Label define where labels differ for same values of a variable by group

[Lars Pete]

Dear all,

I want rel_year_nonneg to have labels as the values of display_rel_year_nonneg, which labmask is unable to do.

Basically, I want labels of rel_year_nonng to differ by cohort.

I can't use "if" condition in label define.

So, I create the a new variable as display_rel_year_nonneg as follows. However, labmask doesn't work in this case since display_rel_year_nonneg not constant within groups of rel_year_nonneg.

* Recode rel_year to be non-negative
gen rel_year_nonneg = .
replace rel_year_nonneg = rel_year + 2 if cohort == 2012
replace rel_year_nonneg = rel_year + 3 if cohort == 2013


* Create a helper variable for correct labeling in graphs and tables
gen display_rel_year_nonneg = ""
replace display_rel_year_nonneg = "-2" if rel_year_nonneg == 0 & cohort == 2012
replace display_rel_year_nonneg = "-1" if rel_year_nonneg == 1 & cohort == 2012
replace display_rel_year_nonneg = "0" if rel_year_nonneg == 2 & cohort == 2012
replace display_rel_year_nonneg = "1" if rel_year_nonneg == 3 & cohort == 2012
replace display_rel_year_nonneg = "2" if rel_year_nonneg == 4 & cohort == 2012
replace display_rel_year_nonneg = "3" if rel_year_nonneg == 5 & cohort == 2012


replace display_rel_year_nonneg = "-3" if rel_year_nonneg == 0 & cohort == 2013
replace display_rel_year_nonneg = "-2" if rel_year_nonneg == 1 & cohort == 2013
replace display_rel_year_nonneg = "-1" if rel_year_nonneg == 2 & cohort == 2013
replace display_rel_year_nonneg = "0" if rel_year_nonneg == 3 & cohort == 2013
replace display_rel_year_nonneg = "1" if rel_year_nonneg == 4 & cohort == 2013
replace display_rel_year_nonneg = "2" if rel_year_nonneg == 5 & cohort == 2013


. labmask rel_year_nonneg, values(display_rel_year_nonneg)
display_rel_year_nonneg not constant within groups of rel_year_nonneg
r(198);

end of do-file

I appreciate your help. Thanks.

[daniel klein]

labmask (probably from SJ, perhaps from SSC) imposes no restriction beyond those of label define and the general concept of value labels. You cannot map the same integer to different text (strings). You need a fundamentally different approach. I would set up a reproducible example demonstrating how you create the graph and then explain what you want to change about it.

[Nick Cox]

I write as the author of labmask. As daniel klein implies, it can be found to be documented at

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

and also on SSC as part of the labutil package. It is required for some other commands to work. I would prefer that people give the Stata Journal reference.

So far, that is just a small question of how to explain community-contributed commands you refer to, following our longstanding request at https://www.statalist.org/forums/help#stata 12.1

Simply and fundamentally, I can only echo Daniel's explanation: value labels correspond one to one with values of a variable once a set of value labels has been defined and associated with a variable with integer values. Nothing that labmask or any other command can do can in any sense undermine or work around that definition. That is not a matter of labmask "not working" but a matter of it following the rules.

What you can do is create different variables using say separate and then define and assign different value labels for each variable. But again with a data example and more explanation, a simpler way of doing what you want may became clear.

[Nick Cox]

Looking at your code again and trying to get some sense of what you're doing.

You have cohorts for 2012 and 2013 and perhaps other years. So variables that code time before or after some origin could be got more directly by something like

* Recode rel_year to be non-negative
gen rel_year_nonneg = rel_year + cond(cohort == 2012, 2, cond(cohort == 2013, 3, .))  

* Create a helper variable for correct labeling in graphs and tables
gen display_rel_year_nonneg = . 
replace display_rel_year_nonneg = rel_year_nonneg - 2 if cohort == 2012
replace display_rel_year_nonneg = rel_year_nonneg - 3 if cohort == 2013

For the second block of code, I can't imagine any advantage in creating a string variable, even if you want value labels as the next step.

[Lars Pete]

Hi Daniel and Nick for your helpful responses.

