31 jan 2008 kl. 23.37 skrev Steve Rowe:
I think this is a critical area for pre-coding design.
I made a section on the main Wiki page for Design, and added under
it a link to a new whiteboard page for discussing this topic:
<http://cwiki.apache.org/confluence/display/MAHOUT/Collection(De-)Serialization
>
I don't think we need a strong data access API from the start, but I
think it is the first thing we should be working with. Unless someone
have some code they want to cotribute, I suggest some of quick and
dirty RAM hogging ARFF glueball.
Can we define a minimal data access API required by the initial
algorithms?
karl
