Using the C++ Math Library

Array Indexing

Array indexing is implemented through the interaction of three classes: mwArray, mwSubArray, and mwIndex. When applied to an mwArray, operator() returns an mwSubArray. The mwSubArray ``remembers'' the indexing operation and defers evaluation until the result is either assigned or referred to.

You can pass an integer, double, mwArray, the contents of a cell or structure field, or an indexing expression as an argument to operator().

The mwArray class interface contains a series of operator() member functions that support n-dimensional indexing. Several of the functions are listed here.

This pair of operator() member functions supports one-dimensional indexing and indexing into arrays with more than 32 dimensions. The second non-const signature supports calls that are targets of the assignment operator and modify the contents of an array.

mwArray operator()(const mwVarargin &a) const;
mwSubArray operator()(const mwVarargin &a);

This pair of operator() member functions supports two-dimensional indexing The second non-const signature supports calls that are valid targets for the assignment operator.

mwArray operator()(const mwArray &a1, const mwArray &a2) const;
mwSubArray operator()(const mwArray &a1,
                     const mwArray &a2);

This pair of operator() member functions supports the maximum number of arguments. To index into more than 32 dimensions, you must construct an mwVarargin object.

mwArray operator()(const mwArray &a1,
                   const mwArray &a2,
                   const mwArray &a3,
                  .
                  .
                  .
                   const mwArray &a32) const;
mwSubArray operator()(const mwArray &a1,
                     const mwArray &a2,
                     const mwArray &a3,
                    .
                    .
                    .
                     const mwArray &a32);

Indexing and Subscripts

Cell Content Indexing