[R] Cycling and different xerror values in cp of rpart, which result should I trust?...

[Prof Brian Ripley]

Do see the posting guide, and do not re-post 
(https://stat.ethz.ch/pipermail/r-help/2014-November/423727.html).

I suspect I am not alone in not knowing what you are talking about (even 
though I am the rpart maintainer).  Please do try to explain clearly 
what you mean with the reproducible example asked for in the posting 
guide, and show us the output you get and want to understand.

And a hint: rpart cross-validation is random, so set a seed and include 
that in your 'reproducible example'.


On 01/12/2014 09:00, Levent TERLEMEZ wrote:
> Dear Users,
> Is it possible to get cycling and different xerror scores in cp of rpart while getting the same tree on every run (or am I missing something or understood wrong)? I build it with an old version of R in a Windows VM installed on a Intel based macbook and reuse the same codes with a newer version of R and required packages on an AMD pc installed with standalone Windows. Both Windows are 64bit, which result should I trust, the old one or new one, macbook or pc?
>
> Thanks in advanced,
> Levent TERLEMEZ.
>
>
> 	[[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
PLEASE do (and no HTML).

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Emeritus Professor of Applied Statistics, University of Oxford
1 South Parks Road, Oxford OX1 3TG, UK