Closed alanpapalia closed 1 year ago
In attempting to compute a 2nd order descent direction via LOBPCG we are occasionally obtaining a NaN value as the least eigenvalue and a vector of NaNs as the corresponding eigenvector. This is currently what is causing a test to fail in https://github.com/MarineRoboticsGroup/cora-plus-plus/pull/19/commits/88648de590e30364e34ab77ca81d80abb3f3dfde
Issue has been resolved AFAIK by https://github.com/MarineRoboticsGroup/cora-plus-plus/pull/19/commits/bb74f162c99761e8b2d55b63aaebed85eb33dc75
In attempting to compute a 2nd order descent direction via LOBPCG we are occasionally obtaining a NaN value as the least eigenvalue and a vector of NaNs as the corresponding eigenvector. This is currently what is causing a test to fail in https://github.com/MarineRoboticsGroup/cora-plus-plus/pull/19/commits/88648de590e30364e34ab77ca81d80abb3f3dfde