Stata: Data Analysis and Statistical Software
   >> Home >> Resources & support >> Users Group meetings >> 1st North American meeting >> Wishes and grumbles

Wishes and grumbles

(by Jeroen Weesie)

During the North American Stata User Group meeting, the usual Wishes and Grumbles session yielded a wide range of requests for new features of Stata. Bill Gould replied to these requests, sometimes with a promise (meaning, it is very likely that it will be done); sometimes with the more cautious reply that StataCorp will carefully consider the feasability of the request; and sometimes with a frown. You are free to interprete this nonverbal form of communication yourself. Then again, given that Stata can be extended relatively easily, someone outside of StataCorp may run off with the idea, and submit a solution to the IDEAS archive or to the Stata Technical Bulletin.

Features to commands

  • promise: ologit, or option
  • promise: stepwise, detail option
  • promise: sort [varlist], stable
  • consideration: add matrix option to correlate
  • consideration: cond(exp,true,false,missing)
  • promise: outfile, nocomma
  • promise: tabulate with expected values (jw: and cochrans' rule)
  • promise: insheet, allstring
  • frown: infile should allow > 2047 variables if some are skipped
  • consideration: merge keyvar using fname, bringin(varlist)
  • ?: border on boxplot
  • promise?: reshape should preserve variable labels (and value labels?)
  • ?: label value varlist lname
  • promise/bug: by : graph should set more conditions

New functions/commands

  • promise: new atan
  • consideration: case / switch statements
  • ?: functions
  • promise: svycox
  • frown: all subset regression
  • frown (3x): debugger
  • consideration: save/restore the workspace
  • consideration: savefile option with estimates
  • consideration: save matrices in with data
  • promise: rowsort varlist [if] [in]
  • ?: converter of enhanced-smcl output to RTF format
  • consideration: constraints as a matrix
  • frown: syntax for sort descending
  • consideration: command to display collinearities in an interpretable format
  • ?: combine returned results / statistics into one matrix
  • consideration: read/write ascii files
  • frown: sparse matrices
  • frown: missing values in matrices
  • operators

Interface issues

  • promise: name of working directory in toolbar; for editor: filename
  • promise: improve layout of variables window
  • promise: increase size of result buffer
  • consideration: In review window: (1) no jumping, and (2) execute a block of commands
  • consideration: audit trail; alternative: user settable number of backups (named fname.dta.bak#)
  • Bill's presentation of his plans for an net-tee-d version of Stata was well-received, especially for tech-support and teaching, though worry was expressed on effects on demand for help.

Misc

  • ?: bs-like commands for em / mi / permutations (Goldstein)
  • well... : presentation-style, interactive, and 3d graphics
  • promise: big-stata, no limits mentioned
  • consideration: more netcourses, for example on panel methods (jw: and survey statistics)

Personally I was pleased to see that many requests were fairly realistic. This is reflected by Bill Gould's many positive requests. Of course, it would be nice that Stata were able to do mixed models, structural equations, bayesian statistics, and boil coffee all at the same time ...

Jeroen

Bookmark and Share 
Meetings
Stata Conference
User Group meetings
Proceedings
Like us on Facebook Follow us on Twitter Follow us on LinkedIn Google+ Watch us on YouTube
Follow us
© Copyright 1996–2013 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index   |   View mobile site