On Tue, Jun 8, 2010 at 12:40 AM, Ted Dunning <[email protected]> wrote:
> Take a look at the source code that he mentioned. THat will always be your
> best cue. Mahout is still very young and is changing rapidly so
> documentation tends to lag badly in some areas.
>
> Finalli I solved it .
Just added this code fragment
BayesParameters params = new BayesParameters(2);
params.set("alpha_i", "1.0");
params.set("dataSource", "hdfs");
I got this idea from the test code of TrainClassifier.
Thanks to all
--
**********************************
JAGANADH G
http://jaganadhg.freeflux.net/blog
