Comments (12)
Evolutionary-Intelligence
commented on May 18, 2024
1
For FMAES, I got 0.152 on my personal computer using the following settings:
options = {
'fitness_threshold': 1e-4,
'max_runtime': 30_000,
'seed_rng': 500,
'max_function_evaluations': 100_000_000,
'sigma': 1.3,
'verbose': 500,
'stagnation': 100
}
I only added the 'stagnation' setting and increased 'sigma' from 0.3 to 1.3 for better exploration.
* Generation 0: best_so_far_y 3.45900e+01, min(y) 3.45900e+01 & Evaluations 16
* Generation 500: best_so_far_y 3.61442e-01, min(y) 3.66357e-01 & Evaluations 8016
* Generation 1000: best_so_far_y 2.96007e-01, min(y) 2.96007e-01 & Evaluations 16016
* Generation 1500: best_so_far_y 2.92347e-01, min(y) 2.92352e-01 & Evaluations 24016
* Generation 2000: best_so_far_y 2.75207e-01, min(y) 2.75319e-01 & Evaluations 32016
* Generation 2500: best_so_far_y 2.62970e-01, min(y) 2.62970e-01 & Evaluations 40016
* Generation 3000: best_so_far_y 2.62729e-01, min(y) 2.62729e-01 & Evaluations 48016
* Generation 3500: best_so_far_y 2.54470e-01, min(y) 2.54471e-01 & Evaluations 56016
* Generation 4000: best_so_far_y 2.54416e-01, min(y) 2.54416e-01 & Evaluations 64016
* Generation 4500: best_so_far_y 2.54416e-01, min(y) 2.54416e-01 & Evaluations 72016
* Generation 4654: best_so_far_y 2.54416e-01, min(y) 2.54416e-01 & Evaluations 74464
....... *** restart *** .......
* Generation 0: best_so_far_y 2.54416e-01, min(y) 1.89882e+01 & Evaluations 74496
* Generation 500: best_so_far_y 2.54416e-01, min(y) 3.33518e-01 & Evaluations 90496
* Generation 1000: best_so_far_y 2.41184e-01, min(y) 2.41184e-01 & Evaluations 106496
* Generation 1500: best_so_far_y 2.38546e-01, min(y) 2.38546e-01 & Evaluations 122496
* Generation 2000: best_so_far_y 2.38490e-01, min(y) 2.38490e-01 & Evaluations 138496
* Generation 2152: best_so_far_y 2.38490e-01, min(y) 2.38490e-01 & Evaluations 143328
....... *** restart *** .......
* Generation 0: best_so_far_y 2.38490e-01, min(y) 2.58909e+01 & Evaluations 143392
* Generation 500: best_so_far_y 2.38490e-01, min(y) 2.38844e-01 & Evaluations 175392
* Generation 1000: best_so_far_y 2.19327e-01, min(y) 2.19327e-01 & Evaluations 207392
* Generation 1500: best_so_far_y 2.17826e-01, min(y) 2.17886e-01 & Evaluations 239392
* Generation 2000: best_so_far_y 2.17233e-01, min(y) 2.17233e-01 & Evaluations 271392
* Generation 2500: best_so_far_y 2.17181e-01, min(y) 2.17181e-01 & Evaluations 303392
* Generation 3000: best_so_far_y 2.17172e-01, min(y) 2.17172e-01 & Evaluations 335392
* Generation 3500: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 367392
* Generation 4000: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 399392
* Generation 4500: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 431392
* Generation 5000: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 463392
* Generation 5500: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 495392
* Generation 6000: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 527392
* Generation 6500: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 559392
* Generation 7000: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 591392
* Generation 7500: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 623392
* Generation 8000: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 655392
* Generation 8500: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 687392
* Generation 9000: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 719392
* Generation 9500: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 751392
* Generation 9583: best_so_far_y 2.17170e-01, min(y) 2.17170e-01 & Evaluations 756640
....... *** restart *** .......
* Generation 0: best_so_far_y 2.17170e-01, min(y) 2.11403e+01 & Evaluations 756768
* Generation 500: best_so_far_y 2.17170e-01, min(y) 2.78004e-01 & Evaluations 820768
* Generation 1000: best_so_far_y 2.17170e-01, min(y) 2.25295e-01 & Evaluations 884768
* Generation 1297: best_so_far_y 2.17170e-01, min(y) 2.25235e-01 & Evaluations 922656
....... *** restart *** .......
