st: RE: bcskew0 transform back

["Nick Cox" <n.j.cox@durham.ac.uk>]

I don't understand this desire. 

The back-transform of a transform (meaning invertible transform) of a
variable is just the original variable. 

If you had gone ln(mpg), then exp(ln(mpg)) gets you back again but you
might as well go directly to mpg. 

These commands are doing something a bit more complicated but the same
principle applies. Or what am I missing? 

Nick 
n.j.cox@durham.ac.uk 

raoul reulen

After i used bcskew0 I want to transform the relevant variable back.
How would I do this? For lnskew0 I do it like this:

.sysuse auto
.lnskew0 lnmpg = mpg
.local gamma = r(gamma)
.replace lnmpg = exp(lnmpg) + `gamma'

but I cannot work out how to do this for bcskew0. Obviously I need to
save r(lambda) to a local, but I can't get it right. Any suggestions?

.sysuse auto
.bcskew0 lnmpg = mpg
.local lambda=r(lambda)
???


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/