Plots for Nested or hierarchical data

[SKK]

I have nested data with counts for each of them and I need to plot them. After some research I found sunburst chart would be a appropriate model to chart them but couldn't find any SAS procedures for it. Is there any procedures like GMAP or GCHART to program Sunburst charts ?

Also is there any other appropriate model for plotting nested data ? 

[PaalNavestad]

Hi there is an easy treemap option in SAS VA. If you do not have that available I think you'll find some treemap options in sgplot. 

See in the SAS blog Graphically speaking. There are usually some very good ideas there.

[SASKiwi]

Is this the sort of chart you had in mind?

https://communities.sas.com/t5/SAS-Communities-Library/Visualizing-nested-data-with-a-sunburst-chart...

[SKK]

Yes Exactly... But I dont have access to SAS VA.. So is there any other way to do this ?

[PaalNavestad]

You can do it with sgplot or the GTL language. I do not have an example handy.
You can alos look at proc tree from SAS/STAT that show the hierarchy very well. You have to juggle a bit.
I think I've seen an example in the graphically speaking blog. Just google graphically speaking sas and it gets you there

[SKK]

Thanks for the information I will look into that blog... Is it possible to create Sub-grouped donut chart for multiple level nested data ?

[PaalNavestad]

I do not know. I seem to remember I did create a set of pies or donuts using sgchart about 15 years ago. But I can't exactly remenber how  I very seldom use pie charts anymore and for the sunflower chart I had problems seeing the use.

Have you just tried using midpoint/group and subgroup in gchart. I would prefer a treemap. There is a mosaic plot in GTL that you can try.

[Rick_SAS]

Statisticians use mosaic plots to display hierarchical data, although in practice it works best for three or fewer levels of nesting. PROC FREQ can display a mosaic plot automatically, The plot is often used to accompany an analysis/discussion of whether there is association between variables.