[R] A useful GUI?

Peter Dalgaard BSA p.dalgaard at biostat.ku.dk
Tue Jul 24 00:15:39 CEST 2001

Frank E Harrell Jr <fharrell at virginia.edu> writes:

> I saw an ad in the July 2001 Amstat News for a complex sample
> survey analysis package called Wesvar.  The package has
> what appears to be a useful feature called a "Workbook"
> by which the user navigates analysis output.  This is a
> hierarchical tree in which the user may click on a part
> of the analysis (table, regression fit, descriptive stats,
> etc.) so as to go directly to that output.  Recent versions
> of SAS have something similar.  In SAS the title of a
> graphic may be what appears in the navigation tree, and
> clicking on that entry will replay the appropriate graph.
> Has anyone had experience with such GUIs and would comment
> on productivity or documentation benefits?  I wonder
> what would be the most platform-independent way to
> implement such a feature, e.g., dynamic html, if it is
> useful to statisticians.

I've tried the SAS stuff (the Analyst application). It's a fairly neat
idea, but you do need to put in a substantial effort to relabel the
stuff and delete any "junk" output, or find yourself with a bunch of
litle icons all having similar generic names like "Probability plot" or
"ANOVA output" (I forget the exact names, but you get the picture).
There was a strange division between SAS programming and Analyst in
that *only* code executed from Analyst would go into the project tree,
nicely incorporating the SAS code that Analyst had written for you,
but excuting SAS code directly would go to a *different* tree, which
did *not* include the code that was executed.

The version I tried (SAS 8.0) was also bug-ridden beyond belief.
Suddenly, the system would get itself into a state where the tree
could no longer be saved, detracting rather badly from the usefulness.
They might have fixed that in the meantime though.

The newer Enterprise Guide stuff is reportedly neater, but we never
succeeded in getting it to run on any of our PC SAS installations
(seemingly, it was not prepared for our particular network

It's the sort of thing that it might be easy to code in Tcl/Tk, and
which I do have some plans of returning to .. in the fullness of time
. when other obligations allow .. etc.... 

   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list