* Generation 0: best_so_far_y 2.17170e-01, min(y) 2.41040e+01 & Evaluations 922912
* Generation 500: best_so_far_y 2.17170e-01, min(y) 2.27704e-01 & Evaluations 1050912
* Generation 968: best_so_far_y 1.97796e-01, min(y) 1.97796e-01 & Evaluations 1170464
....... *** restart *** .......
* Generation 0: best_so_far_y 1.97796e-01, min(y) 2.07147e+01 & Evaluations 1170976
* Generation 500: best_so_far_y 1.97796e-01, min(y) 2.16960e-01 & Evaluations 1426976
* Generation 1000: best_so_far_y 1.84787e-01, min(y) 1.85319e-01 & Evaluations 1682976
* Generation 1500: best_so_far_y 1.52076e-01, min(y) 1.52076e-01 & Evaluations 1938976
* Generation 1636: best_so_far_y 1.52076e-01, min(y) 1.52076e-01 & Evaluations 2008096
....... *** restart *** .......
* Generation 0: best_so_far_y 1.52076e-01, min(y) 1.50320e+01 & Evaluations 2009120
* Generation 500: best_so_far_y 1.52076e-01, min(y) 3.68050e-01 & Evaluations 2521120
* Generation 1000: best_so_far_y 1.52076e-01, min(y) 1.75294e-01 & Evaluations 3033120
* Generation 1483: best_so_far_y 1.52076e-01, min(y) 1.64093e-01 & Evaluations 3526688
....... *** restart *** .......
* Generation 0: best_so_far_y 1.52076e-01, min(y) 1.59880e+01 & Evaluations 3528736
* Generation 348: best_so_far_y 1.52076e-01, min(y) 6.50549e-01 & Evaluations 4239392
....... *** restart *** .......
* Generation 0: best_so_far_y 1.52076e-01, min(y) 1.74142e+01 & Evaluations 4243488
* Generation 500: best_so_far_y 1.52076e-01, min(y) 6.52559e-01 & Evaluations 6291488
* Generation 1000: best_so_far_y 1.52076e-01, min(y) 3.01358e-01 & Evaluations 8339488
* Generation 1016: best_so_far_y 1.52076e-01, min(y) 3.04507e-01 & Evaluations 8400928
....... *** restart *** .......
* Generation 0: best_so_far_y 1.52076e-01, min(y) 1.35338e+01 & Evaluations 8409120
* Generation 181: best_so_far_y 1.52076e-01, min(y) 6.78880e-01 & Evaluations 9883680
....... *** restart *** .......
* Generation 0: best_so_far_y 1.52076e-01, min(y) 8.43856e+00 & Evaluations 9900064
* Generation 375: best_so_far_y 1.52076e-01, min(y) 6.89216e-01 & Evaluations 16027680
....... *** restart *** .......
* Generation 0: best_so_far_y 1.52076e-01, min(y) 1.12617e+01 & Evaluations 16060448
* Generation 141: best_so_far_y 1.52076e-01, min(y) 6.80394e-01 & Evaluations 20647968
....... *** restart *** .......
* Generation 0: best_so_far_y 1.52076e-01, min(y) 8.35395e+00 & Evaluations 20713504
* Generation 195: best_so_far_y 1.52076e-01, min(y) 6.74632e-01 & Evaluations 33427488
....... *** restart *** .......
* Generation 0: best_so_far_y 1.52076e-01, min(y) 1.01387e+01 & Evaluations 33558560
* Generation 259: best_so_far_y 1.52076e-01, min(y) 6.71572e-01 & Evaluations 67375136
....... *** restart *** .......
* Generation 0: best_so_far_y 1.52076e-01, min(y) 1.17326e+01 & Evaluations 67637280
* Generation 124: best_so_far_y 1.52076e-01, min(y) 6.79190e-01 & Evaluations 100000000
0.15207564457445216
from pypop.
Evolutionary-Intelligence
commented on May 18, 2024
Thanks very much for your tips. We are checking your code, in order to help improve it.
from pypop.
