I have the same problem - is the test file sensitive to the order of all
the inputs and outputs or just the outputs?
Is there any way around this? I really want to put a file with any valid
values through a predictor.
Ideally, I'd like to build a model once, and put many new test files
which have no output values filled in through it - that is I want
predictions for the new file. Sorry if this is a dim question, but I
just can't work out how to do it, how do you get predictions for a file
that doesn't have the output values filled in?
Elen Lima wrote:
Thank you for your answer. The files seem to be compatible. I had a
bigger CSV file that I split in two in order to get the same
proportion of true and false in both files. The names and sequences of
variables are the same. However, some of the nominal attributes did
not list their values in the same order. I changed it and it worked.
On 2/6/07, *Peter Reutemann* <fracpete(a)cs.waikato.ac.nz
I'm trying to execute a decision tree (J48)
with two different
one for training and one for test. Both files
have the same
getting the message below when I try to execute:
"Problem evaluating classifier:
train and test set are not compatible"
Does anyone know what is happening? Would anyone be able to help
this? Thank you.
The format doesn't seem to be exactly the same, since you get this
message. Did you check, whether not only the attributes are in the
order, but whether also all the nominal attributes list their
the same order (and have the same amount of values)? A visual diff
like KDiff3 could be handy for comparing the two datasets.
Peter Reutemann, Dept. of Computer Science, University of Waikato, NZ
<http://www.cs.waikato.ac.nz/%7Efracpete/> Ph. +64 (7)
Wekalist mailing list
Wekalist mailing list
The University of Stirling is a university established in Scotland by
charter at Stirling, FK9 4LA. Privileged/Confidential Information may
be contained in this message. If you are not the addressee indicated
in this message (or responsible for delivery of the message to such
person), you may not disclose, copy or deliver this message to anyone
and any action taken or omitted to be taken in reliance on it, is
prohibited and may be unlawful. In such case, you should destroy this
message and kindly notify the sender by reply email. Please advise
immediately if you or your employer do not consent to Internet email
for messages of this kind.