I do have this type of problem too. But everything work out fine if you
ignore it. Any reason behind this?
On 11/26/2011 09:09 AM, Mike Spreitzer wrote:
This is on my MacBook Pro, and I have the current JDK that Apple
provides (1.6.0_29). I just installed the current Maven, which
describes it
This is on my MacBook Pro, and I have the current JDK that Apple provides
(1.6.0_29). I just installed the current Maven, which describes itself as
Apache Maven 3.0.3 (r1075438; 2011-02-28 12:31:09-0500). This is my first
ever install of Maven. I am also just getting started with Mahout. I
For those in the San Francisco area, there will be a Mahout User Meeting on
Nov. 29th at Lucid Imagination's offices. Details and RSVP are at
http://sf-mahout-11-11.eventbrite.com/
For those not in the SF area, I _believe_ we will be recording it and posting
it.
On Fri, Nov 25, 2011 at 4:46 AM, Isabel Drost wrote:
> On 24.11.2011 Ted Dunning wrote:
> > Actually, one of the most reliable ways to kill a database is to use it
> as
> > input or output for even a small Hadoop cluster. Having hundreds of
> > processes all open connections and read at once is
Thanks lot it works .
By using this code
SequenceFile.Reader reader = new SequenceFile.Reader(fs,path1, conf);
IntWritable key = new IntWritable();
WeightedVectorWritable value = new WeightedVectorWritable();
while (reader.next(key, value))
{
System.out.println(value.toString() + " belongs to clus
I think you need to add the --clustering (-cl) option to your KMeans step. By
default, we only calculate the centroids. FWIW, the ClusterDumper can also
dump out the points associated w/ a cluster once you have run the --clustering
step.
Also I don't think the clusterData/part-randomSeed is
On 24.11.2011 Sturm, Martin wrote:
> Since I only want to try it out "standalone" I was hoping that this was
> possible without any Hadoop stuff. Are there any tutorials or examples
> available that show how to load a Dataset? Because I do not even know what
> files are expected here.. cvs?
You ma
On 24.11.2011 Ted Dunning wrote:
> Actually, one of the most reliable ways to kill a database is to use it as
> input or output for even a small Hadoop cluster. Having hundreds of
> processes all open connections and read at once is fairly abusive.
Though that does not mean that data cannot by sy
Team,
While reading the sequencial file . it is returning null
These are the command which i executed.
For Converting the Sequence File to Chunk(sequence vector) :
raghu@Syed:/media/Work/mahout$ bin/mahout seqdirectory -i
/media/Work/mahout/examples/bin/sample/fileList -o
/media/Work/mahout/exampl
Hi Sebastian,
I converted the dataset by simply keeping all user/item pairs that had a
rating of above 3. I'm also using GenericItemBasedRecommender's
mostSimilarItems method instead of the recommend method to make
recommendations.
I'm certainly open to suggestions on better evaluation metrics. I
10 matches
Mail list logo
