How to create a table with dummay as columns and variables as rows (with formatting)

[Yifan Lu]

Hi all,

I recently run into this issue of organizing my summary statistics into a table but distinguishing the dummy variable. per se my dummy takes value of 1 if true and 0 otherwise. I have several variables which I want to look at the average and robust standard error for. And I want to organize my table into this:

	dummy=1	dummy=0
Var 1	mean(Var1) if dummy==1, (RSE)	mean(Var1) if dummy==0, (RSE)
Var 2	mean(Var2) if dummy==1, (RSE)	mean(Var2) if dummy==0, (RSE)

There are of course more than one variable, but I'm confused with how to organize my table into this, especially that Var1, Var2,..., dummy are all variables. I also hope to export this table into a latex form which is why I was trying to use "estout" and "esttab" commands but just couldn't get it into this form.

Thanks in advance!

[Maarten Buis]

sysuse auto, clear
table () (foreign) ,        ///
    stat(mean mpg price)    ///
    stat(semean mpg price)  ///
    nototal nformat(%9.3g)  ///
    sformat("(%s)" semean)
    
collect layout (var#result) (foreign)

After that you can use collect export to save that as a LaTeX table