", "gain.ratio", "information.gain",
"linear.correlation", "oneR", "random.forest.importance", "rank.correlation",
"relief", "symmetrical.uncertainty"),
filter_params = NULL,
selector = c("cutoff.biggest.diff", "cutoff.k", "cutoff.k.percent", "all"),
k = switch(selector, cutoff.k = ceiling(0.5 * ncol(X)), cutoff.k.percent = 0.5, NULL),
verbose = FALSE,
...
)
Y |
Outcome (numeric vector). See |
X |
Predictor variable(s) (data.frame or matrix). See
|
family |
Error distribution to be used in the model:
|
obsWeights |
Optional numeric vector of observation weights. Currently unused. |
id |
Cluster identification variable. Currently unused. |
filter |
A string corresponding to a feature ranking or selecting function
implemented in the FSelector package. One of: |
filter_params |
A named list of tuning parameter arguments specific to
the chosen |
selector |
A string corresponding to a subset selecting function
implemented in the FSelector package. One of:
|
k |
Passed through to the |
verbose |
Should debugging messages be printed? Default: |
... |
Currently unused. |
A logical vector with length equal to ncol(X)
.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.