Options
Bug Report, WEKA:Naive Bayes
Maven
Bug Report, WEKA:Naive Bayes
Rapidminer 4.5
Dear All,
I hope this is a correct bug report.
I wasn't sure where to post.
The WEKA Naive Bayes refuses to run:
it claim's ComplementNaiveBayes: Cannot handle multi-valued nominal attributes!
But I think it should use weka.classifiers.bayes.NaiveBayes, not ComplementNaiveBayes!
I used the following data, and loaded it inside WEKA to confirm that NaiveBayes runs fine there.
Data:
"Day";"Outlook";"Temperature";"Humidity";"Wind";"PlayTennis"
"Dl";"Sunny";"Hot";"High";"Weak";"No"
"D2";"Sunny";"Hot";"High";"Strong";"No"
"D3";"Overcast";"Hot";"High";"Weak";"Yes"
"D4";"Rain";"Mild";"High";"Weak";"Yes"
"D5";"Rain";"Cool";"Normal";"Weak";"Yes"
"D6";"Rain";"Cool";"Normal";"Strong";"No"
"D7";"Overcast";"Cool";"Normal";"Strong";"Yes"
"D8";"Sunny";"Mild";"High";"Weak";"No"
"D9";"Sunny";"Cool";"Normal";"Weak";"Yes"
"Dl0";"Rain";"Mild";"Normal";"Weak";"Yes"
"Dl1";"Sunny";"Mild";"Normal";"Strong";"Yes"
"Dl2";"Overcast";"Mild";"High";"Strong";"Yes"
"Dl3";"Overcast";"Hot";"Normal";"Weak";"Yes"
"Dl4";"Rain";"Mild";"High";"Strong";"No"
Error Log:
P Sep 5, 2009 2:44:06 PM: Initialising process setup
P Sep 5, 2009 2:44:06 PM: [NOTE] No filename given for result file, using stdout for logging results!
P Sep 5, 2009 2:44:06 PM: Checking properties...
P Sep 5, 2009 2:44:06 PM: Properties are ok.
P Sep 5, 2009 2:44:06 PM: Checking process setup...
P Sep 5, 2009 2:44:06 PM: Inner operators are ok.
P Sep 5, 2009 2:44:06 PM: Checking i/o classes...
P Sep 5, 2009 2:44:06 PM: i/o classes are ok. Process output: Model.
P Sep 5, 2009 2:44:06 PM: Process ok.
P Sep 5, 2009 2:44:06 PM: [Warning] W-NaiveBayes: W-NaiveBayes: Deprecated: please use NaiveBayes instead.
P Sep 5, 2009 2:44:06 PM: [Warning] Deprecations: 1 usage of deprecated operators.
P Sep 5, 2009 2:44:06 PM: Process initialised
P Sep 5, 2009 2:44:06 PM: [NOTE] Process starts
P Sep 5, 2009 2:44:06 PM: Process:
Root[1] (Process)
+- CSVExampleSource[1] (CSVExampleSource)
+- W-NaiveBayes[1] (W-NaiveBayes)
P Sep 5, 2009 2:44:06 PM: [Warning] W-NaiveBayes: Deprecated: please use NaiveBayes instead.
P Sep 5, 2009 2:44:06 PM: [Warning] W-NaiveBayes: This learning scheme does not have sufficient capabilities for the given data set: polynominal attributes not supported
G Sep 5, 2009 2:44:06 PM: [Fatal] UserError occured in 1st application of W-NaiveBayes (W-NaiveBayes)
G Sep 5, 2009 2:44:06 PM: [Fatal] Process failed: W-NaiveBayes caused an error: weka.classifiers.bayes.ComplementNaiveBayes: Cannot handle multi-valued nominal attributes!
Root[1] (Process)
+- CSVExampleSource[1] (CSVExampleSource)
here ==> +- W-NaiveBayes[1] (W-NaiveBayes)
Rapidminer 4.5
Dear All,
I hope this is a correct bug report.
I wasn't sure where to post.
The WEKA Naive Bayes refuses to run:
it claim's ComplementNaiveBayes: Cannot handle multi-valued nominal attributes!
