K means partitioning prior to HCPC

[Wenyuan Yin]

Hi all,

I am new to FactoMineR package so forgive me if I am asking some ambiguous questions here. I have a dataset with about 56 variables and variables are mixed types(categorical and continuous). I learned that MFA and FAMD can be used prior to HCPC function for the purpose of dimension reduction. So I ran both of them and use them as the input for HCPC. I got the result but encountered some questions. Since I still had about 38 variables left and each of them has multiple values, when I ran HCPC without any partitioning first, I ended up getting 4 major clusters which have lots of variable values within each and made it a little difficult for me to interpret the definition of each cluster. Then I learned that I can do k means partition before actual HCPC using kk = number. I applied and got the chart like the one below. Basically, it makes hierarchy much more clear but I am wondering is there any way I can get the definition of the sub-cluster 1-25, like the variables contribution to those sub-cluster etc, since the final result I got was still the relationship between individuals/variables and major 4 clusters.

I researched online and it seems cutree function could give different views of branches of the whole tree;meanwhile, keep hierarchy unchanged but I am not sure how to do it.

Let me know if you need more clarification from me.

Thank you so much for your help in advance!

Best,

Wenyuan

                                                                                                                                                                                                 

[François Husson]

In fact, the partition should be approximately the same with or without K-means before hierarchical clustering. K-means is used before hierarchical clustering when there is a lot of people and when hierarchical clustering cannot be performed on all the dataset.

So consider your first analysis. And if there is a lot of variables that describe each cluster, it is not really a problem because variables are sorted from the most important to the least.

FH

[Wenyuan Yin]

Thank you so much for your explanation!

Wenyuan

无病毒。www.avast.com

--
Vous recevez ce message, car vous êtes abonné à un sujet dans le groupe Google Groupes "FactoMineR users".
Pour vous désabonner de ce sujet, visitez le site https://groups.google.com/d/topic/factominer-users/eToX0jkaYnw/unsubscribe.
Pour vous désabonner de ce groupe et de tous ses sujets, envoyez un e-mail à l'adresse factominer-users+unsubscribe@googlegroups.com.
Pour obtenir davantage d'options, consultez la page https://groups.google.com/d/optout.