Colleagues,

We are happy to announce that new versions of geomorph and RRPP have been 
published on CRAN. Please be sure to install both, as they function together.  
Installing from Github will also provide the latest updates.

With these updates we have adjusted and elevated a few internal functions 
regarding effect size calculation and other components, but mostly the updates 
fixed bugs and increased internal function efficiency.  Notably, the 'box.cox' 
function we elevated to principal function a few weeks ago has been replaced by 
powerTrans, to avoid conflicts with the defunct car::box.cox function.

You most likely won't notice much change, with one exception.  For internal 
operations, powerTrans might use an initial Yeo-Johnson transformation for 
values <= 0.  We also found issues with Z-score calculations for statistic 
ranges between 0 and 1, like those produced with PLS correlation coefficients.  
Standardizing these first avoids those issues, but also invokes a Yeo-Johnson 
transformation.  In the event, older effect sizes are desired, they can still 
be obtained.  Here is an example:

myPLS <- two.b.pls(...)
powerTrans(myPLS$rand.r, ...)

In other words, users have more freedom to dictate how effect sizes should be 
calculated.  So, you might notice slight changes in Z-scores, but if you wish 
to obtain the same Z-scores as before this update, it should be easy to do.

Happy computing!

Mike and Dean

Dr. Dean C. Adams
Distinguished Professor
Director, Ecology and Evolutionary Biology Graduate Program
Department of Ecology, Evolution, and Organismal Biology
Iowa State University
www.public.iastate.edu/~dcadams/<http://www.public.iastate.edu/~dcadams/>
phone: 515-294-3834

-- 
You received this message because you are subscribed to the Google Groups 
"Morphmet" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion visit 
https://groups.google.com/d/msgid/morphmet2/SJ0PR04MB83066E2C72161CB24F072220A2C62%40SJ0PR04MB8306.namprd04.prod.outlook.com.

Reply via email to