Change the label/ variable name only in regression output

[Matthew Berg]

Dear Statalist,

Thank you so much for all your previous help in my projects.

I would like to now change the labels of the variables that appear in my regression output.

                                 (Std. err. adjusted for 175 clusters in CASE)
------------------------------------------------------------------------------
             |               Robust
       action| Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         T_C |
    3 years  |   .0014682   .0586037     0.03   0.980    -.1141975    .1171339
             |
 return_numb |
          2  |   .0415479   .0820218     0.51   0.613    -.1203379    .2034337
          3  |  -.0802777   .0850249    -0.94   0.346    -.2480905    .0875352
          4  |   .0261982   .0848565     0.31   0.758    -.1412823    .1936786
          5  |   .0324653    .083936     0.39   0.699    -.1331984    .1981291
          6  |  -.0118075   .0792168    -0.15   0.882    -.1681571     .144542
          7  |    .018131   .0873436     0.21   0.836    -.1542583    .1905204
          8  |  -.0442845   .0773675    -0.57   0.568    -.1969842    .1084151
          9  |  -.0373382   .0917166    -0.41   0.684    -.2183585    .1436821
         10  |   .1418209   .0865785     1.64   0.103    -.0290584    .3127002
             |
        risk |
          2  |   .0302998   .0427611     0.71   0.480    -.0540975     .114697
          3  |  -.0454869   .0468423    -0.97   0.333    -.1379392    .0469654
          4  |  -.0730557   .0513766    -1.42   0.157    -.1744573    .0283459
             |
       round |
          2  |   .0446484   .0467053     0.96   0.340    -.0475335    .1368303
          3  |   .0342588   .0460236     0.74   0.458    -.0565775    .1250951
          4  |  -.0182757   .0466819    -0.39   0.696    -.1104114    .0738601

Particularly, I want to rename T_C "time frame" and return_numb "Return year" in this regression. Specifically, the 2 should say 2021, 3 2020 and so on until 10 -2013.

I tried

label variable return_numb "Return year"
label variable T_C "Time frame"

but when I reran the regression, I still had T_C and return_numb in the regressions.

Then I tried the method described here https://www.statalist.org/forums/for...abel-vairables but this suggested I rename the variable and this would also mean that in future regressions I need to type action i.Time frame instead of the short, preferred i.T_C as before.

Next, in the data edit mode in Stata, I labeled the variable and also included a label under data, but still without success.

YouTube videos suggested exporting the regressions to Excel or Latex and manually changing the variable names. Since I'm running quite a few regressions I would like to avoid this manual work.

Using https://www.statalist.org/forums/for...lable-from-ssc and SSC's elabel,I tried

 elabel define T_C "Time frame"

but received an error message that says "Time frame" invalid name
r(198);


I would be most grateful for any help or advice.

Kind regards,
Matt

[daniel klein]

regerss won't show variable labels; for that you need estimates table or some community-contributed command. regress will show value labels.

Consider:

. sysuse auto
(1978 automobile data)

. 
. label define rep78 ///
>     1 "one"        ///
>     2 "two"        ///
>     3 "three"      ///
>     4 "four"       ///
>     5 "five"

.     
. label values rep78 rep78

. 
. regress price mpg i.rep78

      Source |       SS           df       MS      Number of obs   =        69
-------------+----------------------------------   F(5, 63)        =      4.39
       Model |   149020603         5  29804120.7   Prob > F        =    0.0017
    Residual |   427776355        63  6790100.88   R-squared       =    0.2584
-------------+----------------------------------   Adj R-squared   =    0.1995
       Total |   576796959        68  8482308.22   Root MSE        =    2605.8

------------------------------------------------------------------------------
       price | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         mpg |  -280.2615   61.57666    -4.55   0.000    -403.3126   -157.2103
             |
       rep78 |
        two  |   877.6347   2063.285     0.43   0.672     -3245.51     5000.78
      three  |   1425.657   1905.438     0.75   0.457    -2382.057    5233.371
       four  |   1693.841   1942.669     0.87   0.387    -2188.274    5575.956
       five  |   3131.982   2041.049     1.53   0.130    -946.7282    7210.693
             |
       _cons |   10449.99   2251.041     4.64   0.000     5951.646    14948.34
------------------------------------------------------------------------------

