You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@mahout.apache.org by Ricardo <ri...@meemix.com> on 2010/12/29 14:32:38 UTC
itemsimilarity
Hi:
I am completely new at mahout. I have just installed all the components
needed, however I don't succeed running a query.
I prepared a text file that looks: "user id, item id, rate", and run the
syntax ./mahout itemsimilarity -input ../././text file --output
../././folder
However it doesn't succeed. I've looked all over, but can't find in the NET
a clear example/tutorial that I can follow in order to learn the syntax.
THXS
Ricardo
Re: itemsimilarity
Posted by Sebastian Schelter <ss...@apache.org>.
Hi Ricardo,
please don't forget to answer to user@mahout.apache.org too, your mails
seem to be adressed to me only. Additionally attachments don't work on
the mailing list.
You can happily ignore the files written into the temp-folder, the are
just needed for the job's internals.
You'll find more information here:
https://cwiki.apache.org/confluence/display/MAHOUT/Itembased+Collaborative+Filtering
and here:
http://www.slideshare.net/sscdotopen/mahoutcf
--sebastian
Am 02.01.2011 10:46, schrieb Ricardo:
> THXS.
> I did that, and 3 folders were created under the directory that I added.
> Only one of them (itemIDIndex) had files in it (attached), and are
> unreadable.
> Isn't any example somewhere, where I can practice with existing data to get
> used to the syntax and results?
>
> I tried the links below, but none of them are detailed enough for somebody
> who never used the program before.
>
> https://cwiki.apache.org/confluence/display/MAHOUT/Recommender+Documentation
>
> http://philippeadjiman.com/blog/2009/11/11/flexible-collaborative-filtering-
> in-java-with-mahout-taste/
>
>
> THXS
>
> Ricardo
>
> -----Original Message-----
> From: Sebastian Schelter [mailto:ssc@apache.org]
> Sent: Sunday, January 02, 2011 11:28 AM
> To: Ricardo; user@mahout.apache.org
> Subject: Re: itemsimilarity
>
> Hi Ricardo,
>
> Mahout's jobs need a temporary directory for intermediate files. If you
> don't supply a path for that, they will create a temp folder in the
> directory where you started the job. Additionally a job will not be able
> to write to that directory if it already exists, which is the cause for
> your exception.
>
> Add a parameter --tempDir to your invocation and make sure it points to
> a non existing directory.
>
> --sebastian
>
> Am 02.01.2011 10:22, schrieb Ricardo: THXS.
>> My syntax looks:
>> ./mahout itemsimilarity --input ../cf/taste/.... --output ..//...
>> --similarityClassname SIMILARITY_EUCLIDEAN_DISTANCE --booleanData
>> The output I get:
>> ...Exemption in thread "main"
>> org.apache.hadoop.mapred.FileAlreadyExistsException: Output directory
>> temp/itemIDIndex already exists .....
>>
>> Ricardo
>>
>>
>>
>> -----Original Message-----
>> From: Sebastian Schelter [mailto:ssc@apache.org]
>> Sent: Wednesday, December 29, 2010 5:59 PM
>> To: user@mahout.apache.org
>> Cc: ricardo@meemix.com
>> Subject: Re: itemsimilarity
>>
>> -dev +user
>>
>> Hi Ricardo,
>>
>> can you provide more details about what exactly doesn't work? Do you get
>> an exception or do you get empty results?
>>
>> --sebastian
>>
>> Am 29.12.2010 14:32, schrieb Ricardo:
>>> Hi:
>>> I am completely new at mahout. I have just installed all the components
>>> needed, however I don't succeed running a query.
>>>
>>> I prepared a text file that looks: "user id, item id, rate", and run the
>>> syntax ./mahout itemsimilarity -input ../././text file --output
>>> ../././folder
>>>
>>> However it doesn't succeed. I've looked all over, but can't find in the
>> NET
>>> a clear example/tutorial that I can follow in order to learn the syntax.
>>>
>>> THXS
>>>
>>>
>>>
>>> Ricardo
>>>
>>>
>>>
>>>
>>>
>>>
>>
>>
>>
>
>
>
Re: itemsimilarity
Posted by Sebastian Schelter <ss...@apache.org>.
Hi Ricardo,
Mahout's jobs need a temporary directory for intermediate files. If you
don't supply a path for that, they will create a temp folder in the
directory where you started the job. Additionally a job will not be able
to write to that directory if it already exists, which is the cause for
your exception.
Add a parameter --tempDir to your invocation and make sure it points to
a non existing directory.
--sebastian
Am 02.01.2011 10:22, schrieb Ricardo: THXS.
> My syntax looks:
> ./mahout itemsimilarity --input ../cf/taste/.... --output ..//...
> --similarityClassname SIMILARITY_EUCLIDEAN_DISTANCE --booleanData
> The output I get:
> ...Exemption in thread "main"
> org.apache.hadoop.mapred.FileAlreadyExistsException: Output directory
> temp/itemIDIndex already exists .....
>
> Ricardo
>
>
>
> -----Original Message-----
> From: Sebastian Schelter [mailto:ssc@apache.org]
> Sent: Wednesday, December 29, 2010 5:59 PM
> To: user@mahout.apache.org
> Cc: ricardo@meemix.com
> Subject: Re: itemsimilarity
>
> -dev +user
>
> Hi Ricardo,
>
> can you provide more details about what exactly doesn't work? Do you get
> an exception or do you get empty results?
>
> --sebastian
>
> Am 29.12.2010 14:32, schrieb Ricardo:
>> Hi:
>> I am completely new at mahout. I have just installed all the components
>> needed, however I don't succeed running a query.
>>
>> I prepared a text file that looks: "user id, item id, rate", and run the
>> syntax ./mahout itemsimilarity -input ../././text file --output
>> ../././folder
>>
>> However it doesn't succeed. I've looked all over, but can't find in the
> NET
>> a clear example/tutorial that I can follow in order to learn the syntax.
>>
>> THXS
>>
>>
>>
>> Ricardo
>>
>>
>>
>>
>>
>>
>
>
>
>
Re: itemsimilarity
Posted by Sebastian Schelter <ss...@apache.org>.
-dev +user
Hi Ricardo,
can you provide more details about what exactly doesn't work? Do you get
an exception or do you get empty results?
--sebastian
Am 29.12.2010 14:32, schrieb Ricardo:
> Hi:
> I am completely new at mahout. I have just installed all the components
> needed, however I don't succeed running a query.
>
> I prepared a text file that looks: "user id, item id, rate", and run the
> syntax ./mahout itemsimilarity -input ../././text file --output
> ../././folder
>
> However it doesn't succeed. I've looked all over, but can't find in the NET
> a clear example/tutorial that I can follow in order to learn the syntax.
>
> THXS
>
>
>
> Ricardo
>
>
>
>
>
>