Thanks to Kit Baum, a new update to the margeff package is available through the SSC archive.
To install it, type
ssc install margeff , replace
or
net from http://www.bkae.hu/bartus/stata/
net install margeff, replace
The features of the recent version are:
The default behavior of margeff is again to estimate average partial changes.
Two new options, link(linkname) and level(#) are added. The former gives support for generalized linear models. Parallel to this, the old option eform is not documented, but remains supported (the eform option is interpreted as link(log)).
The partial effects after ologit and oprobit were not correct for intermediary outcomes. This is fixed. The standard errors are not identical with that calculated by mfx; the difference is probably due to the fact that mfx uses numerical instead of analytical calculations. I believe that the margeff results should be trusted.
The gologit2 command is fully supported, partial effects can be obtained whatever link function is specified during estimation of the gologit2 model.
The dummies(varlist) option has an effect only if partial effects at means or other fixed values are requested.
The margeff replay command failed to replay the results. This is fixed
The at(atlist) option produced an error message when atlist contained expressions. This is fixed.
Have fun, and plese keep reporting on bugs
Tamas
------------------------------------------------
Tamas Bartus, PhD
Associate Professor, Institute of Sociology and Social Policy
Corvinus University, Budapest
1093 Budapest, Fovam ter 8.
Phone: +36-1-482-5290 Fax: +36-1-482-5226
Homepage: www.uni-corvinus.hu/bartus
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
