Diagonalizing PairStateInteractions fails for interactionsUpTo = 2 about arc-alkali-rydberg-calculator

[nikolasibalic]

I've reproduced the error. I haven't seen this error before. I will take a look and let you know.

[nikolasibalic]

@wetsixteen Simple solution is just to reduce the number of requested eigenstates from 250 to 200. That's second argument in diagonalise method. So the corrected call that should give result is

calc.diagonalise(np.linspace(0.2,3,300), 200, progressOutput=True)

Let me know if you have problems with this.

Reason: For the parameters set, you have only 231 purely dipole-dipole coupled states. That you can see if you run code with interactoinsUpTo=1 , it will tell you in matrix dimension what is the total number of states you have. Therefore if you ask for 250 eigenstates, it will have to diagonalise also quadrupole-dipole an quadrupole-quadrupole coupled states. For larger interatomic distances, these states essentially don't interact, and their close energy differences might be causing a trouble for ARPAC that seeks eigenstates closest to the given eigenValue (that is, it cannot sort eigenstates by their energies if they are essentially degenerate). That's why originally diagonalisation works, but then for larger inter-atomic distances it breaks, having finished first ~30%.

Future steps: I will try to check if this was the underlying reason and also adjust underlying routines so that in this case they output some user-friendly comment to help users.

@wetsixteen Thank you for submitting this. Please let me know if you experience any other problems.

[vxq9e7bx]

works perfect, thanks!