From: Larry Evans <cppljevans@...>
Date: 2011-01-08T23:40:22+09:00
Subject: Re: Nooby question : multidimensional arrays.

On 01/07/11 21:57, Larry Evans wrote:
[snip]
> Thus, given the sizes, you can create a member variable
> which is the expansion vector, then use that to access the
> elements by calculating the offset from the initial element
> using the dot product:
> 
>     i0*arr.ev[0]+i1*arr.ev[1]+...+in*arr.ev[n]
> 

Also, if you're interested in transposing the vector,
IIRC, that just invovles permuting the expansion vector.
IOW, if the original expansion vector was:

  evo = [e1,e2,e3,...,en-1,en]

then simply reversing it:

  evt = [en,en-1,....,e3,e2,e1]

and using the same vector of values for elements,
would result in a transposed array.  I guess you
could try that for just 2 dimensional array
to see if that works.

  -Larry