Nick, you are right. The sample is from 2010 to 2019 with the following cohorts:

. fre cohort

cohort
-----------------------------------------------------------------------
                          |      Freq.    Percent      Valid       Cum.
--------------------------+--------------------------------------------
Valid   0    Never treted |      11250      49.33      49.33      49.33
        2012              |       1210       5.31       5.31      54.64
        2013              |       1310       5.74       5.74      60.38
        2014              |       1710       7.50       7.50      67.88
        2015              |       2040       8.95       8.95      76.83
        2016              |       3224      14.14      14.14      90.97
        2017              |       2060       9.03       9.03     100.00
        Total             |      22804     100.00     100.00           
-----------------------------------------------------------------------

This is an example:

* Example generated by -dataex-. For more info, type help dataex
clear
input float(cohort year rel_year) long countyfips
2016 2012 -4 2290
2016 2013 -3 2290
2016 2014 -2 2290
2016 2015 -1 2290
2016 2016  0 2290
2016 2017  1 2290
2016 2018  2 2290
2016 2019  3 2290
2013 2010 -3 4001
2013 2011 -2 4001
2013 2012 -1 4001
2013 2013  0 4001
2013 2014  1 4001
2013 2015  2 4001
2013 2016  3 4001
2013 2017  4 4001
2013 2018  5 4001
2013 2019  6 4001
2013 2010 -3 4003
2013 2011 -2 4003
2013 2012 -1 4003
2013 2013  0 4003
2013 2014  1 4003
2013 2015  2 4003
2013 2016  3 4003
2013 2017  4 4003
2013 2018  5 4003
2013 2019  6 4003
2013 2010 -3 4005
2013 2011 -2 4005
2013 2012 -1 4005
2013 2013  0 4005
2013 2014  1 4005
2013 2015  2 4005
2013 2016  3 4005
2013 2017  4 4005
2013 2018  5 4005
2013 2019  6 4005
2013 2010 -3 4007
2013 2011 -2 4007
2013 2012 -1 4007
2013 2013  0 4007
2013 2014  1 4007
2013 2015  2 4007
2013 2016  3 4007
2013 2017  4 4007
2013 2018  5 4007
2013 2019  6 4007
2013 2010 -3 4009
2013 2011 -2 4009
2013 2012 -1 4009
2013 2013  0 4009
2013 2014  1 4009
2013 2015  2 4009
2013 2016  3 4009
2013 2017  4 4009
2013 2018  5 4009
2013 2019  6 4009
2013 2010 -3 4011
2013 2011 -2 4011
2013 2012 -1 4011
end
label values cohort cohort

Are you suggesting that instead of labelling rel_year_nonneg according to different cohorts, use the rel_year_nonneg as it is and then use the variable directly in the graph by doing something like:

levelsof display_rel_year_nonneg, local(labels)
local xlabels
foreach level of local labels {
    local xlabels `xlabels' `level' "`level'"
}

And then in coefplot, using:

   xlabel(`xlabels', angle(0)) ///

?

Thank you.

[Nick Cox]

I am not suggesting anything yet, as I don't understand what you want to do.

First off, I repeat a key detail from earlier in the thread. You should explain community-contributed commands you refer to, following our longstanding request at https://www.statalist.org/forums/help#stata 12.1

Here fre and coefplot are community-contributed.

You evidently want to use coefplot for some purpose, but I can't follow what that is.

[Lars Pete]

Nick Cox Hi Nick, sorry for failing to acknowledge the use of community contributed commands earlier.

As per the request request at https://www.statalist.org/forums/help#stata 12.1, I am using using communty contributed command- coefpplot- to plot coefficients of interactions of treatment with rel_year to get event study for each treatment type. For estimaiton, I use another community contributed command: reghdfe. Earlier, I also used: fre - a community contributed command to show my cohorts. I am using Stata 18.

