Esttab with tabstat by group: how to remove the statistic name?

[Simon Walker]

I am trying to produce a table of means by two grouping variables. I have found a sufficient solution, see MWE below. There is just the issue that the header includes the information "mean" in every column. How can I disable that?


MWE:

local vars turn length weight

sysuse auto.dta, replace
drop if rep78==.

estpost tabstat `vars' if foreign==1, by(rep78) statistics(mean) columns(statistics) nototal
eststo model1
estpost tabstat `vars' if foreign==0, by(rep78) statistics(mean) columns(statistics) nototal
eststo model2


#delimit ;
    esttab model1 model2,
        varwidth(30)
        label unstack nonumber noobs nonote compress
        cells(mean(label("") fmt(2)))
        mtitles("M1" "M2");
#delimit cr

[Andrew Musau]

estout is from SSC. Thanks for the reproducible example. You need:

#delimit ;
    esttab model1 model2, collab(none)
        varwidth(30)
        label unstack nonumber noobs nonote compress
        cells(mean(label("") fmt(2)))
        mtitles("M1" "M2");
#delimit cr