MS Metabolomics Data Processing Utilities

Description Usage Arguments Value References See Also Examples

Apply k-nearest neighbor (KNN) imputation to a matrix or poplin object. This is an interface to the kNN function from the VIM package. Since it is based on Gower's distance, Standardization of input data before imputation would not affect the result.

## S4 method for signature 'matrix'
impute_knn(x, by = c("feature", "sample"), ...)

## S4 method for signature 'poplin'
impute_knn(x, xin, xout, by = c("feature", "sample"), ...)

`x`	A matrix or poplin object.
`by`	Character controlling whether the imputation is performed by k-nearest features or by k-nearest samples. Either "feature" or "sample".
`...`	Additional arguments passed to kNN.
`xin`	Character specifying the name of data to retrieve from `x` when `x` is a poplin object.
`xout`	Character specifying the name of data to store in `x` when `x` is a poplin object.

A matrix or poplin object of the same dimension as x containing the imputed intensities.

Alexander Kowarik, Matthias Templ (2016). Imputation with the R Package VIM. Journal of Statistical Software, 74(7), 1-16. doi:10.18637/jss.v074.i07

Gower, J. C. (1971). A General Coefficient of Similarity and Some of Its Properties. Biometrics, 27(4), 857–871. https://doi.org/10.2307/2528823

Other imputation methods: impute_pca(), impute_randomforest(), impute_univariate(), poplin_impute()

data(faahko_poplin)

if (requireNamespace("VIM", quietly = TRUE)) {
  ## poplin object
  impute_knn(faahko_poplin, xin = "raw", xout = "knn")

  ## matrix
  m <- poplin_raw(faahko_poplin, "raw")
  impute_knn(m)
}