I want to do something like:

*******
eststo a1: reghdfe medcosts paypar#i.rel_year_nonneg pfdp#i.rel_year_nonneg pfdc#i.rel_year_nonneg ppdp#i.rel_year_nonneg ppdc#i.rel_year_nonneg pcdp#i.rel_year_nonneg pcdc#i.rel_year_nonneg pfonly#i.rel_year_nonneg pponly#i.rel_year_nonneg pconly#i.rel_year_nonneg dponly#i.rel_year_nonneg dconly#i.rel_year_nonneg i.rel_year_nonneg pfdp pfdc ppdp ppdc pcdp pcdc pfonly pponly pconly dponly dconly bbdT1std atet_imlcc logpop logMedianIncome logstramccost logpercentpov logUnemploymentRate16 logW, absorb(countyfips year) vce(cluster statefips)

local variables paypar pfdc ppdp ppdc pcdp pponly pconly dconly
local glist ""

foreach var in `variables' {

    local varlabel : variable label `var'
    di "`var': `varlabel'"

    local newvarlabel = subinstr("`varlabel'", "(sum) ", "", .)

    local `var' 1.`var'#0.rel_year_nonneg ///
                  1.`var'#1.rel_year_nonneg ///
                  1.`var'#2.rel_year_nonneg ///
                  1.`var'#3.rel_year_nonneg ///
                  1.`var'#4.rel_year_nonneg ///
                  1.`var'#5.rel_year_nonneg ///
                  1.`var'#6.rel_year_nonneg ///
                  1.`var'#7.rel_year_nonneg ///
                  1.`var'#8.rel_year_nonneg ///
                  1.`var'#9.rel_year_nonneg
    coefplot a1, keep(``var'') yline(0) ciopts(recast(rcap)) citop vertical mcolor(maroon) ///
    title("`newvarlabel'", size(medium)) ///
    ylabel(, angle(0)) ///
    xlabel(`xlabels', angle(0)) ///
    xtitle("Relative Years", size(medium))

    graph save `var'_graph.gph, replace

    local glist "`glist' `var'_graph.gph"
}

graph combine `glist', xsize(8) ysize(5)

The aim is to plot event study for each treatment type. paypar is main treatment indicator.

 
Treat type///////////////Cohort
Never treated
2012
2013
2014
2015
2017
2017

Never Treated
1
0
0
0
0
0
0

pfdc
0
0
1
0
0
1
0

ppdp
0
0
0
0
0
1
0

ppdc
0
0
1
0
0
1
0

pcdp
0
0
0
0
0
0
1

pp only
0
0
0
0
1
0
1

pc only
0
0
0
0
1
0
0

cc only
0
0
1
1
1
1
0

paypar
0
1
1
1
1
1
1

[Nick Cox]

Thanks for the extra detail, but sorry, I can't add helpfully. Your code doesn't provide a self-contained reproducible example, I am not familiar with key commands that you use, and I can't follow what you want. It seems quite likely that you have a soluble problem, but I can't work out what it is, and so leave this thread open for others (as it always was).

[Lars Pete]

Hi Nick Cox,

Thanks for your input. Please allow me to try to provide a self-contained reproducible example, for you as well as others:

I will exclude the additional controls for simplicity.

This is the specification that I am trying to run. Using community contributed command ``reghdfe"

xtset countyfips year  
 reghdfe thpv atet_paypar atet_pfdp atet_pfdc atet_ppdp atet_ppdc atet_pcdp atet_pcdc atet_pfonly atet_pponly atet_pconly atet_dponly atet_dconly pfdp pfdc ppdp ppdc pcdp pcdc pfonly pponly pconly dponly dconly bbdT1std logpop, a(countyfips year) vce(cluster statefips)

which give me ATT for each of the treatment types: paypar pfdc ppdp ppdc pcdp pponly pconly dconly.

atet_`var' is generated as:

foreach var of local treatments {
    gen atet_`var' = (year >= cohort & `var' == 1)
}

I wish to produce an event study for each treatment type for phsr.

So, I wanted to interact each treatment `var' with i.rel_year to get the event study plot:

reghdfe thpv paypar#i.rel_year pfdp#i.rel_year pfdc#i.rel_year ppdp#i.rel_year ppdc#i.rel_year pcdp#i.rel_year pcdc#i.rel_year pfonly#i.rel_year pponly#i.rel_year pconly#i.rel_year dponly#i.rel_year dconly#i.rel_year bbdT1std logpop, a(countyfips year) vce(cluster statefips)

I tried to create rel_year_nonneg since factor interactions don't allow negative values.

