SharedTable uses String for numerical data storage about bonej2 HOT 6 CLOSED

@ctrueden @mdoube The reason we're using Double instead of double[] is that (correct me if I'm wrong) with the former we can have null cells in the table. They denote special empty cells, which don't have values yet. Using NaN or 0 is problematic, because sometimes you might actually get those values as results, and then you can't tell between results and empty cells.

from bonej2.

It is allowed to have repeated rowHeaders, now used in mdoube@6617ad0

from bonej2.

shared table now stores numerical values as Doubles, and empty cells as null
mdoube@1685724

from bonej2.

@mdoube Why are you using DefaultColumn<Double>? Why not DoubleColumn? It is far more space efficient, backed by a double[] instead of a List of boxed Double objects.

from bonej2.

@ctrueden good point, will rectify for the next release of the table

from bonej2.

@rimadoma raises an important point. Swapping in DoubleColumn breaks lots of tests, perhaps for this reason. Reverted to DefaultColumn.

from bonej2.