On 2012-04-02 03:07, Elizabeth Tseng wrote:
Hi,
I've just started using rpy/rpy2. I have a SVM model trained using the
kernlab package. I stored the model in a R file called
"MPDR.training_size2000.RData" (so if I call 'load' in R, the object
'model' is the trained model).
I'm trying to load the model through rpy2, then run predict() on it.
My code is as follows:
import rpy2.rpy_classic as rpy
import rpy2.robjects as robj
rpy.set_default_mode(rpy.NO_CONVERSION)
model = rpy.r.load("MPDR.training_size2000.RData")
rpy.r.predict(model, <some data here>)
and I get the error:
no applicable method for 'predict' applied to an object of class
"character"
It seems like I'm not loading the model correctly. What can be done?
The R function loads objects in R's "global environment"; what is
returned by load() is the names/symbols of the objects loaded AFAIR.
Your model is somewhere in R's global environment.
L.
PS: rpy2 has a more Python-like interface to that
http://rpy.sourceforge.net/rpy2/doc-2.2/html/robjects_serialization.html
Thanks,
--Elizabeth
------------------------------------------------------------------------------
This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here
http://p.sf.net/sfu/sfd2d-msazure
_______________________________________________
rpy-list mailing list
rpy-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rpy-list
------------------------------------------------------------------------------
This SF email is sponsosred by:
Try Windows Azure free for 90 days Click Here
http://p.sf.net/sfu/sfd2d-msazure
_______________________________________________
rpy-list mailing list
rpy-list@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rpy-list