Here is an example of the dataset generated using dataex:

* Example generated by -dataex-. For more info, type help dataex
clear
input long thpv float(paypar pfdc ppdp ppdc pcdp pponly pconly dconly yrtreat cohort rel_year) double statefips long countyfips float(year atet_paypar atet_pfdc atet_ppdp atet_ppdc atet_pcdp atet_pponly atet_pconly atet_dconly bbdT1std logpop)
 776446 1 0 0 0 0 0 0 0 2016 2016 -1  9  9011 2015 0 0 0 0 0 0 0 0   .5968579 12.508478
 861169 1 0 0 0 0 0 0 0 2016 2016  0  9  9011 2016 1 0 0 0 0 0 0 0    .599049 12.505377
 819500 1 0 0 0 0 0 0 0 2016 2016  1  9  9011 2017 1 0 0 0 0 0 0 0   .6106076  12.50259
 830395 1 0 0 0 0 0 0 0 2016 2016  2  9  9011 2018 1 0 0 0 0 0 0 0   .6123917 12.494195
 519495 1 0 0 0 0 0 0 0 2016 2016  3  9  9011 2019 1 0 0 0 0 0 0 0   .8453003 12.488262
 233122 1 0 0 0 0 0 0 0 2016 2016 -6  9  9013 2010 0 0 0 0 0 0 0 0  .14903036 11.936172
 213128 1 0 0 0 0 0 0 0 2016 2016 -5  9  9013 2011 0 0 0 0 0 0 0 0  .14971557  11.93538
 246739 1 0 0 0 0 0 0 0 2016 2016 -4  9  9013 2012 0 0 0 0 0 0 0 0  .15154384 11.928228
 261170 1 0 0 0 0 0 0 0 2016 2016 -3  9  9013 2013 0 0 0 0 0 0 0 0  .15263605 11.927403
 257132 1 0 0 0 0 0 0 0 2016 2016 -2  9  9013 2014 0 0 0 0 0 0 0 0  .15331204  11.92659
 268154 1 0 0 0 0 0 0 0 2016 2016 -1  9  9013 2015 0 0 0 0 0 0 0 0  .27024326   11.9261
 213052 1 0 0 0 0 0 0 0 2016 2016  0  9  9013 2016 1 0 0 0 0 0 0 0  .27316388  11.92235
 199772 1 0 0 0 0 0 0 0 2016 2016  1  9  9013 2017 1 0 0 0 0 0 0 0   .2815094 11.928083
 199790 1 0 0 0 0 0 0 0 2016 2016  2  9  9013 2018 1 0 0 0 0 0 0 0   .2845484 11.924512
 416343 1 0 0 0 0 0 0 0 2016 2016  3  9  9013 2019 1 0 0 0 0 0 0 0  .28758734 11.923185
 421710 1 0 0 0 0 0 0 0 2016 2016 -6  9  9015 2010 0 0 0 0 0 0 0 0  .06751796  11.68206
 365177 1 0 0 0 0 0 0 0 2016 2016 -5  9  9015 2011 0 0 0 0 0 0 0 0  .06822194  11.68064
 398548 1 0 0 0 0 0 0 0 2016 2016 -4  9  9015 2012 0 0 0 0 0 0 0 0  .06820164 11.676998
 356734 1 0 0 0 0 0 0 0 2016 2016 -3  9  9015 2013 0 0 0 0 0 0 0 0  .06815519  11.67422
 291219 1 0 0 0 0 0 0 0 2016 2016 -2  9  9015 2014 0 0 0 0 0 0 0 0  .06814709 11.667825
 779445 1 0 0 0 0 0 0 0 2016 2016 -1  9  9015 2015 0 0 0 0 0 0 0 0  .06822619 11.665217
 793387 1 0 0 0 0 0 0 0 2016 2016  0  9  9015 2016 1 0 0 0 0 0 0 0 .068179734 11.661578
