As the SDA evolves we're going to eventually have to take advantage of sparse matrix approaches to leverage the performance boosts associated with localization when inverting large matrices (e.g. in the Kalman Filter). Here's an example of a spare matrix package in R http://www.econ.uiuc.edu/~roger/research/sparse/SparseM.pdf
There are other matrix libraries in C and Fortran as well.
As the SDA evolves we're going to eventually have to take advantage of sparse matrix approaches to leverage the performance boosts associated with localization when inverting large matrices (e.g. in the Kalman Filter). Here's an example of a spare matrix package in R http://www.econ.uiuc.edu/~roger/research/sparse/SparseM.pdf There are other matrix libraries in C and Fortran as well.