Out of interest do all of your 34 variables have the potential to change week to week or can some of them be handled through classifications? Also, of the variable where classifications don't work because they change week to week, how many have a known/limited set of possible values? e.g. I'm assumi...