predictNewClasses: Class Prediction

In cdeterman/OmicsMarkeR: Classification and Feature Selection for 'Omics' Datasets

Description

This function evaluates a single fitted model and returns the predicted group memberships of new data.

Usage

predictNewClasses(modelFit, method, orig.data, newdata, param = NULL)

Arguments

modelFit	The fitted model being evaluated
method	String of the model to be evaluated
orig.data	The orginal data before subsetting training sets. Required to have the 'observed' group membership
newdata	The testing data to predict group membership
param	Optional alternate parameters being fit to the model

Value

Returns a list of predicted group membership

Examples

dat.discr <- create.discr.matrix(
    create.corr.matrix(
        create.random.matrix(nvar = 50, 
                             nsamp = 100, 
                             st.dev = 1, 
                             perturb = 0.2)),
    D = 10
)

vars <- dat.discr$discr.mat
groups <- dat.discr$classes

fits <- fs.stability(vars, 
                     groups, 
                     method = c("plsda", "rf"), 
                     f = 10, 
                     k = 3, 
                     k.folds = 10, 
                     verbose = 'none')

newdata <- create.discr.matrix(
    create.corr.matrix(
        create.random.matrix(nvar = 50, 
                             nsamp = 100, 
                             st.dev = 1, 
                             perturb = 0.2)),
    D = 10
)$discr.mat

orig.df <- data.frame(vars, groups)

# see what the PLSDA predicts for the new data
# NOTE, newdata does not require a .classes column
predictNewClasses(fits, "plsda", orig.df, newdata)

cdeterman/OmicsMarkeR documentation built on May 13, 2019, 2:35 p.m.