custom sparse matrix format

[markbangert]

matlabs built-in sparse matrix format suffers from the following limitations only double precision possible (memory issues) uses int64 to store the index value which has a maximum value of intmax('int64')=9223372036854775807 for our purposes intmax('int32')=2147483647 would suffice

• considering worst case optimization and biological optimization where we have to store multiple values per index a custum format might save even more memory!

the question is can we design a format the still allows for efficient dose calculation - and I am afraid the answer is no.... :(

[wahln]

Since this is a won't fix issue anyways, just a very wild idea on this: 1) Create a custom sparse matrix class that overloads the corresponding sparsity-related functions. 2) In these overloaded functions use mex-functions based on the Eigen Template Library's sparse matrix implementation (https://eigen.tuxfamily.org/dox/classEigen_1_1SparseMatrix.html) which allows all kinds of datatypes for indices and values

I guess this would be too much of an effort....

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[stale[bot]]

This issue has been automatically closed because it has not seen any activity in four weeks. This happens usually when the issue has already been solved or it is no longer relevant. If that's not the case, feel free to reopen the issue.