Fixing the units for the DPP

[Kevin Waugh]

I haven't worked it out completely, but I've convinced myself to a minimal level of satisfaction that one can reduce a DPP to a k-DPP by adding in appropriately sized vectors that are orthogonal to everything.

The length and number of said vectors isn't trivial to compute (if at all possible), but at least the units are right now!

[Kevin Waugh]

Drew's response: 

I generally think units are  a big deal, but they aren't in this case.

In the orthogonal case it's somewhat easy to see--- just consider either multiplying by \lambda or by 1 (both in meters) if you include or don't include a particular element respectively.

Kevin is describing a non-orthogonal version of the same.


--d