Venn Diagram in STATA

[Ankit Bhardwaj]

Hi, I want to install venn diagram supporting files in STATA. tried ssc install venn & net install venn, from("http://fmwww.bc.edu/repec/bocode/v").not helpful.pls help

[Maarten Buis]

The command venn does not exist on SSC, so that explains why those commands did not work. You can get a list of all commands whose name start with a v by typing ssc desc v. Maybe you wanted to install venndiag (which seems very old, and you may experience problems exporting the resulting graph) or vennbar?

[John Mullahy]

I have occasionally used pvenn and see that there is now a pvenn2 available.

search pvenn

[Ankit Bhardwaj]

thanks a lot!

[Marc Kaulisch]

You may also want to have a look at -upsetplot- (SSC) see

net describe upsetplot, from(http://fmwww.bc.edu/RePEc/bocode/u)

[Nick Cox]

Thanks for the mention of upsetplot

The Stata Journal publication supersedes the SSC version. and in turn a fix was signalled in https://www.statalist.org/forums/for...n-in-upsetplot

The fix will be published formally in Stata Journal 26(2) due in June or so.

Venn diagrams strict sense have been addressed by various user-programmers:

    Lauritsen, J.M. 1999a.  Drawing Venn diagrams.  Stata Technical Bulletin 47: 3-8.

    Lauritsen, J.M. 1999b.  Drawing Venn diagrams.  Stata Technical Bulletin 48: 3.

    Lauritsen, J.M. 1999c.  Drawing Venn diagrams.  Stata Technical Bulletin 49: 8.

    Lauritsen, J.M. 2000.  An update to drawing Venn diagrams.  Stata Technical Bulletin 54: 17-19.

    Lauritsen, J.M. 2009.  venndiag: module to generate Venn diagrams.  http://fmwww.bc.edu/RePEc/bocode/v 

    Gong, W. and J. Ostermann. 2011.  pvenn: module to create proportional Venn diagram.  http://fmwww.bc.edu/RePEc/bocode/p 

    Over, M. 2022.  pvenn2: Proportional Venn diagram, enhanced version of pvenn.  http://digital.cgdev.org/doc/stata/MO/Misc

[Sergiy Radyakin]

See also this version with patterned fills:
http://www.radyakin.org/stata/venndi...enndiagram.htm

[Nick Cox]

The original question leaves open two key details:

1. How many sets are to be displayed. With four or more intersecting sets that becomes tricky to do and very tricky to do well.

2. Whether the real focus of interest is in the frequencies (proportions, percents, whatever) in each portion of the Venn diagram. It is hard to show numbers well on a Venn diagram, or at least to do that more clearly than in a table.

The paper by Tim Morris and myself alluded to in #6 says much more, We line up with a body of opinion that Venn diagrams can be enormously helpful for discussion of questions in elementary logic or probability, but they can be fairly useless for data display, let alone data analysis.

The display in the link to Sergiy Radyakin 's #7 is a case in point. It shows three intersecting sets and the possibillity of belonging to none. So 2^3 = 8 subsets.The coding is undoubtedly clever but once you're decoded the legend are you better off than with a table -- or than with an alternative display?