R help - Apr 2014 - mvpart - easy way to calculate deviance explained by variables

Dear R-users,

I can use mvpart function of library "mvpart" to get a decision tree.
The
output of decision tree could be something like the following. I can use it
to manually calculate deviance explained by the variable in the first split
as below. There should be an easy way to tabulate deviance explained by all
variables. I cannot seem to find it. The function printcp gives you a nice
table with rel error which you can use to estimate R^2. But printcp gives
output for variously sized trees, not for variables entered into the model.

- Kumar Mainali

the deviance explained by cond  *1 – (2857+491)/3905.98 = 0.1426 or 14.26%.*


* 1) root 295 3905.9860 4.806780  *

*   2) cond>=194.15 77  491.2468 2.493506  *

     4) cond>=309.7 25   62.1600 1.440000 *

     5) cond< 309.7 52  388.0000 3.000000 *

*   3) cond< 194.15 218 2857.1560 5.623853  *

     *6) embed>=82.5 114  891.9649 4.017544  *

*        ...........................................*


Graduate Program in Ecology, Evolution and Behavior
The University of Texas at Austin    |    Austin, Texas 78712, USA
 ᐧ

	[[alternative HTML version deleted]]