LLSimpute {BNArray}R Documentation

local least squares imputation

Description

Impute missing values using local least squares algorithm.

Usage

LLSimpute(df.all, df.ori, n.changed)

Arguments

df.all an object of data.frame which contains whole microarray data
df.ori an object of data.frame which contains microarray data of differentially expressed genes
n.changed the number of differentially expressed genes

Details

LLSimpute employs local least squares algorithm to impute missing values in microarray data. Local least squares algorithm estimates missing values in a gene express vector according to the vector's most similar complete express vectors. The missing values are computed as the linear combination of the corresponding data in the nearest neighbor genes. Normally the coefficients could be computed but in rare cases when one gene misses too many values in total experiments, they take values NA. Therefore, after calling LLSimpute, we have to FinalImpute function to ensure a complete database for constructing Bayesian networks.

Value

A data.frame object is returned with imputed data.

Author(s)

Xiaohui Chen cxh1984@interchange.ubc.ca,
Kaida Ning aprilnkd@hotmail.com,
Ming Chen mchen@zju.edu.cn.

References

Hyunsoo Kim, Gene H. Golub and Haesum Park (2005) Missing value estimation for DNA microarray gene expression data: local least squares imputation, Bioinformatcs, 21, 187-198

See Also

FinalImpute

Examples

## Not run: 
        library(BNArray)
        data(total.data)
        setwd("/home")

##SHOULD call LLSimpute first
        ori.compact<-LLSimpute(total.data$df.all,total.data$df.ori,total.data$n.changed)## End(Not run)

[Package BNArray version 1.0 Index]