 751004 1 0 0 0 0 0 0 0 2016 2016  1  9  9015 2017 1 0 0 0 0 0 0 0  .07270537 11.664435
 277926 1 0 0 0 0 0 0 0 2016 2016  2  9  9015 2018 1 0 0 0 0 0 0 0  .07343097 11.670152
 267766 1 0 0 0 0 0 0 0 2016 2016  3  9  9015 2019 1 0 0 0 0 0 0 0  .07415653 11.668064
 442726 1 0 1 0 0 0 0 0 2016 2016 -6 10 10001 2010 0 0 0 0 0 0 0 0   .1979445 11.997263
 479896 1 0 1 0 0 0 0 0 2016 2016 -5 10 10001 2011 0 0 0 0 0 0 0 0  .20274556 12.011127
 566630 1 0 1 0 0 0 0 0 2016 2016 -4 10 10001 2012 0 0 0 0 0 0 0 0  .20651238 12.025282
 576451 1 0 1 0 0 0 0 0 2016 2016 -3 10 10001 2013 0 0 0 0 0 0 0 0    .211773 12.036322
 590834 1 0 1 0 0 0 0 0 2016 2016 -2 10 10001 2014 0 0 0 0 0 0 0 0   .2173106  12.05098
 620632 1 0 1 0 0 0 0 0 2016 2016 -1 10 10001 2015 0 0 0 0 0 0 0 0  .22382417  12.06019
 626205 1 0 1 0 0 0 0 0 2016 2016  0 10 10001 2016 1 0 1 0 0 0 0 0  .23101535  12.06765
 582625 1 0 1 0 0 0 0 0 2016 2016  1 10 10001 2017 1 0 1 0 0 0 0 0  .24679697 12.082905
 606637 1 0 1 0 0 0 0 0 2016 2016  2 10 10001 2018 1 0 1 0 0 0 0 0   .2582541 12.092618
 627597 1 0 1 0 0 0 0 0 2016 2016  3 10 10001 2019 1 0 1 0 0 0 0 0  .26971123  12.10506
1048604 1 0 1 0 0 0 0 0 2016 2016 -6 10 10003 2010 0 0 0 0 0 0 0 0  1.3984373 13.196504
1087700 1 0 1 0 0 0 0 0 2016 2016 -5 10 10003 2011 0 0 0 0 0 0 0 0  1.4008727  13.20275
1315554 1 0 1 0 0 0 0 0 2016 2016 -4 10 10003 2012 0 0 0 0 0 0 0 0  1.4032295 13.209614
1220167 1 0 1 0 0 0 0 0 2016 2016 -3 10 10003 2013 0 0 0 0 0 0 0 0   1.409748  13.21597
1257082 1 0 1 0 0 0 0 0 2016 2016 -2 10 10003 2014 0 0 0 0 0 0 0 0  1.4182273  13.22097
1339643 1 0 1 0 0 0 0 0 2016 2016 -1 10 10003 2015 0 0 0 0 0 0 0 0  1.4303068   13.2262
1388254 1 0 1 0 0 0 0 0 2016 2016  0 10 10003 2016 1 0 1 0 0 0 0 0  1.8787745  13.22948
1353205 1 0 1 0 0 0 0 0 2016 2016  1 10 10003 2017 1 0 1 0 0 0 0 0   1.882569  13.23532
1308996 1 0 1 0 0 0 0 0 2016 2016  2 10 10003 2018 1 0 1 0 0 0 0 0  1.8949186 13.234504
1290841 1 0 1 0 0 0 0 0 2016 2016  3 10 10003 2019 1 0 1 0 0 0 0 0   1.907268 13.233462
 539061 1 0 1 0 0 0 0 0 2016 2016 -6 10 10005 2010 0 0 0 0 0 0 0 0   .1916036 12.191695
 567746 1 0 1 0 0 0 0 0 2016 2016 -5 10 10005 2011 0 0 0 0 0 0 0 0  .19672586   12.2041
 611380 1 0 1 0 0 0 0 0 2016 2016 -4 10 10005 2012 0 0 0 0 0 0 0 0   .2019925  12.21888
 612391 1 0 1 0 0 0 0 0 2016 2016 -3 10 10005 2013 0 0 0 0 0 0 0 0   .2093621 12.233984
