A structural constraint puts restrictions on the structure of the data rather than the data values themselves. We have not put any restrictions on the number of rows or columns – just that they have to be the same.

For example, we can restrict the matrix to be squared; that is, the number of rows must be equal to the number of columns.