Windows is not "case sensitive" so you may have troubles with that command.
I suggest to put different names for input and ouput files.
On 30/05/11 5:16 PM, Nicoara Barbu Talpes wrote:
I have two csv files that have blog articles that I want to classify
over a numerical feature. The two files have four features each: three
numerical and one nominal (the blog article). I want to apply two
filters to the files: NominalToString and StringToWordVector and then
train the test set on the training set.
The problem I encounter is after applying the two filters on the two
data sets separately, when i try to classify the test set on the
training set i get the error that the two are not compatible. This means
the vectors that represent the 'blog article' features after being
applied the stringtovector filter are of different lengths for the two
I tried batch filtering from the cli with the command, but to the
following command i get: 'input file formats differ'.
attribute.NominalToString -C 3 -b -i "C:\weka-files\TRAINING.csv" -o
C:\weka-files\train.arff -r "C:\weka-files\POOL.csv" -s
How to proceed?
Use the FilteredClassifier - it "learns" the filter on the training
data, filters the training data before passing it to the classifier, and
then uses the learned filter to process any test instances before
passing them to the classifier for classification.
Wekalist mailing list
Send posts to: Wekalist(a)list.scms.waikato.ac.nz
List info and subscription status: