dot-imputeMissingGeneExpression: Imputing Missing Data

In JRC-COMBINE/PhysioSpaceMethods: Creates and uses physio spaces as a dimension reduction mapping

Description

.imputeMissingGeneExpression is an internal function called by inputChecker. It uses different methods to impute any missing value of the InputData or Space (or else missing values would break the pipeline in calculatePhysioMap).

Usage

.imputeMissingGeneExpression(InptGEX, METHOD = "PCA")

Arguments

InptGEX	Input matrix with missng values.
METHOD	Method to use in imputation. Available methods are KNN and PCA. Default is 'PCA'.

Value

A matrix with the same dimensions as InptGEX, with missing values imputed.

Examples

## Not run: 
 MatToImpute <-
   matrix(
     rnorm(n = 100000, mean = 0, sd = 100),
     ncol = 10,
     dimnames = list(1:10000, 1:10)
   )
 MatToImpute[sample(x = 1:length(MatToImpute),
                    size = length(MatToImpute) / 20)] <- NA
 ImputedMat <-
   PhysioSpaceMethods:::.imputeMissingGeneExpression(InptGEX = MatToImpute,
   METHOD = "PCA")

## End(Not run)

JRC-COMBINE/PhysioSpaceMethods documentation built on July 27, 2021, 12:53 p.m.