{smcl}
{* 18sep2002}{...}
{.-}
help for {cmd: mimerge}{right:(SJ3-3: st0000)}
{.-}

{title:Merge multiple (imputed) datasets}

{p 8 27 2}
{cmd:mimerge} [{it:varlist}] {cmd:using} {it:filename-prefix}
[{cmd:,} {cmd:keep(}{it:varlist}{cmd:)} {cmdab:uniq:ue} {cmdab:uniqm:aster}
{cmdab:uniqu:sing} {cmdab:nol:abel} 
{cmdab:nok:eep} {cmd:_merge(}{it:varname}{cmd:)}]


{title:Description}

{p 4 4 2}
{cmd:mimerge} joins corresponding observations from each of the {help miset}
datasets with each of the {cmd:using} datasets into single observations.


{title:Options}

{p 4 8 2}
{cmd:keep(}{it:varlist}{cmd:)} specifies the variables to be kept from the
{cmd:using} data.  If {cmd:keep()} is not specified, all variables are kept.

{p 4 8 2}
{cmd:unique}, {cmd:uniqmaster}, and {cmd:uniqusing} specify that the
match variable(s) in a match merge uniquely identifies the observations.

{p 8 8 2}
{cmd:unique} specifies that the match variable(s) uniquely identifies
the observations in the master data and in the using data.  For most
match merges, you should specify {cmd:unique}.  {cmd:merge} does nothing
differently when you specify the option, unless the assumption that you
are making is false.  In that case, an error message is issued and the
data are not merged.

{p 8 8 2}
{cmd:uniqmaster} specifies that the match variable(s) uniquely identifies
the observations in memory, the master data, but not necessarily the
ones in the using data.

{p 8 8 2}
{cmd:uniqusing} specifies that the match variable(s) uniquely identifies
the observations in the using data, but not necessarily the ones in the
master data.

{p 4 8 2} 
{cmd:nolabel} prevents Stata from copying the value label definitions from the
using dataset.

{p 4 8 2} 
{cmd:nokeep} causes merge to ignore observations in the using data that have
no corresponding observation in the master.

{p 4 8 2}
{cmd:_merge(}{it:varname}{cmd:)} specifies the name of the variable that
will mark the source of the resulting observation. The default is
{cmd:_merge(_merge)}.

{p 4 4 2}
These options are all inherited from the parent {cmd:merge} command.


{title:Remarks}

{p 4 4 2}
The master datasets should be {help miset} before using {cmd:mimerge}.
The number of using datasets must be the same as that of the master datasets.
The names follow {cmd:mi} naming conventions, see help {help miset}.


{title:Examples}

{p 4 4 2}
Assume {cmd:foo1.dta} to {cmd:foo5.dta} have variables {cmd:id} and
{cmd:drkfre}, and {cmd:bar1.dta} to {cmd:bar5.dta} have variables {cmd:id}
and {cmd:smk}. By issuing

{p 4 8 2}{cmd:. miset using foo}{p_end}
{p 4 8 2}{cmd:. mido sort id}{p_end}
{p 4 8 2}{cmd:. misave foo, replace}{p_end}
{p 4 8 2}{cmd:. miset using bar}{p_end}
{p 4 8 2}{cmd:. mido sort id}{p_end}
{p 4 8 2}{cmd:. mimerge id using foo}{p_end}
{p 4 8 2}{cmd:. misave foobar}

{p 4 4 2}
the dataset {cmd:foo}{it:m}{cmd:.dta} is merged to {cmd:bar}{it:m}{cmd:.dta}
to form a new dataset {cmd:foobar}{it:m}{cmd:.dta} ({it:m} = 1,...,5), which
contains variables {cmd:id}, {cmd:drkfre}, and {cmd:smk}.


{title:Authors}

{p 4 4 2}
Ning Li, Philip Greenwood, and John Carlin, Clinical Epidemiology &
Biostatistics Unit, Murdoch Children's Research Institute and University of
Melbourne. Email 
{browse "mailto:jbcarlin@unimelb.edu.au":jbcarlin@unimelb.edu.au}


{title:Also see}

{p 4 14 2}
Online: help for {help merge}, {help miset}, {help miappend}, {help misave},
{help mido}, {help mici}, {help mifit}, {help milincom}, {help mireset}{p_end}