Chang-SHAO
commented on May 18, 2024
I just run your codes and get the results:
100%|██████████| 300/300 [01:02<00:00, 4.77it/s, best_result_so_far=0.158]
message: Optimization terminated successfully.
success: True
status: 0
fun: 0.1583149714373654
x: [ 4.046e-01 1.729e+00 ... -5.250e-01 4.071e-01]
nit: 283
jac: [ 1.756e-05 -2.446e-05 ... -2.330e-06 9.231e-06]
hess_inv: [[ 1.666e+00 6.581e-02 ... 5.589e-01 -1.330e-01]
[ 6.581e-02 5.340e-01 ... 1.823e-01 -1.440e-01]
...
[ 5.589e-01 1.823e-01 ... 1.251e+00 -4.515e-01]
[-1.330e-01 -1.440e-01 ... -4.515e-01 3.689e+00]]
nfev: 18239
njev: 299
- Generation 0: best_so_far_y 1.65875e+01, min(y) 1.65875e+01 & Evaluations 16
- Generation 500: best_so_far_y 4.43191e-01, min(y) 4.44579e-01 & Evaluations 8016
- Generation 1000: best_so_far_y 3.63186e-01, min(y) 3.63186e-01 & Evaluations 16016
- Generation 1500: best_so_far_y 3.53268e-01, min(y) 3.53329e-01 & Evaluations 24016
- Generation 2000: best_so_far_y 3.05547e-01, min(y) 3.05707e-01 & Evaluations 32016
- Generation 2500: best_so_far_y 2.95432e-01, min(y) 2.95478e-01 & Evaluations 40016
- Generation 3000: best_so_far_y 2.79785e-01, min(y) 2.79901e-01 & Evaluations 48016
- Generation 3500: best_so_far_y 2.69350e-01, min(y) 2.69350e-01 & Evaluations 56016
- Generation 4000: best_so_far_y 2.65173e-01, min(y) 2.65265e-01 & Evaluations 64016
- Generation 4500: best_so_far_y 2.62260e-01, min(y) 2.62262e-01 & Evaluations 72016
- Generation 5000: best_so_far_y 2.61340e-01, min(y) 2.61340e-01 & Evaluations 80016
...............................
....... *** restart *** ....... - Generation 0: best_so_far_y 2.53208e-01, min(y) 1.40257e+01 & Evaluations 1939728
- Generation 500: best_so_far_y 2.47293e-01, min(y) 2.47293e-01 & Evaluations 1955728
- Generation 1000: best_so_far_y 2.16235e-01, min(y) 2.16235e-01 & Evaluations 1971728
- Generation 1500: best_so_far_y 2.15913e-01, min(y) 2.15913e-01 & Evaluations 1987728
- Generation 1637: best_so_far_y 2.15913e-01, min(y) 2.15913e-01 & Evaluations 1992080
....... *** restart *** ....... - Generation 0: best_so_far_y 2.15913e-01, min(y) 1.71281e+01 & Evaluations 1992144
- Generation 500: best_so_far_y 2.15913e-01, min(y) 3.12021e-01 & Evaluations 2024144
- Generation 1000: best_so_far_y 2.15913e-01, min(y) 2.74514e-01 & Evaluations 2056144
- Generation 1500: best_so_far_y 2.15913e-01, min(y) 2.35233e-01 & Evaluations 2088144
- Generation 2000: best_so_far_y 1.57412e-01, min(y) 1.57556e-01 & Evaluations 2120144
- Generation 2500: best_so_far_y 1.55672e-01, min(y) 1.55673e-01 & Evaluations 2152144
- Generation 3000: best_so_far_y 1.55385e-01, min(y) 1.55386e-01 & Evaluations 2184144
- Generation 3500: best_so_far_y 1.55307e-01, min(y) 1.55307e-01 & Evaluations 2216144
- Generation 4000: best_so_far_y 1.55283e-01, min(y) 1.55284e-01 & Evaluations 2248144
- Generation 4500: best_so_far_y 1.55274e-01, min(y) 1.55274e-01 & Evaluations 2280144
- Generation 5000: best_so_far_y 1.55269e-01, min(y) 1.55269e-01 & Evaluations 2312144
- Generation 5500: best_so_far_y 1.55267e-01, min(y) 1.55267e-01 & Evaluations 2344144
- Generation 6000: best_so_far_y 1.55266e-01, min(y) 1.55266e-01 & Evaluations 2376144
- Generation 6500: best_so_far_y 1.55265e-01, min(y) 1.55265e-01 & Evaluations 2408144
- Generation 7000: best_so_far_y 1.55265e-01, min(y) 1.55265e-01 & Evaluations 2440144
- Generation 7500: best_so_far_y 1.55264e-01, min(y) 1.55264e-01 & Evaluations 2472144
- Generation 8000: best_so_far_y 1.55264e-01, min(y) 1.55264e-01 & Evaluations 2504144
- Generation 8500: best_so_far_y 1.55264e-01, min(y) 1.55264e-01 & Evaluations 2536144
- Generation 9000: best_so_far_y 1.55264e-01, min(y) 1.55264e-01 & Evaluations 2568144
- Generation 9500: best_so_far_y 1.55264e-01, min(y) 1.55264e-01 & Evaluations 2600144
- Generation 10000: best_so_far_y 1.55264e-01, min(y) 1.55264e-01 & Evaluations 2632144
- Generation 10500: best_so_far_y 1.55264e-01, min(y) 1.55264e-01 & Evaluations 2664144
- Generation 11000: best_so_far_y 1.55264e-01, min(y) 1.55264e-01 & Evaluations 2696144
- Generation 11500: best_so_far_y 1.55264e-01, min(y) 1.55264e-01 & Evaluations 2728144
- Generation 12000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 2760144
