Re: Limit number of instances when using "Supplied test set"

Thanks for sharing the data with me. The issue is that both your datasets
are CSV files and WEKA's CSV reader can be a bit difficult to work with.
In your case, when it loads the test set, it does not have a sufficient
buffer size to collect all relevant values of nominal attributes in your
test data. The corresponding error is only visible in Program->LogWindow
of the GUIChooser (or the console, if you run WEKA with a console window):

java.io.IOException: Read unknown nominal value Ffor attribute aa5 (line:
138). Try increasing the size of the memory buffer (-B option) or
explicitly specify legal nominal values with the -L option.
	weka.core.converters.CSVLoader.makeInstance(CSVLoader.java:1014)
	weka.core.converters.CSVLoader.getNextInstance(CSVLoader.java:839)
	
weka.core.converters.ConverterUtils$DataSource.hasMoreElements(ConverterUt
ils.java:373)
	
weka.gui.explorer.ClassifierPanel$20.run(ClassifierPanel.java:1854)

	at
weka.core.converters.CSVLoader.makeInstance(CSVLoader.java:1014)
	at
weka.core.converters.CSVLoader.getNextInstance(CSVLoader.java:839)
	at
weka.core.converters.ConverterUtils$DataSource.hasMoreElements(ConverterUt
ils.java:373)
	at
weka.gui.explorer.ClassifierPanel$20.run(ClassifierPanel.java:1854)

It is possible to increase the buffer size to fix this. After clicking on
the "Open  file..." button in the window for specifying the "Test
Instances",  tick the box "Invoke options dialog", select the file, and
then click "Open". You will then get a GenericObjectEditor window to
configure the CSVLoader. Increase the buffer size from 100 to a
sufficiently large value (e.g., 1,000 in your case) and click "OK".
Clicking through to the "Start" button and running the classifier, you
will see that all 3,097 instances will be classified now.

(There should clearly be an explicit alert window in this case, and it
should not be necessary to check the log...)

Cheers,
Eibe

-----Original Message-----
From: [email protected] <[email protected]>
Sent: Thursday, 16 September 2021 5:47 AM
To: [email protected]
Subject: [Wekalist] Re: Limit number of instances when using "Supplied
test set"

Hi Eibe,

Here is the google drive link:
https://drive.google.com/drive/folders/1K4323o0MRDThhCZAJK_t_4q8qjzQ8sAI?u
sp=sharing

Could you help to take a look at the datasets? I'm trying to use dataset1
as training (random forest, class shows in the last column), test on
dataset2 and vice versa. But only a small part of test dataset has been
tested.

Best,
Shengyuan
_______________________________________________
Wekalist mailing list -- [email protected] Send posts to
[email protected] To unsubscribe send an email to
[email protected] To subscribe, unsubscribe, etc., visit
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette:
http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
_______________________________________________
Wekalist mailing list -- [email protected]
Send posts to [email protected]
To unsubscribe send an email to [email protected]
To subscribe, unsubscribe, etc., visit https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html