evaluateSVM: function to do prediction and evaluate based on the results...

In lizhongliu1996/PredictMisc: package that use three methods to predict methylation regions

Description

function to do prediction and evaluate based on the results of support vector machine

Usage

evaluateSVM(caret_fit, testTable, respCol_index, whichRep_int,
  whichCVfold_int)

Arguments

caret_fit	results of random forest or support vector machine
testTable	test data frame that each column represents a CpG probe while each row represents a sample, each cell is a M value, first column is phenodata
respCol_index	col number of response variable
whichRep_int	repetition num in use right now, this decides on the traning data to be used
whichCVfold_int	crossvalidation num in use right now, this decides on the traning data to be used

Value

A data frame of containing the train and test beta matrix

• auc_results : auc value of prediction

• Sensitivity : Sensitivity value of prediction

• Specificity : Specificity value of prediction

• ... : more index

Examples

## Not run: 
   data(ExampleMvalue_test)
   data(svm_Fit)

   test <- evaluateSVM(
     caret_fit = svm_Fit,
     testTable = ExampleMvalue_test,
     respCol_index = 1,
     whichRep_int = 1,
     whichCVfold_int = 1
   )
 
## End(Not run)

lizhongliu1996/PredictMisc documentation built on Aug. 23, 2019, 5:55 a.m.