Icon

Multivariate

3 workflows in this folder and its subfolders

FAMD

Factor analysis of mixed data - allow aaplying a PCA kernel to mixed datas numericals and strings. Please note that you have to remove string variables with too many categories such as individual idents before treatment. Usually this treatment is used before a clustering method as Kmeans. https://en…

Icon
optimal number of groups

Calculate the optimal number of groups for a future Kmeans with classical intergoup / intragroup variance simulation. If the maximum number of groups declared <= 2, first calculate the maximum reasonable number of groups (can be very long) using a confirmatory logistic regression or dicriminant anal…

QualClassif

Complete classification admitting numerical and categorical variables in entry. Perform AFMD then Xmeans and logistic regression in the end to validate the Xmeans results and reprocess the classification. Generate sql instructions in order to reproduce the classification in a most automatized way di…