Gurobi taking over 3 days and no solution yet

Dejene Boru

unread,

Apr 4, 2017, 4:04:47 AM4/4/17

to Gurobi Optimization

Hi all,

I am using Gurobi 7.0 python API to solve MIP model. It has been running for 3 days and is not converging. I tried to set different tuning parameters but didn't help.

I would greatly appreciate any suggestions regarding this issue.

Here is a partial log of Gurobi

Changed value of parameter DisplayInterval to 60

Prev: 5 Min: 1 Max: 2000000000 Default: 5

Changed value of parameter MIPGap to 0.05

Prev: 0.0001 Min: 0.0 Max: 1e+100 Default: 0.0001

Changed value of parameter NumericFocus to 1

Prev: 0 Min: 0 Max: 3 Default: 0

Optimize a model with 15636 rows, 64248 columns and 241280 nonzeros

Model has 108 SOS constraints

Model has 88480 general constraints

Variable types: 32664 continuous, 31584 integer (31584 binary)

Coefficient statistics:

Matrix range [6e-09, 1e+02]

Objective range [4e-03, 5e+02]

Bounds range [1e+00, 1e+00]

RHS range [1e+00, 1e+02]

Presolve removed 38432 rows and 35956 columns

Presolve time: 23.06s

Presolved: 156612 rows, 177636 columns, 597370 nonzeros

Presolved model has 56224 SOS constraint(s)

Variable types: 61392 continuous, 116244 integer (90192 binary)

