mahout-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From chyi-kwei yau <chyikwei....@gmail.com>
Subject Re: ** Random forest: attrib description string for org.apache.mahout.classifier.df.tools.Describe **
Date Fri, 21 Sep 2012 15:53:04 GMT
Hi,

1) you should remove all lines start with @. Just start with the data lines.
2) In the parameter "-d"
     -d "N 4 C"
    You need to specify which one is the "Label"
    So if the last one is the label, maybe change it to
    -d "N 4 L"

Hope this can help you and please let me know if there is anything
wrong in my description.

Best,
Chyi-Kwei

On Fri, Sep 21, 2012 at 11:33 AM, Rajesh Nikam <rajeshnikam@gmail.com> wrote:
> I want to use Random forest for arff file using BuildForest. This requires
> info file generated using org.apache.mahout.classifier.df.tools.Describe.
> However facing issue how to give description string.
>
> Please let me what is missing.
>
> $ hadoop@ml55:/usr/local/mahout/trunk/examples$ hadoop jar
> target/mahout-examples-0.8-SNAPSHOT-job.jar
> org.apache.mahout.classifier.df.tools.Describe -p ./testdata/hello.arff -f
> ./testdata/hello.info -d "N 4 C"
>
> gives following error:
>
>>>> Exception in thread "main"
> org.apache.mahout.classifier.df.data.DescriptorException: Bad Token : 4
>
> when tried following parameter it gives following error:
>
> $ hadoop jar examples/target/mahout-examples-0.8-SNAPSHOT-job.jar
> org.apache.mahout.classifier.df.tools.Describe -p ./testdata/hello.arff -f
> ./testdata/hello.info -d "N N N N C"
>
> Exception in thread "main" java.lang.IllegalArgumentException: Wrong number
> of attributes in the string
>      at
> com.google.common.base.Preconditions.checkArgument(Preconditions.java:92)
>
>
> ------ start sample arff file ------
>
> @relation hello
>
> @attribute a numeric
> @attribute b numeric
> @attribute c numeric
> @attribute d numeric
> @attribute class {'normal', 'anomaly'}
>
> @data
>
> 1,32,43,4,normal
> 21,22,3,4,normal
> 3,2,3,4,anomaly
> 45,12,33,4,anomaly
> 16,22,34,4,anomaly
>
> ------ end sample arff file ------
>
> Thanks
> Rajesh

Mime
View raw message