- Generation 12500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 2792144
- Generation 13000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 2824144
- Generation 13500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 2856144
- Generation 14000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 2888144
- Generation 14500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 2920144
- Generation 15000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 2952144
- Generation 15500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 2984144
- Generation 16000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3016144
- Generation 16500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3048144
- Generation 17000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3080144
- Generation 17500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3112144
- Generation 18000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3144144
- Generation 18500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3176144
- Generation 19000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3208144
- Generation 19500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3240144
- Generation 20000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3272144
- Generation 20500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3304144
- Generation 21000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3336144
- Generation 21500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3368144
- Generation 22000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3400144
- Generation 22500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3432144
- Generation 23000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3464144
- Generation 23500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3496144
- Generation 24000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3528144
- Generation 24500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3560144
- Generation 25000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3592144
- Generation 25500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3624144
- Generation 26000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3656144
- Generation 26500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3688144
- Generation 27000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3720144
- Generation 27500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3752144
- Generation 28000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3784144
- Generation 28500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3816144
- Generation 29000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3848144
- Generation 29500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3880144
- Generation 30000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3912144
- Generation 30500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3944144
- Generation 31000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 3976144
- Generation 31500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4008144
- Generation 32000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4040144
- Generation 32500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4072144
- Generation 33000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4104144
- Generation 33500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4136144
- Generation 34000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4168144
- Generation 34500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4200144
- Generation 35000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4232144
- Generation 35500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4264144
- Generation 36000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4296144
- Generation 36500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4328144
- Generation 37000: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4360144
- Generation 37500: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4392144
- Generation 37780: best_so_far_y 1.55263e-01, min(y) 1.55263e-01 & Evaluations 4410000
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 1.66687e+01 & Evaluations 4410128
