reg2docx -- Report regression results to formatted table in DOCX file

[Tony Li]

reg2docx is now available on SSC, again, thanks to Kit Baum

with this command, we can report regression results directly to docx
with this command, you do not need to bother to install and set for the wordconvert command

It is used after est store. Users can estimate different regression models. After that they can save the regression results with est store
command. Then, users can call reg2docx to design a formatted table output for all the saved models to a docx file. The docx file can be appended to
other docx files generated by putdocx, sum2docx, corr2docx, t2docx and etc, either using the append or replace option.

A demonstration is as following:


clear

set obs 1000

gen x1 = uniform()

gen x2 = uniform()

gen x3 = uniform()

gen x4 = uniform()

gen x5 = uniform()

gen x6 = uniform()

gen ind = mod(_n,10)

tab ind, gen(ind)

gen y = 0.4+.5*x1+.6*x2+.7*x3+.8*x4+rnormal()*3

replace y = y-.7*x5-.8*x6

forvalue i = 1(1)10 {

replace y = y+sqrt(`i')*ind`i'

}

reg y x1 x5 x6 ind2-ind10

est store m1

reg y x1 x2 x5 x6 ind2-ind10

est store m2

reg y x1 x2 x3 x5 x6 ind2-ind10

est store m3

reg y x1 x2 x3 x4 x5 x6 ind2-ind10

est store m4

reg2docx m1 m2 m3 m4 using d:/mytable2.docx, replace indicate("ind=ind*") drop(x2 x3) r2(%9.3f) ar2(%9.2f) order(x6 x5) b(%9.3f) t(%7.2f) ///
title(table2: OLS regression results) mtitles("model 1" "model 2" "" "model 4")

[Felix Kaysers]

Hi, thank you for this amazing package. It looks super promising. One thing that's keeping me from making the switch from esttab is the advanced ability for value labels and dropping base levels for categorical / dummy / binary variables. Is there any chance this will come to reg2docx or are these limitations of putdocx?