SvmFit: support vector machine

In NormalizeMets: Analysis of Metabolomics Data

Description

Classification using support vector machine (svm) algorithm

Usage

SvmFit(featuredata, groupdata, kernel = "radial", cost = 1, gamma = NULL,
  crossvalid = TRUE, k = 5, tune = FALSE, pred = TRUE,
  pfeaturedata = featuredata, pgroupdata = groupdata, rocplot = TRUE,
  saveoutput = FALSE, outputname = "svm", main = NULL, ...)

Arguments

featuredata	featuredata A data frame in the featuredata format. This is a dataframe with metabolites in columns and samples in rows. Unique sample names should be provided as row names. See NormalizeMets Vignette for details.
groupdata	A vector with group names.
kernel	The kernel used. The default is the radial basis function with type C-classification.
cost	of constraint violation, defaults to 1.
gamma	parameter used for the kernel
crossvalid	A logical indicating whether cross-validation needs to be conducted
k	An integer specifying the k-fold cross-validation. Default is set to 5.
tune	A logical with the default set to FALSE. If TRUE, a grid search will be conducted to tune the hyperparameters, over parameter ranges supplied by the user.
pred	whether the predictions should be made
pfeaturedata	The test dataset for the predictions. The default is featuredata
pgroupdata	The test groupdata for the predictions. The default is groupdata
rocplot	A logical indicating whether a receiver operating characteristic curve needs to be plotted, along with the area under the curve (AUC) printed.
saveoutput	A logical indicating whether the outputs should be saved in the format write.svm
outputname	The name of the output file if the output has to be saved.
main	Plot title.
...	Arguments to be passed on to other methods.

Value

If tune=FALSE, an object of class "svm" svm containing the fitted model or if tune=TRUE, an object of class tune

Author(s)

Alysha M De Livera, Gavriel Olshansky

See Also

tune

Examples

 data(alldataC)
SvmFit(featuredata=alldataC$featuredataC, 
       groupdata=alldataC$groupdataC,
       crossvalid=TRUE,
       k=5,
       rocplot = TRUE)

NormalizeMets documentation built on May 1, 2019, 10:26 p.m.