impute.KNN_TN,proFIAset-method {proFIA} | R Documentation |
Impute the missing values in an FIA experiment using a Weighted K-Nearest Neighbours on Truncated Distribution described by Jasmit S. Shah et al.
## S4 method for signature 'proFIAset' impute.KNN_TN(object, k = 0.6, classes = c("split", "unique"))
object |
A proFIAset object. |
k |
The number of neighbors considered, can be a fraction then in this case the k will be taken for each class as the frac of the effective of the class. 3 at minima because comparison is based on correlation. |
classes |
how to handle imputation for different classes, if 'split', the classes are taken separately, if 'unique', the imputation is done on the full data matrix. |
A proFIAset object with the missing values imputated.
Distribution based nearest neighbor imputation for truncated high dimensional data with applications to pre-clinical and clinical metabolomics studies, J.S Shah 2017, BMC Bioinformatics.
if(require(plasFIA)){ data(plasSet) ###Reinitializing the data matrix plasSet<-makeDataMatrix(plasSet,maxo=FALSE) plasSet<-impute.KNN_TN(plasSet,2) }