powerit
C++ implementation for power iterations
|
Options for compute()
.
More...
#include <core.hpp>
Public Attributes | |
int | iterations = 500 |
double | tolerance = 0.000001 |
int | num_threads = 1 |
int powerit::Options::iterations = 500 |
Maximum number of iterations to perform. Note that the algorithm may converge before reaching this limit.
int powerit::Options::num_threads = 1 |
Number of threads to use for the matrix multiplication in compute()
. The parallelization scheme depends on the definition of the #POWERIT_CUSTOM_PARALLEL function-like macro.
For compute_core()
, the parallelization scheme depends on the provided multiply()
, and this option has no effect.
double powerit::Options::tolerance = 0.000001 |
Tolerance used to determine convergence. The error is defined as the L2 norm of the difference between eigenvectors at successive iterations; if this drops below tolerance
, we assume that the algorithm has converged.