Mark sample with all values present for later use -------------------------------------------------
^marker^ [varlist] [weight] [^if^ exp] [^in^ range] [^,^ ^g^en^(^markvar^)^ ^st^rok]
Description -----------
^marker^ creates a 0-1 variable markvar, such that markvar has value
1 if all values are present (not missing) for every variable in varlist and have sensible non-zero weights if weights are specified and satisfy any ^if^ or ^in^ condition specified
and
0 otherwise.
The name of markvar defaults to ^_marker^, providing that is not already in use.
The main use of ^marker^ is to define samples so that the results of commands that automatically ignore any observations with missing values on any of the variables specified (e.g. ^regress^) are comparable with the results of other commands that do not do this (e.g. ^summarize^). Once the sample has been defined, it can be specified by
... ^if ^markvar^ == 1^
or even more concisely by
... ^if ^markvar
Options -------
^gen(^markvar^)^ specifies the name of a new variable for markvar.
^strok^ indicates that non-missing string values (i.e. values other than the empty string "") are acceptable for any string variables in varlist as defining present observations.
Examples --------
. ^marker rep78^ . ^su if _marker^
. ^regress a b c d e^ . ^marker a b c d e, g(regdata)^ . ^su if regdata^
Author ------
Nicholas J. Cox, University of Durham n.j.cox@@durham.ac.uk
Also see: --------
On-line: help for @mark@ Manual: [R] mark