The following post is on a problem of finding elements of a matrix based on a flat representation of the matrix stored in a vector. A flat representation means that dimensionality (i.e. number of rows and number of columns of the matrix) of the matrix is lost, and all matrix elements are just written into a vector follwing a certain order. For a general matrix this does not make any sense, but for a special class of matrices, such as symmetric matrices, this does really give us some benefits.

The problem

Given the following problem. Suppose we have a large symmetric matrix of a certain dimension n of which the lower triangular part is stored in a vector. We want to find all row- and column-indices of elements of the original matrix that fullfill a certain property.

More Information

Please continue reading here