factorize-movielens-1M.sh is a small example that invokes HADOOP locally
to showcase how ALS factorizes a small dataset sitting in the local
parallelALS runs on cluster, that is its main purpose. Just invoke it
via mahout parallelALS or directly via hadoop jar ... just like any
other Hadoop job. Be aware that its slow as it needs a lot of iterations
which impose a lot of overhead on Hadoop.
If your data is small, then you can also use an SVDRecommender with
ALSWRFactorizer to factorize your data in-memory on a single machine.
On 19.01.2013 01:35, Kamal Ali wrote: