Unable to use estadd with listcoef after logit and mlogit

[Avery Cohn]

I've searched the web, Statalist, and helpfiles, but I'm stumped with the following problem. I'm unable to use the estadd command to append listcoef postestimation results to e() following logit and following mlogit. In both cases, I get the error "option matrix not allowed r(198); ." A check of ado dir reveals that I have spost13_ado (updated 2/10/15) and estout installed. I'm running Stata/MP 13.1. Any idea what the problem might be?

[Richard Williams]

Avery, can you provide code and output, preferably using a replicable example? (And also using code tags). It is always possible you have a syntax error we can't see. But also, estadd was written before spost13 came out and may be having problems with the latest version of listcoef.

[Richard Williams]

The estadd help explicitly refers to spost9. Looks to me like the spost9 listcoef had a matrix option and the spost13 version does not. I'm guessing that stuff that used to get captured by the matrix option is now stored in r(table) -- but estadd doesn't know about that.

[Avery Cohn]

Richard, Thank you. You're second posted allowed me to solve my problem. The package spost13 does now store what used to be the matrix option in r(table). For my purposes, what (according to the helpfiles) could be accomplished in spost9 with the code:

logit y x
estadd listcoef [varlist] [, listcoef_options ]

Can now be accomplished with the code:

logit y x
listcoef [varlist] [, listcoef_options] 
estadd r(table)

I guess it would be good for this to be documented for future users. I'm not sure if there is any difference in potential functionality.

[Richard Williams]

That is the risk with relying on other people's routines. Few, if any, user-written programs support version control the way official Stata commands do. So there is always the risk that a program you are using will change and break yours. In this case it would have been hard for estadd to do anything else, but in other situations people may be wise to borrow code (with thanks and attribution) and write their own programs.

[Avery Cohn]

Agreed. Thanks again, Richard. This thread can be closed now. I searched to see if I am able to do that myself, but I couldn't find instructions.