> We are considering having an arbitrary number of dimensions in the general
> case, but we will probably need to have rank-specific implementations for
> performance, and the question is how high is the rank in practice?
The limit in IDL (my preferred analysis package) is 8 and recently I found a
good reason to construct such an array. But this is rare and otherwise I
don't think I've ever gone past 4 or maybe 5.
---
Mark Hadfield
m.hadfield@xxxxxxxxxxx http://katipo.niwa.cri.nz/~hadfield
National Institute for Water and Atmospheric Research