 585033 1 0 1 0 0 0 0 0 2016 2016 -2 10 10005 2014 0 0 0 0 0 0 0 0   .2187953 12.254653
 641854 1 0 1 0 0 0 0 0 2016 2016 -1 10 10005 2015 0 0 0 0 0 0 0 0   .4064615 12.276633
 540390 1 0 1 0 0 0 0 0 2016 2016  0 10 10005 2016 1 0 1 0 0 0 0 0   .4195234 12.299003
 519547 1 0 1 0 0 0 0 0 2016 2016  1 10 10005 2017 1 0 1 0 0 0 0 0  .44898435 12.325126
 551036 1 0 1 0 0 0 0 0 2016 2016  2 10 10005 2018 1 0 1 0 0 0 0 0   .4677485 12.342682
 592776 1 0 1 0 0 0 0 0 2016 2016  3 10 10005 2019 1 0 1 0 0 0 0 0   .4865126 12.364017
2924649 1 0 0 0 0 0 0 1 2013 2013 -3 11 11001 2010 0 0 0 0 0 0 0 0   1.938559 13.307552
3117944 1 0 0 0 0 0 0 1 2013 2013 -2 11 11001 2011 0 0 0 0 0 0 0 0   1.938494 13.332265
3118172 1 0 0 0 0 0 0 1 2013 2013 -1 11 11001 2012 0 0 0 0 0 0 0 0  1.9671515   13.3562
3323737 1 0 0 0 0 0 0 1 2013 2013  0 11 11001 2013 1 0 0 0 0 0 0 1  1.9908676  13.37789
3395032 1 0 0 0 0 0 0 1 2013 2013  1 11 11001 2014 1 0 0 0 0 0 0 1   2.009736 13.393097
3504760 1 0 0 0 0 0 0 1 2013 2013  2 11 11001 2015 1 0 0 0 0 0 0 1  2.0357604  13.41033
3203978 1 0 0 0 0 0 0 1 2013 2013  3 11 11001 2016 1 0 0 0 0 0 0 1  2.0677426 13.426203
3127250 1 0 0 0 0 0 0 1 2013 2013  4 11 11001 2017 1 0 0 0 0 0 0 1  2.0771666 13.450056
3418858 1 0 0 0 0 0 0 1 2013 2013  5 11 11001 2018 1 0 0 0 0 0 0 1   2.113357 13.462186
3615449 1 0 0 0 0 0 0 1 2013 2013  6 11 11001 2019 1 0 0 0 0 0 0 1  2.1495476  13.46699
1362687 0 0 0 0 0 0 0 0    .    0  . 12 12001 2010 0 0 0 0 0 0 0 0  .53723234 12.418503
1388962 0 0 0 0 0 0 0 0    .    0  . 12 12001 2011 0 0 0 0 0 0 0 0  .53861946 12.426938
1452457 0 0 0 0 0 0 0 0    .    0  . 12 12001 2012 0 0 0 0 0 0 0 0   .5400068 12.434387
1680573 0 0 0 0 0 0 0 0    .    0  . 12 12001 2013 0 0 0 0 0 0 0 0  .54255617 12.438557
1905455 0 0 0 0 0 0 0 0    .    0  . 12 12001 2014 0 0 0 0 0 0 0 0   .5465304 12.450942
2776831 0 0 0 0 0 0 0 0    .    0  . 12 12001 2015 0 0 0 0 0 0 0 0   .5506309 12.464958
2594392 0 0 0 0 0 0 0 0    .    0  . 12 12001 2016 0 0 0 0 0 0 0 0   .5578689 12.480677
2844522 0 0 0 0 0 0 0 0    .    0  . 12 12001 2017 0 0 0 0 0 0 0 0   .5691513  12.49476
1445744 0 0 0 0 0 0 0 0    .    0  . 12 12001 2018 0 0 0 0 0 0 0 0   .5843479 12.505962
2576393 0 0 0 0 0 0 0 0    .    0  . 12 12001 2019 0 0 0 0 0 0 0 0   .5995446 12.502626
  29755 0 0 0 0 0 0 0 0    .    0  . 12 12003 2010 0 0 0 0 0 0 0 0  -.2548468 10.207843
