egen.hlp

是一个经济学管理应用软件 很难找的 但是经济学学生又必须用到

{pmore2}
produces the mode (within {it:varlist}) for {it:varname}, which may be numeric
or string.  The mode is the value occurring most frequently.  If two or more
modes exist, the mode produced will be a missing value.  To avoid this, the
{opt minmode}, {opt maxmode}, or {opt nummode()} options may be used to
specify choices for selecting among the multiple modes.  {opt minmode} returns
the lowest value, and {opt maxmode} returns the highest value.
{opt nummode(#)} will return the {it:#}th mode, counting from the
lowest up.  Missing values are excluded from determination of the mode unless
{opt missing} is specified.  Even so, the value of the mode is recorded for
observations for which the values of {it:varname} are missing unless they are
explicitly excluded, that is, by 
{bind:{cmd:if} {it:varname} {cmd:< .} or {cmd:if} {it:varname} {cmd:!= ""}}.

{phang2}
{opt mtr(year income)}{p_end}
{pmore2}
may not be combined with {cmd:by}.  It returns the U.S. marginal income tax
rate for a married couple with taxable income {it:income} in year {it:year},
where 1930 {ul:<} {it:year} {ul:<} 2004.  {it:year} and {it:income} may be
specified as variable names or constants; e.g., {bind:{cmd:mtr(1993 faminc)}},
{cmd:mtr(surveyyr 28000)}, or {cmd:mtr(surveyyr faminc)}.  A blank or comma
may be used to separate {it:income} from {it:year}.

        {opth pc(exp)} [{cmd:, prop}]{right:(allows {cmd:by} {varlist}{cmd::})  }
{pmore2}
returns {it:exp} (within {it:varlist}) scaled to be a percentage of the total,
between 0 and 100.  The {opt prop} option returns {it:exp} scaled to be a
proportion of the total, between 0 and 1.

        {opth pctile(exp)} [{cmd:, p(}{it:#}{cmd:)}]{right:(allows {cmd:by} {varlist}{cmd::})  }
{pmore2}
creates a constant (within {it:varlist}) containing the {it:#}th percentile
of {it:exp}.  If {opt p(#)} is not specified, 50 is assumed, meaning medians.
Also see {cmd:median()}.

        {opth rank(exp)} [{cmd:,} [{opt f:ield}|{opt t:rack}|{opt u:nique}]]{right:(allows {cmd:by} {varlist}{cmd::})  }
{pmore2}
creates ranks (within {it:varlist}) of {it:exp}; by default, equal
observations are assigned the average rank.  The {cmd:field} option calculates
the field rank of {it:exp}: the highest value is ranked 1, and there is no
correction for ties.  That is, the field rank is 1 + the number of values
that are higher.  The {opt track} option calculates the track rank of
{it:exp}:  the lowest value is ranked 1, and there is no correction for ties.
That is, the track rank is 1 + the number of values that are lower.  The
{opt unique} option calculates the unique rank of {it:exp}: values are ranked
1,...,{it:#}, and values and ties are broken arbitrarily.  Two values that
are tied for second are ranked 2 and 3.

{phang2}
{opth rowfirst(varlist)}{p_end}
{pmore2}
may not be combined with {cmd:by}.  It gives the first nonmissing value in
{it:varlist} for each observation (row).  If all values in {it:varlist} are
missing for an observation, {newvar} is set to missing.

{phang2}
{opth rowlast(varlist)}{p_end}
{pmore2}
may not be combined with {cmd:by}.  It gives the last nonmissing value in
{it:varlist} for each observation (row).  If all values in {it:varlist} are
missing for an observation, {newvar} is set to missing.

{phang2}
{opth rowmax(varlist)}{p_end}
{pmore2}
may not be combined with {cmd:by}.  It gives the maximum value (ignoring
missing values) in {it:varlist} for each observation (row).  If all values in
{it:varlist} are missing for an observation, {newvar} is set to missing.

{phang2}
{opth rowmean(varlist)}{p_end}
{pmore2}
may not be combined with {cmd:by}.  It creates the (row) means of the
variables in {it:varlist}, ignoring missing values; for example, if three
variables are specified and, in some observations, one of the variables is
missing, in those observations {newvar} will contain the mean of the two
variables that do exist.  Other observations will contain the mean of all
three variables.  Where none of the variables exist, {it:newvar} is set to
missing.

{phang2}
{opth rowmin(varlist)}{p_end}
{pmore2}
may not be combined with {cmd:by}.  It gives the minimum value in {it:varlist}
for each observation (row).  If all values in {it:varlist} are missing for an
observation, {newvar} is set to missing.

{phang2}
{opth rowmiss(varlist)}{p_end}
{pmore2}
may not be combined with {cmd:by}.  It gives the number of missing variables
in {it:varlist} for each observation (row).  

{phang2}
{opth rownonmiss(varlist)} [{cmd:,} {opt s:trok}]{p_end}
{pmore2}
may not be combined with {cmd:by}.  It gives the number of nonmissing
values in {it:varlist} for each observation (row){hline 2}this is the value used
by {opt rowmean()} for the denominator in the mean calculation.

