-------------------------------------------------------------------------------
help for nmissing, npresent               (SJ3-4:dm67_2; SJ5-4:dm67_3 in press)
-------------------------------------------------------------------------------

Numbers of missing (present) values

        nmissing [varlist] [if exp] [in range] [, min(#) obs piasm trim ]

        npresent [varlist] [if exp] [in range] [, min(#) obs piasm trim ]

    by ...: may be used with nmissing and npresent; see help by.


Description

    nmissing lists the number of missing values in each variable in varlist
    (or optionally in each observation). Missing means numeric missing (that
    is, sysmiss . or one of the extended missing values .a to .z) for numeric
    variables and empty or "" for string variables.  See help on missing for
    further information.

    npresent lists the number of present (non-missing) values in each
    variable in varlist (or optionally in each observation).


Options

    min(#) specifies that only numbers at least # should be listed.  Default
        1. The argument may be zero or any positive integer or any expression
        evaluating to zero or a positive integer. As a special case, _all or
        * means the number of variables in varlist or, with obs, all the
        observations specified.

    obs specifies that counting is for observations, not variables.

    piasm specifies that period is also string missing: that is, "." counts
        as missing. Some Stata users use "." to indicate missing.

    trim specifies that whether string values are missing is to be determined
        after trimming leading and trailing spaces. For example, with trim "
        " (a single blank) would count as missing.

    trim piasm would specify that e.g. " . " counts as missing.


Saved results 

    If the obs option is not used, r(varlist) holds the list of variables
        containing more than min() missing values. If the obs option is used,
        r(varlist) just echoes varlist.


Examples

    . nmissing

    . nmissing rep78

    To see which observations have all missing values:

    . nmissing, min(*) obs

    To get details on variables with 10 or more missing values:

    . nmissing, min(10)
    . d `r(varlist)'

    . npresent


Author

    Nicholas J. Cox, Durham University, U.K.
    n.j.cox@durham.ac.uk


Acknowledgements 

    A question by Eric Uslaner led to the addition of r(varlist) as a saved
    result.


Also see

    Online:  help for missing; dropmiss (if installed)