Hi Alan,
can you try to specify the full path to the input directory? You should
also be able to omit the full classname and use bin/mahout
recommenditembased instead.
You could also try to manually start the hadoop job like this:
hadoop jar<path_to_job-jar> org.apache.mahout.cf.taste.hadoop.item.RecommenderJob --output
useroutput -n 10 --usersFile users_test.txt -s SIMILARITY_PEARSON_CORRELATION --input userdata/
--sebastian
On 18.11.2010 11:16, Alan Said wrote:
> Hi all,
> While trying to run the example from https://cwiki.apache.org/confluence/display/MAHOUT/Itembased+Collaborative+Filtering
>
> With the following command
> bin/mahout org.apache.mahout.cf.taste.hadoop.item.RecommenderJob --output useroutput
-n 10 --usersFile users_test.txt -s SIMILARITY_PEARSON_CORRELATION --input userdata/
>
> I keep getting this error:
> Exception in thread "main" org.apache.hadoop.mapreduce.lib.input.InvalidInputException:
Input path does not exist: userdata/
>
> I've tried all kinds of path combinations, still same error.
>
> Another weird thing is if I put --input userdata/ as the first argument I get
> Unexpected usage:<command> [Generic Options] [Job-Specific Options]
>
> I'm obviously missing something, so any help would be appreciated.
>
> /Alan
>
|