On May 25, 2017, at 10:33 AM, Marina Santini <marinaromestockholm@gmail.com> wrote:

 If you set a parameter in
that window, a switch will appear next to the name of the classifier.
For instance, if you set useKernelEstimator to True, the switch -K
will appear next to NaiveBayes.

Peter Reutemann’s “Command to code”
http://weka.8497.n7.nabble.com/unofficial-Weka-package-command-to-code-td37040.html
might be useful if you are a little used to reading java.

I also wrote something that could be useful, from a prior post

http://www195.pair.com/mik3hall/WekaSerialConverter.java   [ More on that if of interest at http://www195.pair.com/mik3hall/weka.html ]
is code I wrote to convert pre-Weka 3.8 models to 3.8 ones by re-running them.
I added a “-o” option to just get the command line options from an existing model, like…

java -cp .:weka38.jar WekaSerialConverter -o iris_rf.ser.model
weka.classifiers.trees.RandomForest -P 100 -I 100 -num-slots 1 -K 0 -M 1.0 -V 0.001 -S 1

So if you create a model of a classification configuration you can run this Weka based code to see what the parameters were that you used.
You can possibly figure out how the parameters correspond to the gui options from that.

Michael Hall





Hi,

I would like to take the liberty to suggest 2 improvements to the Weka
Explorer interface.

1) Preprocess tab: At present, it is hard to understand whether the
Remove filter button works correctly. In my experience, it does not.
It would be helpful to erase the name of the previously applied filter
when pressing the Remove filter button and display "None", as a
confirmation that the filter has been correctly un-applied from the
preprocessing of a dataset.

2) Refer to all parameter windows: I make an example to explain this
need. Example:
When selecting Naive Bayes in the Classify tab, if you click on the
name on the classifier, a window is displayed with all the available
parameters available for that classifier. If you set a parameter in
that window, a switch will appear next to the name of the classifier.
For instance, if you set useKernelEstimator to True, the switch -K
will appear next to NaiveBayes.

Sometimes it is not straightforward to understand to which parameter
in the window the switch corresponds to (eg, in the case of SMO).
Would it be possible to display also the switch in the parameter
window? For example: useKernelEstimator (-K): False

If I have been unclear, please do not hesitate to let me know.

Thanks in advance

Keep up the good work.

Marina
_______________________________________________
Wekalist mailing list
Send posts to: Wekalist@list.waikato.ac.nz
List info and subscription status: https://list.waikato.ac.nz/mailman/listinfo/wekalist
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html