SparseFactorAnalysis (version 1.0)
Scaling Count and Binary Data with Sparse Factor Analysis
Description
Multidimensional scaling provides a means of uncovering a latent structure underlying observed data, while estimating the number of latent dimensions. This package presents a means for scaling binary and count data, for example the votes and word counts for legislators. Future work will include an EM implementation and extend this work to ordinal and continuous data.