- Generation 500: best_so_far_y 1.55263e-01, min(y) 2.63335e-01 & Evaluations 4474128
- Generation 1000: best_so_far_y 1.55263e-01, min(y) 2.06230e-01 & Evaluations 4538128
- Generation 1345: best_so_far_y 1.55263e-01, min(y) 2.06176e-01 & Evaluations 4582160
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 1.29086e+01 & Evaluations 4582416
- Generation 500: best_so_far_y 1.55263e-01, min(y) 1.86254e-01 & Evaluations 4710416
- Generation 998: best_so_far_y 1.55263e-01, min(y) 1.81902e-01 & Evaluations 4837648
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 1.95505e+01 & Evaluations 4838160
- Generation 500: best_so_far_y 1.55263e-01, min(y) 2.00258e-01 & Evaluations 5094160
- Generation 1000: best_so_far_y 1.55263e-01, min(y) 1.77227e-01 & Evaluations 5350160
- Generation 1060: best_so_far_y 1.55263e-01, min(y) 1.77227e-01 & Evaluations 5380368
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 1.04358e+01 & Evaluations 5381392
- Generation 500: best_so_far_y 1.55263e-01, min(y) 2.43145e-01 & Evaluations 5893392
- Generation 1000: best_so_far_y 1.55263e-01, min(y) 1.84885e-01 & Evaluations 6405392
- Generation 1010: best_so_far_y 1.55263e-01, min(y) 1.84885e-01 & Evaluations 6414608
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 1.10192e+01 & Evaluations 6416656
- Generation 342: best_so_far_y 1.55263e-01, min(y) 6.37732e-01 & Evaluations 7115024
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 1.21271e+01 & Evaluations 7119120
- Generation 500: best_so_far_y 1.55263e-01, min(y) 6.20709e-01 & Evaluations 9167120
- Generation 1000: best_so_far_y 1.55263e-01, min(y) 2.58374e-01 & Evaluations 11215120
- Generation 1228: best_so_far_y 1.55263e-01, min(y) 2.51251e-01 & Evaluations 12144912
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 1.13383e+01 & Evaluations 12153104
- Generation 500: best_so_far_y 1.55263e-01, min(y) 3.86129e-01 & Evaluations 16249104
- Generation 861: best_so_far_y 1.55263e-01, min(y) 3.61371e-01 & Evaluations 19198224
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 8.38923e+00 & Evaluations 19214608
- Generation 500: best_so_far_y 1.55263e-01, min(y) 4.27539e-01 & Evaluations 27406608
- Generation 708: best_so_far_y 1.55263e-01, min(y) 4.22711e-01 & Evaluations 30798096
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 1.42631e+01 & Evaluations 30830864
- Generation 270: best_so_far_y 1.55263e-01, min(y) 6.60824e-01 & Evaluations 39645456
....... *** restart *** ....... - Generation 0: best_so_far_y 1.55263e-01, min(y) 9.53739e+00 & Evaluations 39710992
The results of multi-start optimization and FMAES are 0.1583149714373654 and 0.155263, respectively.
My system is based on python3.11.
from pypop.
FiksII
commented on May 18, 2024
Hmm... Strange. Could you send me numpy, scipy, pypop, numba versions. @Chang-SHAO
from pypop.
Chang-SHAO
commented on May 18, 2024
numpy==1.25.2
scipy==1.11.3
pypop7==0.0.72
numba==0.58.0
from pypop.
Chang-SHAO
commented on May 18, 2024
python==3.11.4
from pypop.
FiksII
commented on May 18, 2024
@Chang-SHAO
This is super weird.
I installed python 3.11.4. Windows 10 x64.
celluloid==0.2.0
colorama==0.4.6
contourpy==1.1.1
cycler==0.12.1
filelock==3.12.4
fonttools==4.43.1
fsspec==2023.9.2
imageio==2.31.5
Jinja2==3.1.2
joblib==1.3.2
kiwisolver==1.4.5
llvmlite==0.41.0
MarkupSafe==2.1.3
matplotlib==3.8.0
mpmath==1.3.0
networkx==3.1
numba==0.58.0
numpy==1.25.2
packaging==23.2
pandas==2.1.1
Pillow==10.0.1
pyparsing==3.1.1
pypop7==0.0.72
python-dateutil==2.8.2
pytz==2023.3.post1
scikit-learn==1.3.1
scipy==1.11.3
seaborn==0.13.0
six==1.16.0
sympy==1.12
threadpoolctl==3.2.0
torch==2.1.0
tqdm==4.66.1
typing_extensions==4.8.0
tzdata==2023.3
And have got the same results
"multi-start optimization is 0.161725642456162, FMAES - 0.172185"
if I use numpy generator, which might be more stable, I have got 0.1534:
message: Optimization terminated successfully.
success: True
status: 0
fun: 0.15348349998395383
x: [ 1.810e+00 -1.906e-02 ... -1.579e-01 -9.263e-01]
nit: 488
jac: [ 3.576e-07 -1.401e-05 ... 1.729e-06 -2.235e-08]
hess_inv: [[ 8.521e-01 1.285e-02 ... 3.749e-02 5.405e-01]
[ 1.285e-02 1.087e-03 ... -2.505e-05 3.619e-02]
...
