Data imputation for visualisation
This repository contains two function which can be used for data imputation before data visualisation by elastic graphs, elastic maps or any other techniques.
Function degup calculates fraction of missing data in each record and each feature and then remove the record or feature with greatest fraction of missing data. This procedure is repeated until there is records or features with missed data.
Function kNNImpute imputes data by weighted mean of k nearest neighbour. Nearest neighbours are defined by known values and intervals of distribution of unknown values.
Function svdWithGaps imputes data by decomposition of data matrix 'data' into singular vectors and later reconstruct all values.
Please refer as Mirkes, EM, Zinovyev A, Gorban, AN “Data imputation”, available online on https://github.com/Mirkes/DataImputation, accessed Date_of_Access
Supported by the University of Leicester (UK), Institut Curie (FR), the Ministry of Education and Science of the Russian Federation, project № 14.Y26.31.0022