st: re: yeo-johnson power transform

[Kit Baum <baum@bc.edu>]

(1) copy boxcox.ado to yeojohnson.ado and change program name.

(2) to handle Yeo-Johnson of LHS only, copy boxco_l.ado to  
yeojo_l.ado and alter yeojohnson.ado call appropriately.

(3) In yeojo_l.ado,



program define LHS
        args todo b lnf

        tempname theta ssq diffL        /* theta is theta for Lhs */
        tempvar err yt

        scalar `theta' = `b'[1,1]               /* this is theta */
        scalar `diffL'=reldif( `theta', 0)

        if `diffL'> 1e-10 {
                qui gen double `yt' =( ( $ML_y1^`theta'-1)/`theta' )
        }
        else {
                qui gen double `yt' = ln( $ML_y1 )
        }


Last block of code should say something like

	qui gen double `yt' = .
        if `diffL'> 1e-10 {
                qui replace `yt' =( ( ($ML_y1+1)^`theta'-1)/ 
`theta' ) if $ML_y1 >= 0
		qui replace `yt' = ( ( (abs($ML_y1)+1)^(2*`theta')-1)/(2*`theta' )  
if $ML_y1 < 0
        }
        else {
                qui replace `yt' = ln( $ML_y1 )
        }


per the formula given at http://rweb.stat.umn.edu/R/library/alr3/html/ 
powtran.html

The yeojohnson command should then be able to handle arbitrary "y"  
values (check to see if boxcox.ado screens out negative values, and  
if so remove that screen from the code).

Kit Baum, Boston College Economics
http://ideas.repec.org/e/pba1.html
An Introduction to Modern Econometrics Using Stata:
http://www.stata-press.com/books/imeus.html


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/