Closed quanp closed 7 years ago
FYI, this is what I am getting:
---------------------------
CASSCF calculation
---------------------------
* nstate : 1
* nclosed : 89
* nact : 11
* nvirt : 364
=== CASSCF iteration (molden) ===
* Using the second-order algorithm
0 0 -2132.21131564 7.72e-06 3.09
res : 2.71e-04 lamb: 1.00e+00 eps : -2.40e-07 step: 4.16e-04 2.71
res : 8.28e-05 lamb: 1.00e+00 eps : -2.50e-07 step: 4.45e-04 2.71
res : 3.72e-05 lamb: 1.00e+00 eps : -2.51e-07 step: 4.57e-04 2.90
res : 1.95e-05 lamb: 1.00e+00 eps : -2.51e-07 step: 4.64e-04 2.77
res : 8.78e-06 lamb: 1.00e+00 eps : -2.51e-07 step: 4.65e-04 2.84
res : 3.67e-06 lamb: 1.00e+00 eps : -2.51e-07 step: 4.63e-04 2.80
res : 1.74e-06 lamb: 1.00e+00 eps : -2.51e-07 step: 4.62e-04 2.83
res : 8.72e-07 lamb: 1.00e+00 eps : -2.51e-07 step: 4.62e-04 2.81
res : 3.47e-07 lamb: 1.00e+00 eps : -2.51e-07 step: 4.62e-04 2.87
1 0 -2132.21131577 1.55e-07 70.62
* Second-order optimization converged. *
* Integral transformation done. Elapsed time: 10.32
guess 0: closed .1......... open 1.11.1..1..
Thank you for your fast respond. I will check this. But what I see from your log file, BAGEL runs fine because the active space defined in the Molden file is already converged by Molcas, thus it only takes two iterations for Bagel to converge.
I can confirm that the bug is fixed.
Best regards. Quan
Great!
Hello,
I really like the idea of using Molden file from other programs as an input for Bagel. I'm testing a system containing a metal center and I found a bug while trying to read a Molden file containing g basis functions (generated by Molcas). You can see the input, molden, and log files here: https://github.com/quanp/Bagel-test
It seems that Bagel ignores the g function, giving in total 455 basis functions, while it should be 464 basis functions. After that, Bagel crashes with a segmentation fault
gdb gives:
Thank you in advance, Quan