buildConfMat: Build confounder design matrices with and without intercepts

In combi: Compositional omics model based visual integration

Description

Build confounder design matrices with and without intercepts

Usage

buildConfMat(confounders)

Arguments

confounders

A dataframe of confounding variables #' For the preliminary trimming, we do not include an intercept, but we do include all the levels of the factors using contrasts=FALSE: we want to do the trimming in every subgroup, so no hidden reference levels For the filtering we just use a model with an intercept and treatment coding, here the interest is only in adjusting the offset

Value

a list with components

confModelMatTrim	A confounder matrix without intercept, with all levels of factors present. This will be used to trim out taxa that have zero abundances in any subgroup defined by confounders
confModelMat	A confounder matrix with intercept, and with reference levels for factors absent. This will be used to fit the model to modify the independence model, and may include continuous variables

combi documentation built on Nov. 8, 2020, 5:34 p.m.