Description: Matrix decomposition techniques are used to find main features of a dataset. These features can be used to describe the data or to find clusters of objects. The decompositions have problems when the matrices have missing data or are from different exposures. This article examines the issue and some papers dealing with it.