Categorical.first should throw exception on length match

[ykallus]

import riptable as rt
ds1 = rt.Dataset({'a': rt.arange(1, 4)})
print(ds1.cat(rt.arange(10000)//4000).first(ds1['a']))

*key_0                 a
------   ---------------
     0                 1
     1   140380335215024
     2   140379305999776

compare to behavior with mean instead of first

[ykallus]

related to https://github.com/rtosholdings/riptable/issues/124?