. 
. estimates table , varlabel

---------------------------------------
                Variable |   Active    
-------------------------+-------------
           Mileage (mpg) | -280.26149  
                         |             
      Repair record 1978 |             
                    two  |  877.63472  
                  three  |   1425.657  
                   four  |   1693.841  
                   five  |  3131.9822  
                Constant |  10449.991  
---------------------------------------

To get something similar you could

label variable return_numb "Return year"
label variable T_C         "Time frame"

elabel define return_numb:return_numb_lbl ///
    (1/10) (=strofreal(2023-#))

where the elabel syntax is short for

label define return_numb_lbl 1 "2022" 2 "2021" ... 10 "2013"
label values return_numb return_numb_lbl

[Matthew Berg]

Dear Daniel,

Thank you very much for your reply. Following your suggestion, I implemented

label variable return_numb "Return year"
label variable T_C         "Time frame"

elabel define return_numb:return_numb_lbl ///
    (1/10) (=strofreal(2022-#))

and it worked in the regress regression output in Stata

                                 (Std. err. adjusted for 175 clusters in CASE)
------------------------------------------------------------------------------
             |               Robust
       action | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         T_C |
    3 years  |   .0014682   .0586037     0.03   0.980    -.1141975    .1171339
             |
 return_numb |
       2021  |   .0415479   .0820218     0.51   0.613    -.1203379    .2034337
       2020  |  -.0802777   .0850249    -0.94   0.346    -.2480905    .0875352
       2019  |   .0261982   .0848565     0.31   0.758    -.1412823    .1936786
       2018  |   .0324653    .083936     0.39   0.699    -.1331984    .1981291
       2017  |  -.0118075   .0792168    -0.15   0.882    -.1681571     .144542
       2016  |    .018131   .0873436     0.21   0.836    -.1542583    .1905204
       2015  |  -.0442845   .0773675    -0.57   0.568    -.1969842    .1084151
       2014  |  -.0373382   .0917166    -0.41   0.684    -.2183585    .1436821
       2013  |   .1418209   .0865785     1.64   0.103    -.0290584    .3127002
             |
        risk |
          2  |   .0302998   .0427611     0.71   0.480    -.0540975     .114697
          3  |  -.0454869   .0468423    -0.97   0.333    -.1379392    .0469654
          4  |  -.0730557   .0513766    -1.42   0.157    -.1744573    .0283459
             |
       round |
          2  |   .0446484   .0467053     0.96   0.340    -.0475335    .1368303
          3  |   .0342588   .0460236     0.74   0.458    -.0565775    .1250951
          4  |  -.0182757   .0466819    -0.39   0.696    -.1104114    .0738601

However not in my exported file, using outreg2 from ssc

outreg2 using myfile.doc, replace ctitle(Basic) dec(3)

I obtained the following when I clicked on myfile.doc:
As you can see the years are no longer displayed but instead it's gone back to return_numb etc.

	(1)
VARIABLES	Basic
1.T_C	0.001
	(0.059)
2.return_numb	0.042
	(0.082)
3.return_numb	-0.080
	(0.085)
4.return_numb	0.026
	(0.085)
5.return_numb	0.032
	(0.084)
6.return_numb	-0.012
	(0.079)
7.return_numb	0.018
	(0.087)
8.return_numb	-0.044
	(0.077)
9.return_numb	-0.037
	(0.092)
10.return_numb	0.142
	(0.087)
2.risk	0.030
	(0.043)
3.risk	-0.045
	(0.047)
4.risk	-0.073
	(0.051)
2.round	0.045
	(0.047)
3.round	0.034
	(0.046)
4.round	-0.018
	(0.047)

Constant	0.308***
	(0.087)
Observations	574
R-squared	0.026

Robust standard errors in parentheses
*** p<0.01, ** p<0.05, * p<0.1

I omitted parts of the regression output for better readability (which is why the constant, observations and r^2 don't line up anymore).

Further, I am using outreg2 (from SSC) because I have many factor variables and outreg2 suppresses the base dummies while estout does not anymore (nobaselevels option no longer available).

I would be most grateful for any help or comments.

Thank you very much.

[daniel klein]

I have not used outreg2 for about a decade or so. My guess is that it has an option (or more) to display variable and value labels that you need to specify. I cannot tell you what it is. Read

help outreg2

[Matthew Berg]

Thank you! It's the option label for anyone else wondering.