Root relaxation: objective 8.107765e+03, 43154 iterations, 7.21 seconds

Nodes | Current Node | Objective Bounds | Work

Expl Unexpl | Obj Depth IntInf | Incumbent BestBd Gap | It/Node Time

0 0 8107.76482 0 1560 - 8107.76482 - - 35s

0 0 8113.55917 0 2033 - 8113.55917 - - 43s

0 0 8122.67580 0 1892 - 8122.67580 - - 48s

0 0 8153.77305 0 3337 - 8153.77305 - - 72s

0 0 8154.17370 0 3558 - 8154.17370 - - 76s

0 0 8178.65739 0 4600 - 8178.65739 - - 108s

0 0 8179.66430 0 5335 - 8179.66430 - - 116s

0 0 8240.79427 0 5698 - 8240.79427 - - 163s

0 0 8244.59263 0 6109 - 8244.59263 - - 169s

0 0 8254.17490 0 6820 - 8254.17490 - - 187s

0 0 8254.68897 0 7203 - 8254.68897 - - 194s

0 0 8260.43990 0 8107 - 8260.43990 - - 219s

0 0 8261.19857 0 8359 - 8261.19857 - - 228s

0 0 8265.89965 0 9497 - 8265.89965 - - 261s

0 0 8266.11819 0 9843 - 8266.11819 - - 270s

0 0 8269.14022 0 10402 - 8269.14022 - - 297s

0 0 8269.95092 0 10506 - 8269.95092 - - 308s

0 0 8271.77155 0 11152 - 8271.77155 - - 332s

0 0 8271.93986 0 11241 - 8271.93986 - - 337s

0 0 8272.49449 0 11377 - 8272.49449 - - 356s

0 0 8272.55326 0 11461 - 8272.55326 - - 360s

0 0 8273.42833 0 11231 - 8273.42833 - - 371s

0 0 8273.66656 0 11190 - 8273.66656 - - 376s

0 0 8273.80475 0 11295 - 8273.80475 - - 390s

0 0 8273.82647 0 11360 - 8273.82647 - - 393s

0 0 8274.03401 0 11343 - 8274.03401 - - 398s

0 0 8274.04047 0 11354 - 8274.04047 - - 399s

0 0 8274.20784 0 11671 - 8274.20784 - - 405s

0 0 8274.21229 0 11622 - 8274.21229 - - 406s

0 0 8274.22937 0 11652 - 8274.22937 - - 409s

0 0 8274.24002 0 11742 - 8274.24002 - - 411s

0 0 8274.31095 0 12288 - 8274.31095 - - 425s

0 0 8274.31547 0 12354 - 8274.31547 - - 426s

0 0 8274.61183 0 12377 - 8274.61183 - - 432s

0 0 8274.61901 0 12441 - 8274.61901 - - 434s

0 0 8275.02957 0 12407 - 8275.02957 - - 439s

0 0 8275.03085 0 12452 - 8275.03085 - - 441s

0 0 8275.10496 0 12497 - 8275.10496 - - 445s

0 0 8275.10658 0 12601 - 8275.10658 - - 446s

0 0 8275.15883 0 12651 - 8275.15883 - - 451s

0 0 8275.15883 0 12641 - 8275.15883 - - 453s

0 2 8275.15883 0 12641 - 8275.15883 - - 468s

5 8 8329.42467 3 12002 - 8280.63857 - 844 487s

32 32 8501.18358 14 11134 - 8280.63857 - 520 563s

47 48 8915.30451 21 8555 - 8280.63857 - 583 609s

67 62 8911.90258 31 10169 - 8280.63857 - 625 666s

80 74 8913.26809 36 9839 - 8280.63857 - 680 726s

101 93 8914.55183 42 8799 - 8280.63857 - 704 783s

122 116 8920.28761 49 7427 - 8280.63857 - 756 848s

138 128 8923.57435 54 6081 - 8280.63857 - 803 918s

159 149 8928.17314 64 5908 - 8280.63857 - 852 971s

178 168 8936.50851 71 5285 - 8280.63857 - 887 1034s

210 193 8953.73128 84 4253 - 8280.63857 - 876 1099s

233 215 8946.69816 90 3822 - 8280.63857 - 857 1148s

.....

Regards,

Sonja Mars

unread,

Apr 4, 2017, 4:40:38 AM4/4/17

to gur...@googlegroups.com

Hi,

Can you post the last part of the log file (after three days)? 30 lines or so should be enough. What you just posted is the very beginning and in this beginning it looks like Gurobi has a hard time finding a feasible solution.

It would be interesting to see how this looks like after 3 days and how many nodes Gurobi has processed so far.

Thanks,
Sonja

----
Dr. Sonja Mars
Gurobi Optimization - Technical Support

Dejene Boru

unread,

Apr 4, 2017, 5:03:49 AM4/4/17

to Gurobi Optimization

Hi,

Unfortunately, I have stopped Gurobi and deleted the log for the 3 days yesterday but here is the last log for the one I started sometime yesterday.

Gurobi 7.0.2 (linux64, Python) logging started Mon Apr 3 10:55:23 2017

Changed value of parameter DisplayInterval to 60

Prev: 5 Min: 1 Max: 2000000000 Default: 5

Changed value of parameter MIPGap to 0.05

Prev: 0.0001 Min: 0.0 Max: 1e+100 Default: 0.0001

Optimize a model with 15636 rows, 64248 columns and 241280 nonzeros

Model has 108 SOS constraints

Model has 88480 general constraints

Variable types: 32664 continuous, 31584 integer (31584 binary)

Coefficient statistics:

Matrix range [6e-09, 1e+02]

Objective range [2e-03, 5e+02]

Bounds range [1e+00, 1e+00]

RHS range [1e+00, 1e+02]

Presolve removed 38432 rows and 35956 columns

Presolve time: 17.37s

Presolved: 156612 rows, 177636 columns, 597370 nonzeros

Presolved model has 56224 SOS constraint(s)

Variable types: 61392 continuous, 116244 integer (90192 binary)

Root relaxation: objective 7.982909e+03, 33418 iterations, 1.19 seconds

Nodes | Current Node | Objective Bounds | Work

Expl Unexpl | Obj Depth IntInf | Incumbent BestBd Gap | It/Node Time

0 0 7982.90923 0 1852 - 7982.90923 - - 23s

0 0 8000.38428 0 2338 - 8000.38428 - - 29s

0 0 8008.22594 0 2385 - 8008.22594 - - 33s

0 0 8032.43896 0 2479 - 8032.43896 - - 37s

0 0 8032.80036 0 2657 - 8032.80036 - - 38s

0 0 8038.96496 0 2789 - 8038.96496 - - 41s

0 0 8038.96546 0 2886 - 8038.96546 - - 42s

0 0 8042.09644 0 2598 - 8042.09644 - - 43s

0 0 8042.09679 0 2670 - 8042.09679 - - 44s

0 0 8042.54253 0 2774 - 8042.54253 - - 46s

0 0 8044.01477 0 2599 - 8044.01477 - - 47s

0 0 8046.74721 0 3019 - 8046.74721 - - 48s

0 0 8047.10696 0 3074 - 8047.10696 - - 49s

0 0 8052.65066 0 3457 - 8052.65066 - - 51s

0 0 8052.73240 0 3594 - 8052.73240 - - 52s

0 0 8053.57920 0 3287 - 8053.57920 - - 54s

0 0 8053.58973 0 3347 - 8053.58973 - - 55s

0 0 8053.93289 0 3960 - 8053.93289 - - 59s

0 0 8053.93603 0 4181 - 8053.93603 - - 60s

0 0 8054.25730 0 3985 - 8054.25730 - - 62s

0 0 8054.26703 0 4312 - 8054.26703 - - 63s

0 0 8054.66486 0 4718 - 8054.66486 - - 65s

0 0 8054.66654 0 4764 - 8054.66654 - - 66s

0 0 8055.20815 0 4880 - 8055.20815 - - 68s

0 0 8055.21022 0 4993 - 8055.21022 - - 69s

0 0 8055.64388 0 5632 - 8055.64388 - - 71s

0 0 8055.64713 0 5738 - 8055.64713 - - 72s

0 0 8056.27161 0 6892 - 8056.27161 - - 75s

0 0 8056.27181 0 6984 - 8056.27181 - - 76s

0 0 8056.74715 0 7162 - 8056.74715 - - 83s

0 0 8056.74779 0 7248 - 8056.74779 - - 85s

0 0 8057.26674 0 7541 - 8057.26674 - - 87s

0 0 8057.26847 0 7536 - 8057.26847 - - 89s

0 0 8057.59577 0 7254 - 8057.59577 - - 91s

:

...........

