Rename estimated coefficients

[Berit Alsson]

Hi,

Im trying to make a table with three different models but the estimated coefficient in CSDID is not named the same as in TWFE and in the Stacked model so my tables looks like this:

	TWFE	Stacked reg	CSDID
Post	123	123
ATT			123

How can I rename "ATT" to "post" for my estimated coefficients? I have tried estout using CSDID, rename(ATT post) and esttab, rename(ATT post) without success.



I'm currently using est save to save my estimates then est store and est out(seen below) to use them in my table. Here is a sample of the beginning of my tables:


estout tsno5 ssno5 csno5 tsc5 ssc5 csc5 using "twfe.tex", replace ///
style(tex) ///
cells(b(fmt(a3) star) se(fmt(a3) par)) ///
starl( * 0.10 ** 0.05 *** 0.010) ///
keep(post_ev ATT) ///
varlabels(post_ev "Post" ALL "Post") ///
mlabels("TWFE" "Stacked" "CS" "TWFE(no cov)" "Stacked(no cov)" "CS(no cov)" ) collabels(none) ///
substitute(_ \_ # \#) ///
prehead( \footnotesize \centering `"\def\sym#1{\ifmmode^{#1}\else\(^{#1}\)\fi}"' ///
"\begin{threeparttable} \begin{tabular}{l*{@E}{c}} \toprule \toprule" ///
"\caption{Estimates for all methods & region: all & km:5}" ///
"& \multicolumn{6}{c}{\textsc{A. Static model}} \") ///
posthead(\cmidrule(lr){2-7} \\) ///
prefoot() ///
postfoot()
.
.
.
.

Many thanks!

[George Ford]

In esttab, I think the name change option is coef(ATT post)

[Berit Alsson]

Hi George!

the command does change the name but only for display, it doesn't save the name. Maybe esttab is not the right function? but I cannot find any rename option for est save or est store?


Thank you

[George Ford]

open csdid.ado. look at line 740. Change that to "Post". Run it the ado. Then run the model to see if that does it.

[Andrew Musau]

Look at how one can rename a coefficient using erepost from SSC in https://www.statalist.org/forums/for...em-to-tex-file. Editing the ado-file will work, but it seems overkill.

[KOBAYASHI Ryuki]

I encountered the same issue with csdid—the aggregated coefficient is always named “ATT,” which prevented it from aligning with my TWFE results in the esttab table.
The following approach resolved it:

// Rename the ATT coefficient using nlcom
nlcom <desired_name> = _b[ATT]

// Store the renamed estimates
est store cs_result

By using nlcom to rename the coefficient and then storing that result, I could display the TWFE result and the csdid result on the same row.
I hope this helps anyone else facing the same problem.

[Andrew Musau]

A remark on the useful advice in #6:

// Rename the ATT coefficient using nlcom
nlcom <desired_name> = _b[ATT]

// Store the renamed estimates
est store cs_result

This syntax does not appear to be correct for nlcom. From the documentation, the syntax should be:

Nonlinear combination of parameters -- one expression

nlcom [name:]exp [, options]

So one would need:

nlcom <desired_name> : _b[ATT]

Example:

sysuse auto, clear
regress mpg weight displacement
*SUGGESTED
nlcom myname= _b[weight]

*CORRECTED
nlcom myname: _b[weight]

Res.:

. regress mpg weight displacement

      Source |       SS           df       MS      Number of obs   =        74
-------------+----------------------------------   F(2, 71)        =     66.79
       Model |  1595.40969         2  797.704846   Prob > F        =    0.0000
    Residual |  848.049768        71  11.9443629   R-squared       =    0.6529
-------------+----------------------------------   Adj R-squared   =    0.6432
       Total |  2443.45946        73  33.4720474   Root MSE        =    3.4561

------------------------------------------------------------------------------
         mpg | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
      weight |  -.0065671   .0011662    -5.63   0.000    -.0088925   -.0042417
displacement |   .0052808   .0098696     0.54   0.594    -.0143986    .0249602
       _cons |   40.08452    2.02011    19.84   0.000     36.05654    44.11251
------------------------------------------------------------------------------

. 
. *SUGGESTED

. 
. nlcom myname= _b[weight]

"=" not allowed in expression
r(198);

. 
. 
. 
. *CORRECTED

. 
. nlcom myname: _b[weight]

      myname: _b[weight]

------------------------------------------------------------------------------
         mpg | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
      myname |  -.0065671   .0011662    -5.63   0.000    -.0088529   -.0042813
------------------------------------------------------------------------------

.

[KOBAYASHI Ryuki]

Dear Mr. Musau,

Thank you very much for pointing out the syntax mistake. I had a typo. The code you provided is correct.

[Parul Gupta]

Masau: could you suggest how to use this after csdid? If I use
nlcom <desired_name> : _b[ATT] after csdid, I get the error "ATT not found".

[Andrew Musau]

You need to post the estat results first.

use https://friosavila.github.io/playingwithstata/drdid/mpdta.dta, clear
csdid lemp lpop , ivar(countyreal) time(year) gvar(first_treat) method(dripw)
estat simple, post
nlcom Post: _b[ATT], post
esttab

Res.:

. 
. estat simple, post
Average Treatment Effect on Treated
------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
         ATT |  -.0417518   .0115028    -3.63   0.000    -.0642969   -.0192066
------------------------------------------------------------------------------

. 
. nlcom Post: _b[ATT], post

        Post: _b[ATT]

------------------------------------------------------------------------------
             | Coefficient  Std. err.      z    P>|z|     [95% conf. interval]
-------------+----------------------------------------------------------------
        Post |  -.0417518   .0115028    -3.63   0.000    -.0642969   -.0192066
------------------------------------------------------------------------------

. 
. esttab

----------------------------
                      (1)   
                            
----------------------------
Post              -0.0418***
                  (-3.63)   
----------------------------
N                           
----------------------------
t statistics in parentheses
* p<0.05, ** p<0.01, *** p<0.001

.