sample.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

{smcl}
{* 23feb2005}{...}
{cmd:help sample}{right:dialog:  {bf:{dialog sample}}}
{hline}

{title:Title}

{p2colset 5 19 21 2}{...}
{p2col :{hi:[D] sample} {hline 2}}Draw random sample{p_end}
{p2colreset}{...}


{title:Syntax}

{p 8 15 2}
{cmd:sample}
{it:#}
{ifin}
[{cmd:,} {opt c:ount} {opt by(groupvars)}]

{p 4 6 2}
{opt by} may be used with {opt sample}; see {helpb by}.
{p_end}


{title:Description}

{pstd}
{opt sample} draws random samples of the data in memory.  "Sampling" here is
defined as drawing observations without replacement; see {helpb bootstrap}
for sampling with replacement.

{pstd}
The size of the sample to be drawn can be specified as a percent or as a
count:

{pin}
    {opt sample} without the {opt count} option draws a {it:#} percent
    pseudorandom sample of the data in memory, thus discarding 100 - {it:#}
    percent of the observations.

{pin}
    {opt sample} with the {opt count} option draws a {it:#}-observation
    pseudorandom sample of the data in memory, thus discarding {cmd:_N} - {it:#}
    observations.  {it:#} can be larger than {help _N}, in which case all
    observations are kept.

{pstd}
In either case, observations not meeting the optional {opt if} and {opt in}
criteria are kept (sampled at 100%).

{pstd}
If you are interested in reproducing results, you must first set the
random-number seed; see {helpb generate}.


{title:Options}

{phang}
{opt count} specifies that {it:#} in {opt sample} {it:#} be
    interpreted as an observation count rather than as a percentage.  Typing
    {opt sample 5} without the {opt count} option means that a 5% sample be
    drawn; typing {opt sample 5, count}, however,
    would draw a sample of 5 observations.

{pmore}
    Specifying {it:#} as greater than the number of observations in the dataset
    is not considered an error.

{pmore}
    {opt count} may be combined with the {opt by} prefix;
    e.g., {cmd:by sex: sample 50, count} would draw a sample of size 50 for
    men and 50 for women.

{phang}
{opt by(groupvars)} specifies that a {it:#}-percent sample be drawn within
    each set of values of {it:groupvars}, thus maintaining the proportion of
    each group.


{title:Examples}

{p2colset 5 37 39 2}{...}
{p2col :{cmd:. sample 10}}(draw 10% sample){p_end}
{p2col :{cmd:. sample 10 if race==0}}(keep all race!=0,
	but sample 10% of race==0){p_end}
{p2col :{cmd:. by race: sample 10}}(sample 10% of each race)

{p2col :{cmd:. sample 50, count}}(draw 50-obs. sample){p_end}
{p2col :{cmd:. sample 50 if race==0, count}}(keep all race!=0,
	but sample 50 obs. of race==0){p_end}
{p2col :{cmd:. by race: sample 50, count}}(sample 50 obs. of each race)
{p2colreset}{...}


{title:Also see}

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

{psee}
Online:  {helpb bootstrap}
{p_end}