double precision       -> float64
null double precision  -> null.Float64
array double precision -> types.Float64Array
array real             -> types.Float64Array

We were so happy to add decimal support we overdid it.

- Fix bug report in comment on #162