  29576 0 0 0 0 0 0 0 0    .    0  . 12 12003 2011 0 0 0 0 0 0 0 0 -.25497955 10.207584
  27521 0 0 0 0 0 0 0 0    .    0  . 12 12003 2012 0 0 0 0 0 0 0 0  -.2551129  10.20692
  29360 0 0 0 0 0 0 0 0    .    0  . 12 12003 2013 0 0 0 0 0 0 0 0 -.25524822 10.204925
  31475 0 0 0 0 0 0 0 0    .    0  . 12 12003 2014 0 0 0 0 0 0 0 0 -.25521025 10.210163
  32020 0 0 0 0 0 0 0 0    .    0  . 12 12003 2015 0 0 0 0 0 0 0 0 -.25519297 10.220376
  32283 0 0 0 0 0 0 0 0    .    0  . 12 12003 2016 0 0 0 0 0 0 0 0 -.23681693 10.238923
  31946 0 0 0 0 0 0 0 0    .    0  . 12 12003 2017 0 0 0 0 0 0 0 0 -.23440316 10.250016
  32602 0 0 0 0 0 0 0 0    .    0  . 12 12003 2018 0 0 0 0 0 0 0 0   -.233249  10.25256
  38478 0 0 0 0 0 0 0 0    .    0  . 12 12003 2019 0 0 0 0 0 0 0 0 -.23209484 10.282198
 354694 0 0 0 0 0 0 0 0    .    0  . 12 12005 2010 0 0 0 0 0 0 0 0  .12248213 12.036777
 312136 0 0 0 0 0 0 0 0    .    0  . 12 12005 2011 0 0 0 0 0 0 0 0  .12266454 12.039085
 300137 0 0 0 0 0 0 0 0    .    0  . 12 12005 2012 0 0 0 0 0 0 0 0   .1231056  12.05302
 299561 0 0 0 0 0 0 0 0    .    0  . 12 12005 2013 0 0 0 0 0 0 0 0  .12402184 12.070155
 342927 0 0 0 0 0 0 0 0    .    0  . 12 12005 2014 0 0 0 0 0 0 0 0  .27104497  12.09208
 347354 0 0 0 0 0 0 0 0    .    0  . 12 12005 2015 0 0 0 0 0 0 0 0  .27345794 12.109132
 275594 0 0 0 0 0 0 0 0    .    0  . 12 12005 2016 0 0 0 0 0 0 0 0   .2778051 12.120308
 268667 0 0 0 0 0 0 0 0    .    0  . 12 12005 2017 0 0 0 0 0 0 0 0   .2874613 12.120314
 267926 0 0 0 0 0 0 0 0    .    0  . 12 12005 2018 0 0 0 0 0 0 0 0  .29713088  12.12963
 265139 0 0 0 0 0 0 0 0    .    0  . 12 12005 2019 0 0 0 0 0 0 0 0   .3068004 12.070854
  58225 0 0 0 0 0 0 0 0    .    0  . 12 12007 2010 0 0 0 0 0 0 0 0 -.25037575  10.25836
  54847 0 0 0 0 0 0 0 0    .    0  . 12 12007 2011 0 0 0 0 0 0 0 0 -.25049835 10.256746
  49735 0 0 0 0 0 0 0 0    .    0  . 12 12007 2012 0 0 0 0 0 0 0 0 -.25068903 10.208174
  52632 0 0 0 0 0 0 0 0    .    0  . 12 12007 2013 0 0 0 0 0 0 0 0 -.25082248  10.19962
  56088 0 0 0 0 0 0 0 0    .    0  . 12 12007 2014 0 0 0 0 0 0 0 0  -.2509296 10.191107
end
label values paypar paypar
label def paypar 0 "Never Treated", modify
label def paypar 1 "Treated", modify
label values cohort cohort
label def cohort 0 "Never treted", modify

[/CODE]
[/CODE]