predict.networkBasedSVM: Predict Method for Network-based SVM Fits

In pathClass: Classification using biological pathways as prior knowledge

Description

Obtains predictions from a fitted networkBasedSVM object.

Usage

  ## S3 method for class 'networkBasedSVM'
predict(object, newdata, ...)

Arguments

object	a fitted object of class inheriting from 'networkBasedSVM'
newdata	a matrix with variables to predict
...	currently ignored.

Value

the predictions.

Author(s)

Marc Johannes [e-mail]

Examples

## Not run: 
library(pathClass)
data(example_data)
matched <- matchMatrices(x=x, adjacency=adjacency.matrix, mapping=mapping)
ad.list <- as.adjacencyList(matched$adjacency)
fit = fit.networkBasedSVM(matched$x[1:5,], y[1:5], DEBUG=TRUE,  adjacencyList=ad.list,
 lambdas=10^(-1:2), sd.cutoff=50)
predict(fit, newdata=matched$x[6:10,])

## End(Not run)

pathClass documentation built on May 29, 2017, 11:44 p.m.