{smcl}
{* 2007-02-26}{...}
{cmd:help jnsw} {right:Version 1.0 2007-02-26}
{hline}
{title:Title}
{p2colset 5 14 14 2}{...}
{p2col:{hi:jnsw} {hline 2}}Fit Johnson's system of transformations by Wheeler's quantile method{p_end}
{p2colreset}{...}
{title:Syntax}
{p 8 17 2}
{cmd:jnsw}
{varname}
{ifin}
[{cmd:,} {it:options}]
{synoptset 20 tabbed}{...}
{synopthdr}
{synoptline}
{syntab:Main}
{synopt:{opth g:enerate(newvar)}}create variable named {it:newvar} that is the fitted normal transformation of {varname}{p_end}
{synopt:{opt di:stribution(type)}}force fit to a specified Johnson distribution type{p_end}
{synopt:{opt lnt:olerance(#)}}tolerance for discriminating from the log-normal line{p_end}
{synopt:{opt tbo:neonly}}use only the upper quantiles{p_end}
{synopt:{opt p:ercentile(#)}}use a subset of the data{p_end}
{synopt:{opt ga:mma(#)}}force fit with a specified value for {it:gamma}{p_end}
{synopt:{opt d:elta(#)}}force fit with a specified value for {it:delta}{p_end}
{synopt:{opt l:ambda(#)}}force fit with a specified value for {it:lambda}{p_end}
{synopt:{opt x:i(#)}}force fit with a specified value for {it:xi}{p_end}
{synoptline}
{p2colreset}{...}
{p 4 6 2}
{cmd:by} may be used with {cmd:jnsw}; see {helpb by}.{p_end}
{title:Description}
{pstd}
{cmd:jnsw} fits Johnson systems of transformations using the quantile method of Wheeler (1980).
{title:Options}
{dlgtab:Main}
{phang}
{cmd:distribution} forces the fit to a user-specified Johnson distribution type;
acceptable types are {it:SL}, {it:SB} and {it:SU} in upper case, lower case or mixed
case. The default is to select the type by reference to the log-normal line.
{phang}
{cmd:lntolerance} specifies the tolerance for discriminating {it:SB} and {it:SU}
distribution types from {it:SL}, based upon difference from the log-normal line. The
default is 0.01, acceptable values are 0.0001 to 0.5, with smaller values resulting in
more fits declared as {it:SL} type.
{phang}
{cmd:tboneonly} requests that only the upper tail contribute to the the intermediary
{it:tb} result; the default is to use the average of two {it:tb}s determined from
both tails.
{phang}
{opt percentile(#)} requests that only the values of {varname} less than or equal to
the # percentile contribute to the fit; the default is 100.
{phang}
{opt gamma}
{opt delta}
{opt lambda}
{opt delta} forces to fit to use user-specified values of the parameter.
{title:Remarks}
{pstd}
{cmd:jnsw} fits parameters of Johnson distributions by the method of quantiles (Wheeler, 1980).
The method selects and fits the Johnson distribution on the basis of five quantiles, one of
which is the median. Flynn (2006) determined that, at least for fitting certain {it:SB}
distributions, Wheeler's quantile method performs better than two other popular percentile
methods that use only four percentiles.
{pstd}
Details of the method for fitting {it:gamma} and {it:delta} are given in Wheeler (1980). In
particular, the default of using both tails to calculate the intermediary statistic, {it:tb},
is based upon a suggestion in the article that this is expected to provide for more accurate
results in fitting {it:delta}. The article illustrates the method using {it:tb} estimated from
only one tail, however, and the {opt tboneonly} option will force this should the user desire to
follow Wheeler's illustrations.
{pstd}
{it:lambda} and {it:xi} are fit using ordinary least-squares linear regression, as suggested in the article.
{pstd}
The {opt percentile} option is made available for use with censored data, as suggested in the article.
{pstd}
For the {it:SL} Johnson type, {cmd:jnsw} deviates from the article in that (i) {it:gamma} is
explicitly fit (Wheeler combines it into {it:delta}, assuming {it:lambda} will be explicitly fit),
(ii) {it:delta} is constrained to be positive in accordance with convention (Wheeler
allows it to be negative for negatively skewed data, again, assuming that {it:lambda} will be allowed
to vary freely), and (iii) {it:lambda} is set to 1 if the data are positively skewed and -1 if
the data are negatively skewed. This latter convention of constraining {it:lambda} to unity, with
its sign reflecting direction of skew, follows Hill, Hill and Holder (1976). That method is
implemented in {help jnsn} (if installed). These deviations from Wheeler's article allow results
between the two methods to be directly comparable. They also allow the results from {cmd:jnsw},
which are returned in return scalars and a return macro, to be used in conjunction with {help ajv}
(if installed) to generate random variates that follow the Johnson distribution fit for {varname}
by {cmd:jnsw}.
{title:References}
{pstd}
M. R. Flynn, Fitting human exposure data with the Johnson SB distribution. {it:Journal of Exposure Science and Environmental Epidemiology} {bf:16}:56{c 150}62, 2006.
{pstd}
I. D. Hill, R. Hill and R. L. Holder, Fitting Johnson curves by moments. {it:Applied Statistics} {bf:25}:180{c 150}89, 1976.
{pstd}
R. E. Wheeler, Quantile estimators of Johnson curve parameters. {it:Biometrika} {bf:67}:725{c 150}28, 1980.
{title:Examples}
{phang}{cmd:. jnsw x}
{phang}{cmd:. jnsw x, di(sb) p(75) l(101) x(0)}
{title:Author}
{pstd}
Joseph Coveney
jcoveney@bigplanet.com
{title:Also see}
{psee}
Manual: {bf:[R] summarize}, {bf:[R] boxcox}, {bf:[R] lnskew0}, {bf:[R] ladder}
{psee}
Online: {helpb jnsn}, {helpb ajv}, {helpb transint}, {helpb summarize}, {helpb boxcox}, {helpb lnskew0}, {helpb ladder},
{helpb xriml}