{pmore2}
String variables may not be specified unless option {opt strok} is also
specified.  If {opt strok} is specified, string variables will be counted as
containing missing values when they contain "".  Numeric variables will be
counted as containing missing when their value is "{ul:>}.".

{phang2}
{opth rowsd(varlist)}{p_end}
{pmore2}
may not be combined with {cmd:by}.  It creates the (row) standard deviations
of the variables in {it:varlist}, ignoring missing values.

{phang2}
{opth rowtotal(varlist)}{p_end}
{pmore2}
may not be combined with {cmd:by}.  It creates the (row) sum of the variables
in {it:varlist}, treating missing as 0.

        {opth sd(exp)}{right:(allows {cmd:by} {varlist}{cmd::})  }
{pmore2}
creates a constant (within {it:varlist}) containing the standard
deviation of {it:exp}.  Also see {opt mean()}.

        {opt seq()} [{cmd:,} {opt f:rom(#)} {opt t:o(#)} {opt b:lock(#)}]{right:(allows {cmd:by} {varlist}{cmd::})  }
{pmore2}
returns integer sequences.  Values start from {opt from()} (default 1) and
increase to {opt to()} (the default is the maximum number of values) in
{opt blocks} (default size 1).  If {opt to()} is less than the maximum number,
sequences restart at {opt from()}.  Numbering may also be separate within groups
defined by {it:varlist} or decreasing if {opt to()} is less than {opt from()}.
Sequences depend on the sort order of observations, following three rules: (1)
observations excluded by {help if} or {help in} are not counted,
(2) observations are sorted by {it:varlist}, if specified; and (3) otherwise,
the order is that when called.  Note that no {it:arguments} are specified.

        {opth skew(varname)}{right:(allows {cmd:by} {varlist}{cmd::})  }
{pmore2}
returns the skewness (within {it:varlist}) of {it:varname}.
When {opt skew()} is specified, {newvar} will always be produced as
{it:type} {opt double}.  The {it:type} specified by the user will be ignored.

{phang2}
{opth std(exp)} [{cmd:,} {opt m:ean(#)} {opt s:td(#)}]{p_end}
{pmore2}
may not be combined with {cmd:by}.  It creates the standardized values of
{it:exp}.  The options specify the desired mean and standard deviation.  The
default is {cmd:mean(0)} and {cmd:std(1)}, producing a variable with mean 0,
standard deviation 1.

{phang2}
{opth tag(varlist)} [{cmd:,} {opt m:issing}]{p_end}
{pmore2}
may not be combined with {cmd:by}.  It tags just one observation in each
distinct group defined by {it:varlist}.  When all observations in a group have
the same value for a summary variable calculated for the group, it will be
sufficient to use just one value for many purposes.  The result will be 1 if
the observation is tagged and never missing, and 0 otherwise.  Values
for any observations excluded by either [{it:{help if}}] or [{it:{help in}}]
are set to 0 (not missing).  Hence, if {opt tag} is the variable
produced by {cmd:egen tag =} {opt tag(varlist)}, the idiom {opt if tag}
is always safe.  {opt missing} specifies that missing values of {it:varlist}
may be included.

        {opth total(exp)}{right:(allows {cmd:by} {varlist}{cmd::})  }
{pmore2}
creates a constant (within {it:varlist}) containing the sum of {it:exp}.  Also
see {cmd:mean()}.


{title:Description}

{pstd}
{cmd:egen} creates {newvar} of the optionally specified storage type equal to
{it:fcn}{cmd:(}{it:arguments}{cmd:)}.  Here {it:fcn}{cmd:()} is a function
specifically written for {cmd:egen}, as documented below or as written by
users.  Note that only {cmd:egen} functions may be used with {cmd:egen}, and
conversely, only {cmd:egen} may be used to run {cmd:egen} functions.

{pstd}
Depending on {it:fcn}{cmd:()}, {it:arguments}, if present, refers to an
expression, {varlist}, or a {it:{help numlist}}, and the {it:options}
are similarly {it:fcn} dependent.  Note that explicit subscripting (using
{cmd:_N} and {cmd:_n}), which is commonly used with {cmd:generate}, should not
be used with {cmd:egen}; see {help subscripting}.


{title:Examples}

{phang}{cmd:. egen avg = mean(chol)}{p_end}
{phang}{cmd:. gen dev = chol - avg}

    {cmd:. egen medgro = median(inc80-inc79)}{right:({it:exp}, {cmd:-} means subtraction)  }

{phang}{cmd:. egen avginc = rowmean(inc78 inc79 inc80)}{p_end}
    {cmd:. egen avginc = rowmean(inc78 - inc80)}{right:({it:varlist}, {cmd:-} means through)  }
{phang}{cmd:. egen taxrate= mtr(1978 inc78)}

{phang}{cmd:. egen stdscor = std(score)}{p_end}
{phang}{cmd:. egen newscor = std(score), mean(100) std(10)}

{phang}{cmd:. egen ttlsales = total(sales), by(region)}

{phang}{cmd:. egen racesex = group(race sex)}{p_end}
{phang}{cmd:. ir deaths smokes pyears, by(racesex)}


{title:Also see}

{psee}
Manual:  {bf:[D] egen}

{psee}
Online:  {helpb collapse}, {helpb generate}
{p_end}