Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Nick Cox <n.j.cox@durham.ac.uk> |
To | "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu> |
Subject | st: RE: mosaic plot with an intensity dimension |
Date | Fri, 24 Feb 2012 15:17:55 +0000 |
-spineplot- is from SJ. Please remember to explain _where_ user-written programs you refer to come from. I don't think the literature on graphics is encouraging on quantitative decoding of colour intensity. Either way, as you know, -spineplot- does not support what you want directly. An alternative plot that is quite easy to arrange, but may or may not work well, is egen group = group(y x), label spineplot z group Nick n.j.cox@durham.ac.uk Dimitriy V. Masterov I would am trying to create the graph of customer data where y is categorical customer type (4 integer values), x is binary new customer variable, and z is a binary purchased or not variable. I would like to create a mosaic plot of y and x to show the sizes of the groups, but somehow express the percentage of customers who purchase the product for each group using color intensity. I want to show that some group has a really high purchase rate, but they are a small fraction of the customers. Right now I do this with user-written spineplot using the text option: bys type new_cust: egen mean=mean(reorder_90) replace mean=mean*100 gen mean2=string(mean, "%4.0f") spineplot type new_cust, text(mean2) percent This uses numbers rather than color intensity, and is harder to read or to see patterns, at least for my brain. A rough example of what I want to see is http://newsmap.jp, where the color intensity indicates the number of related articles. Is there a way of expressing the mean with color intensity? Is there a better way of visualizing this data? * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/