Fit Mixture Models Using the Expectation Maximisation (EM) Algorithm
Count the number of unique items ion a vector x
Prepare data for use with multimix
The E(xpectation) step
Initialise the parameter list.
Map integer index N>0 back to left member of generating pair.
Make initial Z matrix from initial assignment of observations to clust...
Read Z from FORTRAN output. Make into R matrix
Start from random groups of similar size.
Title
The M(aximisation) step
multimix: Fit Mixture Models Using the Expectation Maximisation (EM) A...
Maps integer pairs (u,v) with 0<u<v bijectively to positive integers.
S3 method for plotting multimix results objects
S3 printing method for for multimix parameter results
S3 method for the printing of multimix results
Map integer index N>0 back to right member of generating pair.
A set of functions which use the Expectation Maximisation (EM) algorithm (Dempster, A. P., Laird, N. M., and Rubin, D. B. (1977) <doi:10.1111/j.2517-6161.1977.tb01600.x> Maximum likelihood from incomplete data via the EM algorithm, Journal of the Royal Statistical Society, 39(1), 1--22) to take a finite mixture model approach to clustering. The package is designed to cluster multivariate data that have categorical and continuous variables and that possibly contain missing values. The method is described in Hunt, L. and Jorgensen, M. (1999) <doi:10.1111/1467-842X.00071> Australian & New Zealand Journal of Statistics 41(2), 153--171 and Hunt, L. and Jorgensen, M. (2003) <doi:10.1016/S0167-9473(02)00190-1> Mixture model clustering for mixed data with missing information, Computational Statistics & Data Analysis, 41(3-4), 429--440.