How to Make Market Basket Plots?I have a customer who wants to make association visualizations that I believe are known as market basket plots (see below). I have seen examples of these plots created via script by others. Please reply if you know of a script or add-in that could be shared. Ideas on other ways to get similar information would be great as well (SEM?). Thanks!
Re: How to Make Market Basket Plots?I don't have an answer for you Rob, but I am also interested. These kinds of plots are very useful in Organizational Network Analysis (ONA), looking at interactions between people. They appear similar to Constellation plots, but obviously not identical. I think R has a package to do this, but it would be very nice if JMP also offered it.
Looking forward to an official reply!
Re: How to Make Market Basket Plots?Hi Rob, I had a look into this a little while ago as I wanted to model customers transitioning between STIPs modules. More of a Markov decision model but would display similar to your customers needs.
Similar to Jerry, I didn't find a solution in JMP. There are hints towards fringes of functionality in:
https://community.jmp.com/t5/Discussions/Network-Association-Diagram/td-p/7633
https://community.jmp.com/t5/JMP-Wish-List/Network-Diagram-for-showing-relationships/idi-p/234004
https://www.jmp.com/support/help/14-2/associative-arrays-in-graph-theory.shtml
But definitely not the full solution.
I haven't found the time to acquire and understand SAS' Enterprise Miner or the R package functionality. Perhaps the latter would be best to be put in an Add-In?
In lieu of an official reply, happy to work together on this as there seems to be suitable case studies where we can demonstrate value.
Kind Regards,
Ben
Re: How to Make Market Basket Plots?There is an add-in called partial correlation diagrams that will produce a similar graph based on partial correlations.. You must have numeric data however. Maybe this will help?
Partial Correlation Diagram - JMP User Community
Re: How to Make Market Basket Plots?Pretty cool, and pretty close Chris! Thanks!
