Install Micombine

[Tess Gecha]

Good morning,

How do I install the micombine command on Stata?

Best,
Tess

[Carlo Lazzaro]

Tess:
just type:

net sj 7-4 st0067_3

from within Stata and you'll be taken to the -micombine- community-contributed module helpfile.

[Tess Gecha]

Thank you for your response, Carlo.

When I type in the command, it gives me this:

-------------------------------------------------------------------------------------------------
package st0067_3 from http://www.stata-journal.com/software/sj7-4
-------------------------------------------------------------------------------------------------

TITLE
SJ7-4 st0067_3. Update: Multiple imputation of missing...

DESCRIPTION/AUTHOR(S)
Update: Multiple imputation of missing values

by Patrick Royston, MRC Clinical Trials Unit, London, UK
Support: [email protected]
After installation, type help ice, micombine, and
ice_reformat

INSTALLATION FILES (type net install st0067_3)
st0067_3/ice.ado
st0067_3/ice.hlp
st0067_3/ice_reformat.ado
st0067_3/ice_reformat.hlp
st0067_3/micombine.ado
st0067_3/micombine.hlp
st0067_3/uvis.ado
st0067_3/uvis.hlp
st0067_3/cmdchk.ado
st0067_3/nscore.ado
-------------------------------------------------------------------------------------------------

And when I try the micombine command again, it says "command micombine is unrecognized."

[Carlo Lazzaro]

Tess:
-search micombine-, spot the relevant entry and then install all the stuff.
Worked for me a handful of seconds ago.

[Øyvind Snilsberg]

or just type

net install st0067_3

[Tess Gecha]

Got it - thank you.

It successfully installed, but I am now running into another error message:

micombine xtgee logallalone dib_bpd nperiod nperiod2 sex,i(id) eform robust nolog
imputation identifier _mj not found

Does anyone know what this means?

[Rich Goldstein]

it sounds as though you either do not have multiply imputed data or you did not use the -ice- (or possibly -uvis-) command to impute the data; -ice- automatically sets up "counter" variables including _mj and -micombine- expects to see these; so, assuming you have already imputed the data, how did you impute the data?