Addressing Multidimensional Arrays

A useful method of representing a function of
n variables is to consider the function to assume 
its values at selected points in n-dimensional space.
 Although this picture is of value to the analyst, 
the elements of an n-dimensional array must exist in
conventional storage as a linear array or vector. 
 The means of performing the transformation of a set
of indices locating on array element in n-space 
to the location (address) of the element in its storage
vector is the subject of this paper.  It is noted 
that the index address transformation is computationally
identical to the conversion of a number from 
a fixed to a mixed radix number system.  Several ways
of implementing the transformation are described.

CACM April, 1962

Hellerman, H.

CA620402 JB March 20, 1978  2:29 PM

435	5	615
615	5	615
615	5	615
615	5	615
849	5	615
155	6	615
196	6	615
615	6	615
615	6	615