FWIW, The place I use this the most is when solving PDEs using finite elements with Dirichlet/Periodic boundary conditions. If I want to preserve the symmetry of the matrix before factorization, I end up needing to delete some rows and columns. I could write the code to create the matrix the way I end up using it, but that would require every function that creates a matrix to know about the boundary conditions which complicates things considerably.
In Julia, I would be using SparseMatrixCSC, so the issue is different than for general N-dimensional tensors. It's probably possible implement this functionality for SparseMatrixCSC using the existing functionality for vectors. On Wednesday, February 12, 2014 8:04:24 AM UTC-5, Andreas Lobinger wrote: > > Hello colleagues, > > i couldn't track down, where the Nonsense applies, but in general (also in > Matlab) it's better for organizing the data in a fixed array and track > down, what you access from that. This helps the computer/language system. > > However, it's not fully uncommon to formulate algorithms that > reduce/consume an array in an iterative way and it's just easier to read, > if the non-content is discarded. So this helps the reader or author of > software. > > Implementation for the first is straight forward. > Implementation for the second is (and will be) under discussion as general > CS topic... > >