[ 3.749e-02 -2.505e-05 ... 6.151e-01 -5.427e-02]
[ 5.405e-01 3.619e-02 ... -5.427e-02 3.545e+00]]
nfev: 33062
njev: 542
def multi_start_optimization(func, bounds, n_starts=10, method='BFGS', callback=None, seed=None, verbose=True,
**kwargs):
best_result = None
generator = numpy.random.default_rng(seed=seed)
for _ in (pbar := tqdm(range(n_starts), disable=not verbose)):
x0 = generator.random(*bounds[0].shape) * (bounds[1] - bounds[0]) + bounds[0]
result = scipy.optimize.minimize(func, x0, method=method, callback=callback, **kwargs)
if best_result is None or result.fun < best_result.fun:
best_result = result
pbar.set_postfix({'best_result_so_far': best_result.fun})
return best_result
from pypop.
Evolutionary-Intelligence
commented on May 18, 2024
@FiksII According to the official suggestion of NumPy, it is better to use its generator: https://numpy.org/doc/stable/reference/random/generator.html
from pypop.
FiksII
commented on May 18, 2024
@Evolutionary-Intelligence The problem is I can't achieve FMAES 0.155:
....... *** restart *** .......
* Generation 0: best_so_far_y 1.84766e-01, min(y) 1.28830e+01 & Evaluations 4225392
* Generation 500: best_so_far_y 1.84766e-01, min(y) 4.66329e-01 & Evaluations 5249392
* Generation 1000: best_so_far_y 1.72185e-01, min(y) 1.74094e-01 & Evaluations 6273392
* Generation 1087: best_so_far_y 1.72185e-01, min(y) 1.72500e-01 & Evaluations 6449520
....... *** restart *** .......
* Generation 0: best_so_far_y 1.72185e-01, min(y) 1.31143e+01 & Evaluations 6453616
* Generation 500: best_so_far_y 1.72185e-01, min(y) 3.64424e-01 & Evaluations 8501616
* Generation 780: best_so_far_y 1.72185e-01, min(y) 2.91337e-01 & Evaluations 9644400
....... *** restart *** .......
* Generation 0: best_so_far_y 1.72185e-01, min(y) 1.11503e+01 & Evaluations 9652592
* Generation 500: best_so_far_y 1.72185e-01, min(y) 3.37978e-01 & Evaluations 13748592
* Generation 685: best_so_far_y 1.72185e-01, min(y) 2.82461e-01 & Evaluations 15255920
....... *** restart *** .......
* Generation 0: best_so_far_y 1.72185e-01, min(y) 9.21559e+00 & Evaluations 15272304
* Generation 500: best_so_far_y 1.72185e-01, min(y) 4.49968e-01 & Evaluations 23464304
* Generation 717: best_so_far_y 1.72185e-01, min(y) 4.29491e-01 & Evaluations 27003248
....... *** restart *** .......
* Generation 0: best_so_far_y 1.72185e-01, min(y) 1.20239e+01 & Evaluations 27036016
* Generation 215: best_so_far_y 1.72185e-01, min(y) 6.52917e-01 & Evaluations 34048368
....... *** restart *** .......
* Generation 0: best_so_far_y 1.72185e-01, min(y) 8.29719e+00 & Evaluations 34113904
* Generation 500: best_so_far_y 1.72185e-01, min(y) 6.05617e-01 & Evaluations 66881904
* Generation 802: best_so_far_y 1.72185e-01, min(y) 5.39836e-01 & Evaluations 86608240
....... *** restart *** .......
* Generation 0: best_so_far_y 1.72185e-01, min(y) 1.50296e+01 & Evaluations 86739312
* Generation 102: best_so_far_y 1.72185e-01, min(y) 6.62570e-01 & Evaluations 100000000
from pypop.
Evolutionary-Intelligence
commented on May 18, 2024
For PyPop7, we always use this generator way to control the random number generation, if possible.
from pypop.
