Hi!
I’m using minimum hopping global optimization method in ase. In qn0000x.log file, as the following showed:
BFGSLineSearch: 32[ 54] 21:08:34 -97.530839* 0.0849
BFGSLineSearch: 33[ 56] 21:14:40 -97.531158* 0.0912
BFGSLineSearch: 34[ 57] 21:17:43 -97.531520* 0.0790
BFGSLineSearch: 35[ 58] 21:20:47 -97.531759* 0.0657
BFGSLineSearch: 36[ 59] 21:23:51 -97.531938* 0.0772
BFGSLineSearch: 37[ 61] 21:29:59 -97.532774* 0.1205
BFGSLineSearch: 38[117] 00:15:16 -97.532853* 0.1163
BFGSLineSearch: 39[118] 00:18:19 -97.533302* 0.0787
BFGSLineSearch: 40[119] 00:21:23 -97.533821* 0.0623
BFGSLineSearch: 41[124] 00:36:43 -97.534023* 0.0340
It seems that, between step 37 and 38, the computer uses much time in force calls(from 61 to 117). I am confused that what the role of force calls play? Why force calls number is always larger than step number?
Thank you for attention and help!