Data Visualization - Further Step than Correspondence Analysis

#1
Hi,

I've currently been using correspondence to visualize our different consumer groups with each other.

I have been comparing my column groups (AA, BB, CC) with my row groups (A, B, C). The cross section of groups all have affinities with each other that I've been able to use XLSTAT to visualize the relationship (e.g. AA is closely tied to B).

However, I also have data showing the affinity between AA and BB, AA and CC, A and B, A and C, etc. Correspondence analysis will only show the visualization between the row and columns, but not with each other. I'd like to see at the same time how closely related AA is to BB, etc.

So I have a matrix with columns (AA, BB, CC, A, B, C) and rows (AA, BB, CC, A, B, C). What kind of data visualization analysis can graph all these affinities for me? Which software would be appropriate for my use?

Thanks,
 

gianmarco

TS Contributor
#2
Hello,
I have experience with CA, so I could give you some help, BUT I do not quite follow you when you say:
Correspondence analysis will only show the visualization between the row and columns, but not with each other. I'd like to see at the same time how closely related AA is to BB, etc.
CA allows to display rows AND columns points; the distance between points of the same type (i.e., distance between row points) is telling you something about how much similar the row-profile points actually are.
If you can clarify your point, I could possibly give more help.


I have built a website on Correspondence Analysis (from an archaeological perspective):
http://cainarchaeology.weebly.com/

There, you can also find a description of an R package of mine aimed at facilitating the interpretation of CA results.


Hope this helps,
Best
Gm