exprso: The 'exprso' Package

In tpq/exprso: Rapid Deployment of Machine Learning Algorithms

Description

Welcome to the exprso package!

The exprso function imports data into the learning environment.

See mod to process the data.

See split to split off a test set.

See fs to select features.

See build to build models.

See pl to build models high-throughput.

See pipe to process pipelines.

See buildEnsemble to build ensembles.

See exprso-predict to deploy models.

See conjoin to merge objects.

Usage

exprso(x, y, label = 1, switch = FALSE)

Arguments

x	A matrix of feature data for all samples. Rows should contain samples and columns should contain features.
y	A vector of outcomes for all samples. If class(y) == "character" or class(y) == "factor", exprso prepares data for binary or multi-class classification. Else, exprso prepares data for regression. If y is a matrix, the program uses the outcome in label.
label	A numeric scalar or character string. The column to use as the label if y is a matrix.
switch	A logical scalar. Toggles which class label is called Control in binary classification.

Value

An ExprsArray object.

Examples

library(exprso)
data(iris)
array <- exprso(iris[,1:4], iris[,5])
arrays <- splitSample(array, percent.include = 67)
train <- trainingSet(arrays)
test <- testSet(arrays)
train <- fsANOVA(train, top = 0)
train <- fsPrcomp(train, top = 3)
mach <- buildSVM(train, top = 5, kernel = "linear", cost = 1)
pred <- predict(mach, test)
calcStats(pred)

tpq/exprso documentation built on July 27, 2019, 8:44 a.m.