Adding stats in regression output

[Jade Li]

Hello Everyone,

I would like to ask aquick question (I assume that a data example here is not necessary):

I am using this code:

regress depn_var x1 x2 x3 i.time_y, vce(cluster country)
outreg2 using results_1.xls, replace excel   title("Main Results") nocons dec(3) label

I need to add the mean of the dependent variable (similar to the number of obs and R-squared... at the end of the results (outcome) table). I did not find how to do it using outreg2.

Thank you
Li

[Nick Cox]

I don't know whether a data example is needed to answer your question, but it wouldn't do any harm.

I'll just note that outreg2 from SSC remains a popular download, but questions on it are rarely answered here. Its author left Statalist many years ago. I don't know if he replies to direct email. I don''t use it, so I will stop there.

[Jade Li]

Thank you Mr. Cox,

I am open to other Stata-commands or codes that would be also helpful. I mean a code that generates the Excel table and adds the mean of dependent var too.

Kind wishes
Jade

[Nick Cox]

Sorry, but like everyone else my expertise is bounded. Although re-presenting model fit output is an extraordinarily common activity, it's not what I ever do myself, except by writing my own code, which won't help here, not least because exporting to Excel is not what I want to do. So I don't have advice on translating to similar commands.

But -- much more importantly -- you may get advice to that effect from others and your title is suitably inclusive.

[Andrew Musau]

The procedure does not differ from adding any other statistic. As the statistic is not stored, place it in a local macro. Otherwise, take a look at

help outreg2

[Jade Li]

Dear Mr. Musau,

Thank you, that was very helpful!

I adjusted my code as follows:

  local avDV = e(ymean)
 regress depn_var x1 x2 x3 i.time_y, vce(cluster country) outreg2 using results_1.xls, replace excel   title("Main Results") addstat(avDV, `e(ymean)') nocons dec(3) label

However, the results did not show me the mean of the dependent variable. Is there something wrong with my code?

[Andrew Musau]

I do not know where you get "e(ymean)" from. I suggested that you store the statistic in a local.

qui sum depn_var if e(sample)
local ymean = r(mean)
...., addstat(`ymean')

[Jade Li]

Thank you! I found a similar one from Nick I guess:

  program estadd_mu, eclass
      syntax [ , prefix(name) * ]
      su `e(depvar)' if e(sample), meanonly
    ereturn scalar `prefix'mu = r(mean)
end

and it worked out well!
Kind wishes
Jade