labmask (SSC) may not label 5030000000

[Marc Kaulisch]

Dear Nick Cox , I encounter an unintuitive error which -labmask- (SSC). Do you have an idea what happens there?

-labmask- works well for a few hundret values but it stops at one case.

My code is as follows:

labmask aobj_vorg, val(projekttitel1) lblname(l_aobj)

The unique number of aobj_vorg is 1.054 and the unique number of projekttitel1 is 1.039.

Tracing the error shows the following sequence at the end:

 su `example' if `group' == `i', meanonly
= su __000005 if __000004 == 1053, meanonly
- local label = `values'[`r(min)']
= local label = projekttitel1[1082]
- local value = `varlist'[`r(min)']
= local value = aobj_vorg[1082]
- label def `lblname' `value' `"`label'"', modify
= label def l_aobj 5003000000 `"CEDUS"', modify
  --------------------------------------------------------------------------------------------------------------------------------------------------------------- begin label ---
  - version 10.0
  - local vv : display "version " string(_caller()) ", missing:"
  - gettoken val : 0
  - if (strpos("`val'", "val") > 0 ) {
  = if (strpos("def", "val") > 0 ) {
    gettoken val 0 : 0
    syntax anything [, nofix]
    if "`fix'" != "" {
    local fix ", nofix"
    }
    gettoken var rest : anything
    while `"`rest'"' != "" {
    gettoken lab rest : rest
    local label "`lab'"
    }
    local vlist : list anything - lab
    if "`lab'" == "." {
    local lab ""
    }
    foreach var of varlist `vlist' {
    `vv' _label `val' `var' `lab' `fix'
    }
    }
  - else {
  - `vv' _label `macval(0)'
  = version 7, missing: _label def l_aobj 5003000000 `"CEDUS"', modify
may not label 5003000000
    }

[daniel klein]

This isn't specific to labmask. The largest positive integer that you can attach a label to is 2,147,483,621, which is - for some reason - one more than the documented largest value of a type long variable.

[Nick Cox]

Thanks for Marc Kaulisch for the question and daniel klein for the answer. I agree with Daniel, and indeed the trace in #1 shows that label is objecting to the request.

FYI, labmask if you want to reference it was published via the Stata Journal

. search labmask, sj

Search of official help files, FAQs, Examples, and Stata Journals

SJ-8-2  gr0034  . . . . . . . . . .  Speaking Stata: Between tables and graphs
        (help labmask, seqvar if installed) . . . . . . . . . . . .  N. J. Cox
        Q2/08   SJ 8(2):269--289
        outlines techniques for producing table-like graphs

although the code on SSC is idnntical, namely

*! NJC 1.0.0 20 August 2002

[Marc Kaulisch]

Thanks to both of you, Nick Cox and daniel klein. I was not aware of this. So I have to find a good way to reduce the 10digit numbers to nine digits without risking duplicate entries...

[Mike Lacy]

The built-in -egen- function -group()- will create a new variable that is a many to 1 mapping of your variable -aobj_vorg-. There are other ways to get from 10 digits to something smaller and non-duplicating, but that's probably the easiest. The new variable won't necessarily be 9 digits long; it will range from 1 to however big is needed. If that is OK, and presuming that the actual values of -aobj_vorg- aren't of intrinsic interest, the following might work for you.

egen shorter = group(aobj_vorg)
labmask shorter, val(projekttitel1) lblname(l_aobj)

[Marc Kaulisch]

The built-in -egen- function -group()- will create a new variable that is a many to 1 mapping of your variable -aobj_vorg-. There are other ways to get from 10 digits to something smaller and non-duplicating, but that's probably the easiest. The new variable won't necessarily be 9 digits long; it will range from 1 to however big is needed. If that is OK, and presuming that the actual values of -aobj_vorg- aren't of intrinsic interest, the following might work for you.

egen shorter = group(aobj_vorg)
labmask shorter, val(projekttitel1) lblname(l_aobj)

Hi Mike Lacy , thanks a lot for this excellent idea. To some degree values of aobj_vorg have an intrinsic interest but labels are good enough to counter the downside.