[Bug] Negative eigenvalueSee original GitHub issue
in cma.py CMA.tell() I somehow get to this point:
self._B, self._D = None, None self._C = [[ 8.14737017e-04 2.68770961e-02 -6.44404672e-03] [ 2.68770961e-02 8.40956376e-01 -4.00553778e-01] [ -6.44404672e-03 -4.00553778e-01 2.18242242e+00]] D2 = [ -6.44813495e-05 7.31257026e-01 2.29300099e+00] D = [ nan 0.85513568 1.51426582] This leads to self._C = [[ nan nan nan] [ nan nan nan] [ nan nan nan]] at the end of CMA.tell() Then in CMA._sample_solution() D2, B = np.linalg.eigh(self._C) crashes for obvious reasons LinAlgError: Eigenvalues did not converge
- Created 3 years ago
Top GitHub Comments
This error happens once in a while. Hard to tell when. I’d need to work on a way to make it reproducible. I’ve just looked at https://github.com/CyberAgent/cmaes/pull/34 and it seems this is what I actually get. I’ll try this PR to see if it fixes it.
Also, I’ve seen some indications in the code like
# (p.28) I think it refers to some book. Which one?