You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@mahout.apache.org by magicalo <ma...@yahoo.com> on 2011/12/01 19:24:40 UTC

20newsgroups example does not print verbose output

Hello,

I have ran the 20newsgroups example on my own data set. It runs successfully
and prints the summary output. However, I have enabled the verbose option in 
the script when I run the testclassifier, but I don't see any verbose output.
I am interested in seeing exactly which data was misclassified.
I am running the command in hdfs mode. I was expecting this verbose output 
be spilled on the command line, but I am not getting anything. I am also not 
seeing any log file or other file on the hdfs that would have this comparison:

a       b       c       <--Classified as
89      11      0        |  100     a     = out
1       99      0        |  100     b     = in
0       0       0        |  0       c     = unknown


Based on the TestClassifier.java class, I should be seeing this printed 
somewhere:
log.info("Line Number: " + lineNum + " Line(30): "
+ (line.length() > 30 ? line.substring(0, 30) : line)
+ " Expected Label: " + correctLabel + " Classified Label: "
+ classifiedLabel.getLabel() + " Correct: " + correct);


The exact command that I am running is: 
mahout testclassifier \
  -m ${WORK_DIR}/myproj-bydate/bayes-model \
  -d ${WORK_DIR}/myproj-bydate/bayes-test-input \
  -type bayes \
  -ng 1 \ 
  -source hdfs \
  -v \
  -method mapreduce

Any suggestions? Thanks


Re: 20newsgroups example does not print verbose output

Posted by magicalo <ma...@yahoo.com>.
Ted,

Do you have any idea where the log may be stored or how to view this output? Tx




Re: 20newsgroups example does not print verbose output

Posted by Lance Norskog <go...@gmail.com>.
Good question. Sending these things to logging is useless anyway; I'd
rather have a separate sequencefile or CSV file that I can work with.

On Thu, Dec 1, 2011 at 10:24 AM, magicalo <ma...@yahoo.com> wrote:

> Hello,
>
> I have ran the 20newsgroups example on my own data set. It runs
> successfully
> and prints the summary output. However, I have enabled the verbose option
> in
> the script when I run the testclassifier, but I don't see any verbose
> output.
> I am interested in seeing exactly which data was misclassified.
> I am running the command in hdfs mode. I was expecting this verbose output
> be spilled on the command line, but I am not getting anything. I am also
> not
> seeing any log file or other file on the hdfs that would have this
> comparison:
>
> a       b       c       <--Classified as
> 89      11      0        |  100     a     = out
> 1       99      0        |  100     b     = in
> 0       0       0        |  0       c     = unknown
>
>
> Based on the TestClassifier.java class, I should be seeing this printed
> somewhere:
> log.info("Line Number: " + lineNum + " Line(30): "
> + (line.length() > 30 ? line.substring(0, 30) : line)
> + " Expected Label: " + correctLabel + " Classified Label: "
> + classifiedLabel.getLabel() + " Correct: " + correct);
>
>
> The exact command that I am running is:
> mahout testclassifier \
>  -m ${WORK_DIR}/myproj-bydate/bayes-model \
>  -d ${WORK_DIR}/myproj-bydate/bayes-test-input \
>  -type bayes \
>  -ng 1 \
>  -source hdfs \
>  -v \
>  -method mapreduce
>
> Any suggestions? Thanks
>
>


-- 
Lance Norskog
goksron@gmail.com

Re: 20newsgroups example does not print verbose output

Posted by Grant Ingersoll <gs...@apache.org>.
What do you have for logging in your classpath?

On Dec 1, 2011, at 1:24 PM, magicalo wrote:

> Hello,
> 
> I have ran the 20newsgroups example on my own data set. It runs successfully
> and prints the summary output. However, I have enabled the verbose option in 
> the script when I run the testclassifier, but I don't see any verbose output.
> I am interested in seeing exactly which data was misclassified.
> I am running the command in hdfs mode. I was expecting this verbose output 
> be spilled on the command line, but I am not getting anything. I am also not 
> seeing any log file or other file on the hdfs that would have this comparison:
> 
> a       b       c       <--Classified as
> 89      11      0        |  100     a     = out
> 1       99      0        |  100     b     = in
> 0       0       0        |  0       c     = unknown
> 
> 
> Based on the TestClassifier.java class, I should be seeing this printed 
> somewhere:
> log.info("Line Number: " + lineNum + " Line(30): "
> + (line.length() > 30 ? line.substring(0, 30) : line)
> + " Expected Label: " + correctLabel + " Classified Label: "
> + classifiedLabel.getLabel() + " Correct: " + correct);
> 
> 
> The exact command that I am running is: 
> mahout testclassifier \
>  -m ${WORK_DIR}/myproj-bydate/bayes-model \
>  -d ${WORK_DIR}/myproj-bydate/bayes-test-input \
>  -type bayes \
>  -ng 1 \ 
>  -source hdfs \
>  -v \
>  -method mapreduce
> 
> Any suggestions? Thanks
> 

--------------------------------------------
Grant Ingersoll
http://www.lucidimagination.com