lnskew0.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}
{* 20jan2005}{...}
{cmd:help lnskew0}, {cmd:help bcskew0}{right:dialogs:  {bf:{dialog lnskew0}}  {bf:{dialog bcskew0}}}
{hline}

{title:Title}

{p2colset 5 20 22 2}{...}
{p2col :{hi:[R] lnskew0} {hline 2}}Find zero-skewness log or Box-Cox
transform{p_end}
{p2colreset}{...}


{title:Syntax}

{phang}
Zero-skewness log

{p 8 17 2}
{cmd:lnskew0} {newvar} {cmd:=} {it:{help exp}} {ifin} [{cmd:,} {it:options}]

{phang}
Zero-skewness Box-Cox transform

{p 8 17 2}
{cmd:bcskew0} {newvar} {cmd:=} {it:{help exp}} {ifin} [{cmd:,} {it:options}] 

{synoptset 12 tabbed}{...}
{synopthdr}
{synoptline}
{syntab :Main}
{synopt :{opt d:elta(#)}}increment for derivative of skewness function; default
is {cmd:delta(0.02)} for {opt lnskew0} and {cmd:delta(0.01)} for {opt bcskew0}{p_end}
{synopt :{opt z:ero(#)}}value for determining convergence; default is
{cmd:zero(0.001)}{p_end}
{synopt :{opt l:evel(#)}}set confidence level; default is {cmd:level(95)}{p_end}
{synoptline}
{p2colreset}{...}


{title:Description}

{pstd}
{cmd:lnskew0} creates {newvar} = ln(+/-{it:exp} - k), choosing k and the sign
of {it:{help exp}} so that the skewness of {it:newvar} is zero.

{pstd}
{cmd:bcskew0} creates {it:newvar} = ({it:exp}^L - 1)/L, the Box-Cox power
transformation, choosing L so that the skewness of {it:newvar} is zero.
{it:exp} must be strictly positive.  Also see {helpb boxcox} for maximum
likelihood estimation of L.


{title:Options}

{dlgtab:Main}

{phang}
{opt delta(#)} specifies the increment used for calculating the derivative of
the skewness function with respect to k ({cmd:lnskew0}) or L ({cmd:bcskew0}).
The default values are 0.02 for {cmd:lnskew0} and 0.01 for {cmd:bcskew0}.

{phang}
{opt zero(#)} specifies a value for skewness to determine convergence that
is small enough to be considered zero and is, by default, 0.001.

{phang}
{opt level(#)} specifies the confidence level for the confidence interval for
k ({cmd:lnskew0}) or L ({cmd:bcskew0}).  The confidence interval is calculated
only if {opt level()} is specified.  {it:#} is specified as an integer; 95
means 95% confidence intervals.  The {opt level()} option is honored only if
the number of observations exceeds 7.


{title:Examples}

{phang}{cmd:. lnskew0 lnmpg = mpg}{p_end}
{p 16 16 2}creates {cmd:lnmpg} containing ln({cmd:mpg}-k) with skewness 0.

{phang}{cmd:. lnskew0 lnmpg = mpg, level(95)}{p_end}
{p 16 16 2}... and reports 95% confidence intervals for k.

{phang}{cmd:. bcskew0 bcmpg = mpg}{p_end}
{p 16 16 2}creates {cmd:bcmpg} containing ({cmd:mpg}^L-1)/L with skewness 0.

{phang}{cmd:. bcskew0 bcmpg = mpg, level(90)}{p_end}
{p 16 16 2}... and reports 90% confidence interval for L.


{title:Also see}

{psee}
Manual:  {bf:[R] lnskew0}

{psee}
Online:  {helpb boxcox}, {helpb ladder}, {helpb swilk}
{p_end}