projectR-methods: Generic projectR function

In projectR: Functions for the projection of weights from PCA, CoGAPS, NMF, correlation, and clustering

Description

A function for the projection of new data into a previously defined feature space.

Usage

projectR(data, loadings, dataNames = NULL, loadingsNames = NULL, ...)

## S4 method for signature 'matrix,matrix'
projectR(
  data,
  loadings,
  dataNames = NULL,
  loadingsNames = NULL,
  NP = NA,
  full = FALSE,
  family = "gaussianff",
  bootstrapPval = FALSE,
  bootIter = 1000
)

## S4 method for signature 'matrix,LinearEmbeddingMatrix'
projectR(
  data,
  loadings,
  dataNames = NULL,
  loadingsNames = NULL,
  NP = NA,
  full = FALSE,
  model = NA,
  family = "gaussianff",
  bootstrapPval = FALSE,
  bootIter = 1000
)

## S4 method for signature 'matrix,prcomp'
projectR(
  data,
  loadings,
  dataNames = NULL,
  loadingsNames = NULL,
  NP = NA,
  full = FALSE
)

## S4 method for signature 'matrix,rotatoR'
projectR(
  data,
  loadings,
  dataNames = NULL,
  loadingsNames = NULL,
  NP = NA,
  full = FALSE
)

## S4 method for signature 'matrix,correlateR'
projectR(
  data,
  loadings,
  dataNames = NULL,
  loadingsNames = NULL,
  NP = NA,
  full = FALSE,
  bootstrapPval = FALSE,
  bootIter = 1000
)

## S4 method for signature 'matrix,hclust'
projectR(
  data,
  loadings,
  dataNames = NULL,
  loadingsNames = NULL,
  full = FALSE,
  targetNumPatterns,
  sourceData,
  bootstrapPval = FALSE,
  bootIter = 1000
)

## S4 method for signature 'matrix,kmeans'
projectR(
  data,
  loadings,
  dataNames = NULL,
  loadingsNames = NULL,
  full = FALSE,
  sourceData,
  bootstrapPval = FALSE,
  bootIter = 1000
)

Arguments

data	Target dataset into which you will project. It must of type matrix.
loadings	loadings learned from source dataset.
dataNames	a vector containing unique name, i.e. gene names, for the rows of the target dataset to be used to match features with the loadings, if not provided by rownames(data). Order of names in vector must match order of rows in data.
loadingsNames	a vector containing unique names, i.e. gene names, for the rows ofloadings to be used to match features with the data, if not provided by rownames(loadings). Order of names in vector must match order of rows in loadings.
...	Additional arguments to projectR
NP	vector of integers indicating which columns of loadings object to use. The default of NP=NA will use entire matrix.
full	logical indicating whether to return the full model solution. By default only the new pattern object is returned.
family	VGAM family function for model fitting (default: "gaussianff")
bootstrapPval	logical to indicate whether to generate p-values using bootstrap, not available for prcomp and rotatoR objects
bootIter	number of bootstrap iterations, default = 1000
model	Optional arguements to choose method for projection
targetNumPatterns	desired number of patterns with hclust
sourceData	data used to create cluster object

Details

loadings can belong to one of several classes depending on upstream analysis. Currently permitted classes are matrix, CogapsResult, CoGAPS, pclust, prcomp, rotatoR, and correlateR.

Value

A matrix of sample weights for each input basis in the loadings matrix (if full=TRUE, full model solution is returned).

Examples

projectR(data=p.ESepiGen4c1l$mRNA.Seq,loadings=AP.RNAseq6l3c3t$Amean, 
dataNames = map.ESepiGen4c1l[["GeneSymbols"]])

library("CoGAPS")
CR.RNAseq6l3c3t <- CoGAPS(p.RNAseq6l3c3t, params = new("CogapsParams",
nPatterns=5))
projectR(data=p.ESepiGen4c1l$mRNA.Seq,loadings=CR.RNAseq6l3c3t,
dataNames = map.ESepiGen4c1l[["GeneSymbols"]])

pca.RNAseq6l3c3t<-prcomp(t(p.RNAseq6l3c3t))
pca.ESepiGen4c1l<-projectR(data=p.ESepiGen4c1l$mRNA.Seq, 
loadings=pca.RNAseq6l3c3t, dataNames = map.ESepiGen4c1l[["GeneSymbols"]])

pca.RNAseq6l3c3t<-prcomp(t(p.RNAseq6l3c3t))
r.RNAseq6l3c3t<-rotatoR(1,1,-1,-1,pca.RNAseq6l3c3t$rotation[,1:2])
pca.ESepiGen4c1l<-projectR(data=p.ESepiGen4c1l$mRNA.Seq, 
loadings=r.RNAseq6l3c3t, dataNames = map.ESepiGen4c1l[["GeneSymbols"]])

c.RNAseq6l3c3t<-correlateR(genes="T", dat=p.RNAseq6l3c3t, threshtype="N", 
threshold=10, absR=TRUE)
cor.ESepiGen4c1l<-projectR(data=p.ESepiGen4c1l$mRNA.Seq, loadings=c.RNAseq6l3c3t, 
NP="PositiveCOR", dataNames = map.ESepiGen4c1l[["GeneSymbols"]])

projectR documentation built on Nov. 8, 2020, 7:50 p.m.