[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: saving dprobit results for use in a table

From   Martin Weiss <[email protected]>
To   [email protected]
Subject   Re: st: saving dprobit results for use in a table
Date   Sat, 30 Aug 2008 19:30:19 +0200

Type -eret li- after -dporbit- to see that e(dfdx) contains the marginal effects while -eststo- still captures the coefficients of the -probit- regression. -eststo- can display both results simulataneously, though:

sysuse auto, clear
cap ssc inst estout
eststo clear
dprobit f w p dis
mat def A=e(dfdx)
estadd mat margeff=A
esttab, aux(margeff)


Quoting David Merriman <[email protected]>:

I would like to save dprobit coefficients to report in a table.

I tried using estimates store and then esttab [code is below] but
esstab returned the probit rather than the dprobit coefficients.

This question came up (on statalist) in 2004 and Buzz Burhans posted
some matrix code to do this but (A) the code is complicated and I
don't really understand it and (B) I am wondering if anything simpler
has developed in the meantime.

Code I used follows

#delimit cr
foreach var of varlist $border_tax_paid_vars  {
dprobit `var' $border_inc_distance_vars $land_use_vars
$B_demographic_vars  [pw=(1/os_obs_weight)] if place==2, r
estimates store E8`var'
adjust $land_use_vars $B_demographic_vars   if place==2, pr
#delimit ;

[code omitted]

esttab E3indiana_tax_paid E4indiana_tax_paid E3illinois_tax_paid
E3cook_tax_paid E4cook_tax_paid E3chicago_tax_paid E4chicago_tax_paid
using "C:\Documents and Settings\dmerrim\My Documents\rwj\summer
2008\stata\tables\some revised tables1.csv", noisily  replace
label star(* 0.10 ** 0.05 *** 0.01)
stats(N r2_p)b(%4.2f)varwidth(34)  modelwidth(7)
title("Estimates of marginal change in probability of each stamp"
"from a small change in each independent variable" "using data from 96
Chicago TAZ");
*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index