But I think it should use weka.classifiers.bayes.NaiveBayes, not ComplementNaiveBayes!
I used the following data, and loaded it inside WEKA to confirm that NaiveBayes runs fine there.
Data:
"Day";"Outlook";"Temperature";"Humidity";"Wind";"PlayTennis"
"Dl";"Sunny";"Hot";"High";"Weak";"No"
"D2";"Sunny";"Hot";"High";"Strong";"No"
"D3";"Overcast";"Hot";"High";"Weak";"Yes"
"D4";"Rain";"Mild";"High";"Weak";"Yes"
"D5";"Rain";"Cool";"Normal";"Weak";"Yes"
"D6";"Rain";"Cool";"Normal";"Strong";"No"
"D7";"Overcast";"Cool";"Normal";"Strong";"Yes"
"D8";"Sunny";"Mild";"High";"Weak";"No"
"D9";"Sunny";"Cool";"Normal";"Weak";"Yes"
"Dl0";"Rain";"Mild";"Normal";"Weak";"Yes"
"Dl1";"Sunny";"Mild";"Normal";"Strong";"Yes"
"Dl2";"Overcast";"Mild";"High";"Strong";"Yes"
"Dl3";"Overcast";"Hot";"Normal";"Weak";"Yes"
"Dl4";"Rain";"Mild";"High";"Strong";"No"
Error Log:
P Sep 5, 2009 2:44:06 PM: Initialising process setup
P Sep 5, 2009 2:44:06 PM: [NOTE] No filename given for result file, using stdout for logging results!
P Sep 5, 2009 2:44:06 PM: Checking properties...
P Sep 5, 2009 2:44:06 PM: Properties are ok.
P Sep 5, 2009 2:44:06 PM: Checking process setup...
P Sep 5, 2009 2:44:06 PM: Inner operators are ok.
P Sep 5, 2009 2:44:06 PM: Checking i/o classes...
P Sep 5, 2009 2:44:06 PM: i/o classes are ok. Process output: Model.
P Sep 5, 2009 2:44:06 PM: Process ok.
P Sep 5, 2009 2:44:06 PM: [Warning] W-NaiveBayes: W-NaiveBayes: Deprecated: please use NaiveBayes instead.
P Sep 5, 2009 2:44:06 PM: [Warning] Deprecations: 1 usage of deprecated operators.
P Sep 5, 2009 2:44:06 PM: Process initialised
P Sep 5, 2009 2:44:06 PM: [NOTE] Process starts
P Sep 5, 2009 2:44:06 PM: Process:
Root[1] (Process)
+- CSVExampleSource[1] (CSVExampleSource)
+- W-NaiveBayes[1] (W-NaiveBayes)
P Sep 5, 2009 2:44:06 PM: [Warning] W-NaiveBayes: Deprecated: please use NaiveBayes instead.
P Sep 5, 2009 2:44:06 PM: [Warning] W-NaiveBayes: This learning scheme does not have sufficient capabilities for the given data set: polynominal attributes not supported
G Sep 5, 2009 2:44:06 PM: [Fatal] UserError occured in 1st application of W-NaiveBayes (W-NaiveBayes)
G Sep 5, 2009 2:44:06 PM: [Fatal] Process failed: W-NaiveBayes caused an error: weka.classifiers.bayes.ComplementNaiveBayes: Cannot handle multi-valued nominal attributes!
Root[1] (Process)
+- CSVExampleSource[1] (CSVExampleSource)
here ==> +- W-NaiveBayes[1] (W-NaiveBayes)
0
Answers
we have a bugtracker at bugs.rapid-i.com. This is the right place for such nice and detailed bug reports
Greetings,
Sebastian
one short hint: You just could replace the WEKA Bayes with the rapid miner Naive Bayes learner. This should work anyway. And the learner doing exactly the same.
Greetings,
Sebastian
Since my forum account doesn't work?
unfortunately the forum and the bug tracker use incompatible table formats. You will have to create a new login if you want to report bugs.
Greetings,
Sebastian
PS: I took a look on the matter and saw, that the W-NaiveBayes operator is deprecated. So this bug has got reaaally low priority, because we have some bugs on not deprecated things, which come first...