On Fri, Feb 22, 2008 at 12:36 PM, marco hofmann <[EMAIL PROTECTED]> wrote:
> Hi all,
> I have a problem with the predict command. Here is an example:
>
> from rpy import *
> r.library("nnet")
> model = r("Fxy~x+y")
>
> df = r.data_frame(x = r.c(0,2,5,10,15)
> ,y = r.c(0,2,5,8,10)
> ,Fxy = r.c(0,2,5,8,10))
NNModel is being converted to python and r.predict is looking for the
R object, which is no longer available. You'll need to turn off
conversion.
> NNModel = r.nnet(model, data = df
> , size =10, decay =1e-3
> , lineout=True, skip=True
> , maxit=1000, Hess =True)
>
> XG = r.expand_grid(x = r.seq(0,7,1), y = r.seq(0,7,1))
> x = r.seq(0,7,1)
> y = r.seq(0,7,1)
>
> fit = r.predict(NNModel,XG)
>
> Running that skript delivers an error:
>
> Traceback (most recent call last):
> File "F:\test\Python\test.py", line 11, in <module>
> fit = r.predict(NNModel,XG)
> rpy.RPy_RException: Fehler in UseMethod("predict") : keine anwendbare
> Methode für "predict"
>
>
> But the same skript in R shows no problems. It seems that I got not the
> right way to speech to the predict function, but I can't see why.
>
> Here's the code in R:
>
> library(nnet)
>
> model <- Fxy~x+y
>
> df <- data.frame(x=c(0,2,5,10,15)
> ,y=c(0,2,5,8,10)
> ,Fxy=c(0,2,5,8,10))
>
> NNModel <- nnet(model, data =df, size =10
> , decay =1e-3, lineout=TRUE
> , skip=TRUE, maxit=1000, Hess =TRUE)
>
> XG <- expand.grid(x=seq(0,7,1),y=seq(0,7,1))
> x <- seq(0,7,1)
> y <- seq(0,7,1)
> fit <- predict(NNModel,XG)
>
>
> which shows no problems. Where is the difference? I want to use the nnet
> Library to interpolate several 3D Datas. The python code loops over
> Data files and sorts them.. Maybe I have to put the whole code to R?
>
> I have to say that I don't complete understand the predict function- But
> in R does it what I want. I looked in the thread of Thomas few days
> ago, but it does not really help me.
It is not a problem with R. See the message here for more details on
what I mentioned above:
http://sourceforge.net/mailarchive/forum.php?thread_name=BAY134-W138C87347F6BA30C7D4BFDA41C0%40phx.gbl&forum_name=rpy-list
Sean
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
rpy-list mailing list
rpy-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rpy-list
