Statalist


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

st: conformability error in micombine


From   "Kelly Richardson" <[email protected]>
To   [email protected]
Subject   st: conformability error in micombine
Date   Tue, 13 Nov 2007 11:11:01 -0600

Dear Statalisters,

I am running the following model and I get a conformability error:

micombine mlogit chwt_cat new_retire obese over_weight under_weight
cesd_rd funstat medcond1_ medcond2_ married age_con high_school_grad
some_college college_grad post_college black latino other_race smoke
alc_high alc_med time1 time2 time3 alive vigex if previous_retire==0 &
currwork==1 & other_collar==1, robust cluster(unique_id) obsid(_i)
eform(odds ratio)


When I run the nearly identical model (with the exception of selecting
blue collar workers rather than other collar workers) I have no
problem.

micombine mlogit chwt_cat new_retire obese over_weight under_weight
cesd_rd funstat medcond1_ medcond2_ married age_con high_school_grad
some_college college_grad post_college black latino other_race smoke
alc_high alc_med time1 time2 time3 alive vigex if previous_retire==0 &
currwork==1 & blue_collar==1, robust cluster(unique_id) obsid(_i)
eform(odds ratio)


I looked at the archives and noticed a lot of people were having
trouble with missing data. I looked to see if any variables had
completely missing data for the subset and found several of the
education variables were blank. So, I ran the model again without the
education variables and I got the same error message.

micombine mlogit chwt_cat new_retire obese over_weight under_weight
cesd_rd funstat medcond1_ medcond2_ married age_con high_school_grad
black latino other_race smoke alc_high alc_med time1 time2 time3 alive
vigex if previous_retire==0 & currwork==1 & other_collar==1, robust
cluster(unique_id) obsid(_i) eform(odds ratio)

I added set trace in the .do file and it stopped here:

- if "`genxb'"!="" {
  = if ""!="" {
    predict `xbtmp' if `touse'==1 & `J'==`i', xb
    replace `xb'=`xbtmp' if `touse'==1 & `J'==`i'
    drop `xbtmp'
    }
  - matrix `Q`i''=e(b)
  = matrix __000009=e(b)
  - if `i'==1 {
  = if 2==1 {
    matrix `Q'=e(b)
    matrix `W'=e(V)
    }
  - else {
  - matrix `Q'=`Q'+e(b)
  = matrix __000004=__000004+e(b)
conformability error
    matrix `W'=`W'+e(V)
    }


I don't know if that gives you enough information to answer the
question or not. I'm certainly willing to send the entire log to
someone if they need it. Any suggestions you all might have would be
enthusiastically welcomed.

Thanks,
Kelly
*
*   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/



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