Evolutionary-Intelligence
commented on May 18, 2024
For BFGS, I obtained 0.148 on my personal computer:
100%|██████████████████████████████████████████████████████| 300/300 [04:24<00:00, 1.13it/s, best_result_so_far=0.148]
fun: 0.14828481985383515
hess_inv: array([[ 2.0961215 , 0.08509032, 0.18346933, ..., 0.16121667,
0.06686123, -0.43784459],
[ 0.08509032, 1.91309737, 0.75694706, ..., 0.05655528,
0.08857216, 0.31681217],
[ 0.18346933, 0.75694706, 0.69920324, ..., 0.05734176,
0.11169623, -0.15906565],
...,
[ 0.16121667, 0.05655528, 0.05734176, ..., 0.59763052,
-0.10901236, 0.11503432],
[ 0.06686123, 0.08857216, 0.11169623, ..., -0.10901236,
0.89061476, -0.24634906],
[-0.43784459, 0.31681217, -0.15906565, ..., 0.11503432,
-0.24634906, 4.17223822]])
jac: array([-8.75052065e-05, 3.37455422e-05, 2.41957605e-06, 2.34860927e-05,
-2.36947089e-05, 1.23474747e-05, 2.92062759e-06, -1.43740326e-05,
5.03454357e-05, 3.47476453e-05, -6.09885901e-05, 5.98505139e-05,
-1.31204724e-05, 3.48389149e-05, -1.76597387e-05, -2.08821148e-05,
-6.03646040e-05, 8.12429935e-05, 3.06293368e-05, -6.67702407e-05,
1.09653920e-05, 8.75815749e-06, -1.92467123e-05, 1.03972852e-05,
1.18333846e-05, -6.42407686e-05, 6.33951277e-05, 5.14611602e-05,
-8.22097063e-05, 4.25986946e-06, -2.04294920e-05, 3.18437815e-05,
-7.43083656e-05, -1.32974237e-05, -3.81935388e-05, -1.77603215e-05,
1.90306455e-05, 8.14720988e-06, 9.55574214e-05, 1.44578516e-05,
1.95633620e-05, 1.09635293e-05, -3.16649675e-07, -1.98744237e-05,
7.72066414e-06, -1.95614994e-05, -1.54785812e-05, 3.71038914e-06,
6.73905015e-06, -7.23451376e-06, 2.01575458e-05, 3.55895609e-05,
1.08387321e-05, -2.13719904e-05, 7.06315041e-05, -3.15345824e-06,
4.14438546e-06, 2.66432762e-05, 1.62962824e-05, 4.97139990e-06])
message: 'Optimization terminated successfully.'
nfev: 18361
nit: 282
njev: 301
status: 0
success: True
x: array([ 0.44604777, 1.57291812, 2.18368842, -0.21170684, 4.09407358,
0.95095942, 1.61002277, 0.93623952, 1.63431024, 4.04614979,
1.21756344, 0.70340479, 1.26349757, 0.65616069, 4.51423851,
0.90463141, 0.58072188, 1.0397659 , 1.81215946, 1.65111476,
1.66346183, -0.11780882, -0.03126143, 2.29387165, 3.90712599,
1.08005201, 0.92082673, 5.02446848, 5.12294812, 2.28257676,
0.52091617, 5.63332664, 5.58009323, 4.29602413, 3.57000464,
6.21840832, 6.77843958, 1.8425023 , 5.69346366, 5.03920492,
-0.33451392, -0.32414526, -0.40593616, -0.40431025, -0.3642978 ,
-0.439838 , -0.45072511, -0.40888369, -0.74354353, -0.29325016,
-0.48155137, -0.36315907, -0.50587948, -0.13964949, -0.51739519,
-0.2651726 , -0.4835695 , -0.12372969, -0.45856249, 0.38577874])
from pypop.
FiksII
commented on May 18, 2024
@Evolutionary-Intelligence I am completely confused
from pypop.
Related Issues (9)
- n_parents issue HOT 1
- Possible bug after restart of RMES HOT 1
- Optimizers don't take into account lower and upper boundaries HOT 1
- Feature request. Early stopping HOT 5
- Parllelization HOT 1
- How to get candidate solutions from each iteration? HOT 3
- Add Arxiv article resource link to readme HOT 2
- No module named 'pypop7.optimizers.bo' HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from pypop.