Millions of clusters are not likely to make sense.

Tens of thousands might just make sense.


On Fri, Jun 14, 2013 at 9:48 AM, Neetha <[email protected]> wrote:

> Thank you for the reply. At Ted: if we are talking in the sense of a
> millions of users there will be a millions of cluster , do this clustering
> be feasible.
>
>
> On Fri, Jun 14, 2013 at 1:27 AM, Ted Dunning <[email protected]>
> wrote:
>
> >  [image: Boxbe] <https://www.boxbe.com/overview> This message is
> eligible
> > for Automatic Cleanup! ([email protected]) Add cleanup rule<
> https://www.boxbe.com/popup?url=https%3A%2F%2Fwww.boxbe.com%2Fcleanup%3Ftoken%3DUidJJCGBsgyNZdyg04nO%252BNvneXvGTMg50O7JlejSkQAOd91gWBRYa6rImYY52P8PP2QNQf2o6SMVPDmLny0W8ELvhxQpm7qeEdJw16b0QIVsH6MiPq6MiWqm4aWRqUNMYY3hHYtjfotF2DiEYRkFXQ%253D%253D%26key%3D5eB29OSPchFmdH044S6TA0UftcTY%252FTd7ebJrWBroYQA%253D&tc_serial=14371471469&tc_rand=1859759626&utm_source=stf&utm_medium=email&utm_campaign=ANNO_CLEANUP_ADD&utm_content=001>|
> More
> > info<
> http://blog.boxbe.com/general/boxbe-automatic-cleanup?tc_serial=14371471469&tc_rand=1859759626&utm_source=stf&utm_medium=email&utm_campaign=ANNO_CLEANUP_ADD&utm_content=001
> >
> >
> > Thanks Grant.  Exactly correct.
> >
> > Some pig or hive action is indicated here.  Or write a map-reduce where
> the
> > reducer does the vector generation.
> >
> >
> >
> > On Thu, Jun 13, 2013 at 7:13 PM, Grant Ingersoll <[email protected]
> > >wrote:
> >
> > > I think Ted was implying just write a script to aggregate the Movielens
> > > data by user id.  Should be pretty straightforward.
> > >
> > > On Jun 13, 2013, at 10:05 AM, Neetha <[email protected]> wrote:
> > >
> > > > Thank you, for the reply. How can we group the user.
> > > >
> > > >
> > > > On Thu, Jun 13, 2013 at 3:41 PM, Ted Dunning <[email protected]>
> > > wrote:
> > > >
> > > >> [image: Boxbe] <https://www.boxbe.com/overview> This message is
> > > eligible
> > > >> for Automatic Cleanup! ([email protected]) Add cleanup rule<
> > >
> >
> https://www.boxbe.com/popup?url=https%3A%2F%2Fwww.boxbe.com%2Fcleanup%3Ftoken%3DGYex%252FPN%252FsEWDwuSs%252F9AS43g45aYbNc1OMuaZA7xu3TRldhNItvxAspHuwKeaedBKYvZ5Ah5DVIK7%252F%252B0qQSbX3CvYa7lvPle4%252BTdcv5k4cI%252BL4yoMK8by1Rm7UhZnW7TcvFw%252FeqoeYWXhz%252BgDPSUIWA%253D%253D%26key%3D0Lbb2Ob2N7oax0oxeBQTRLmrOCps42qosLO9Gh82kvs%253D&tc_serial=14367563490&tc_rand=1983549237&utm_source=stf&utm_medium=email&utm_campaign=ANNO_CLEANUP_ADD&utm_content=001
> > >|
> > > More
> > > >> info<
> > >
> >
> http://blog.boxbe.com/general/boxbe-automatic-cleanup?tc_serial=14367563490&tc_rand=1983549237&utm_source=stf&utm_medium=email&utm_campaign=ANNO_CLEANUP_ADD&utm_content=001
> > > >
> > > >>
> > > >> You need to group by user before converting to vector to get
> sensible
> > > >> clustering.
> > > >>
> > > >>
> > > >> On Wed, Jun 12, 2013 at 1:06 PM, Grant Ingersoll <
> [email protected]
> > > >>> wrote:
> > > >>
> > > >>> The CSVVectorIterator in the Integration package will take in a CSV
> > > file
> > > >>> and produce vectors.  It assumes that each row is the equivalent
> of a
> > > >>> DenseVector (does MovieLens fit that?)  If you need otherwise, I'd
> > > >> suggest
> > > >>> starting with the code and modifying to fit your needs.
> > > >>>
> > > >>>
> > > >>> -Grant
> > > >>>
> > > >>> On Jun 12, 2013, at 6:11 AM, Neetha <[email protected]> wrote:
> > > >>>
> > > >>>> Hi,
> > > >>>>
> > > >>>>
> > > >>>> I am using 1m movielens.
> > > >>>>
> > > >>>> I need to run the K-means clustering using mahout and hadoop.
> > > Actually,
> > > >>>> 1st step in the clustering is to convert into a sequence file,
> then
> > > >> into
> > > >>>> vector format and then apply the clustering algorithm. My doubt
> is,
> > Is
> > > >>>> there any need to convert the movielens rating.csv file into a
> > > sequence
> > > >>>> file. If needed what are the commands for applying clustering
> > > technique
> > > >>>> using mahout and the hadoop.
> > > >>>>
> > > >>>> Thanking you,
> > > >>>> Neetha Suan Thampi
> > > >>>
> > > >>> --------------------------------------------
> > > >>> Grant Ingersoll | @gsingers
> > > >>> http://www.lucidworks.com
> > > >>>
> > > >>>
> > > >>>
> > > >>>
> > > >>>
> > > >>>
> > > >>
> > > >>
> > >
> > > --------------------------------------------
> > > Grant Ingersoll | @gsingers
> > > http://www.lucidworks.com
> > >
> > >
> > >
> > >
> > >
> > >
> >
> >
>

Reply via email to