Re: timesSquaredJob cardinality exception

[Sean Owen]

In the recommender world we use (hashed) item IDs as dimensions in
sparse vectors and matrices. Conceptually the cardinality of the
matrix and vector is infinite for these purposes, but we settle for
Integer.MAX_VALUE. So, I believe you'll want to use this same value
for all cardinalities. Being sparse, of course, these entities won't
actually have a billion entries.