make_ensemble: Generate predictions based on an ensemble of classification...

In katrinakalantar/metahostlab: Metagenomic Host Labelling of Unlabelled Data (one line, title case)

Description

Generate predictions based on an ensemble of classification algorithms;

Usage

make_ensemble(x, y, library, multiple = FALSE)

Arguments

x	training data; matrix of features/covariates used for prediction.
y	training labels; must be factor of non-integer values i.e. "one"/"two" instead of 1/2
library	A vector of strings indicating the algorithms to be included in the ensemble. Available algorithms can be queried here: https://topepo.github.io/caret/available-models.html
multiple	Boolean value indicating whether to run multiple iterations of cross-validation to generate ensemble predictions; default = FALSE; If TRUE, will perform 5 iterations

Value

A list containing: "MF", samples to be kept by majority filter; "CF", samples to be kept by consensus filter, "full_res", the full list of discordant predictions

katrinakalantar/metahostlab documentation built on May 26, 2019, 2:31 p.m.