293972 270887 8590.03983 181 2240 - 8542.70752 - 478 74381s

294270 271180 8594.40031 219 1744 - 8542.70752 - 478 74459s

294418 271302 8595.41513 223 2042 - 8542.70752 - 478 74518s

294703 271607 8626.91169 253 2365 - 8542.70752 - 478 74573s

295029 271889 8638.40908 269 1698 - 8542.70752 - 478 74670s

295283 272132 8644.14032 285 1263 - 8542.70752 - 478 74726s

295588 272392 8650.95468 310 991 - 8542.70752 - 478 74785s

295912 272705 infeasible 338 - 8542.71219 - 478 74889s

296108 272886 8960.86849 81 2915 - 8542.71219 - 478 74945s

296789 273566 8971.47753 158 820 - 8542.71219 - 478 75069s

297448 274136 8971.92739 208 761 - 8542.71219 - 477 75172s

297806 274512 8973.56851 233 777 - 8542.71219 - 477 75222s

298151 274843 8973.47156 260 605 - 8542.71219 - 477 75300s

298397 275091 9000.05387 301 493 - 8542.71400 - 477 75377s

298557 275218 8548.26724 51 3165 - 8542.71400 - 478 75467s

298798 275441 8555.77668 100 1583 - 8542.71400 - 478 75524s

299064 275694 8561.54634 107 1955 - 8542.71400 - 478 75578s

299362 275941 8561.82770 116 1820 - 8542.71400 - 478 75658s

299530 276105 8562.79572 127 1546 - 8542.71400 - 478 75719s

299874 276385 8563.59882 154 1579 - 8542.71400 - 478 75798s

300089 276558 8564.00150 174 1567 - 8542.71400 - 478 75855s

300396 276834 8564.30911 203 1455 - 8542.71400 - 478 75910s

300655 277100 infeasible 225 - 8542.71632 - 478 75993s

300856 277284 8551.62670 74 3214 - 8542.71632 - 478 76045s

301148 277588 8561.15911 103 2927 - 8542.71632 - 478 76101s

301383 277780 8565.15660 120 2178 - 8542.71632 - 478 76154s

301648 278042 8570.29091 140 2033 - 8542.71632 - 478 76230s

301842 278242 8585.16251 146 2048 - 8542.71632 - 478 76286s

302146 278507 8576.63085 163 1838 - 8542.71632 - 478 76352s

302473 278812 8601.69800 185 1623 - 8542.71632 - 478 76423s

302795 279126 8601.56509 215 1495 - 8542.71632 - 478 76476s

303145 279496 8603.46754 235 1447 - 8542.71632 - 478 76527s

303515 279768 infeasible 260 - 8542.71922 - 477 76604s

303732 279972 8550.61544 72 2877 - 8542.71977 - 478 76655s

304066 280298 8976.01750 122 914 - 8542.71977 - 477 76743s

304283 280496 8976.87290 155 1331 - 8542.71977 - 478 76802s

304834 281047 9073.56967 201 1438 - 8542.71977 - 478 76936s

304940 281128 9027.18783 199 1399 - 8542.72117 - 478 77017s

305090 281301 8556.54592 77 2272 - 8542.72117 - 478 77071s

305454 281611 8566.26072 109 1225 - 8542.72117 - 478 77159s

305682 281816 8567.05787 148 1069 - 8542.72117 - 478 77209s

306024 282157 8567.37199 170 793 - 8542.72117 - 478 77257s

306346 282488 8579.65288 187 524 - 8542.72117 - 478 77336s

306581 282648 8567.73941 202 908 - 8542.72117 - 478 77387s

306892 282967 8571.34056 227 878 - 8542.72117 - 478 77442s

307173 283231 infeasible 260 - 8542.72274 - 478 77496s

307455 283509 8565.69434 87 1779 - 8542.72274 - 478 77594s

307680 283696 8567.81338 107 1983 - 8542.72274 - 478 77651s

307961 283932 8570.36934 131 1606 - 8542.72274 - 478 77740s

308174 284115 8569.35227 142 1335 - 8542.72274 - 478 77798s

308510 284437 8572.04358 182 1216 - 8542.72274 - 478 77853s

308873 284803 infeasible 227 - 8542.72438 - 478 77923s

309100 284985 8647.35714 48 4878 - 8542.72438 - 478 78007s

309265 285135 8659.78904 69 3347 - 8542.72438 - 478 78065s

309569 285435 8669.53769 106 1904 - 8542.72438 - 479 78136s

309826 285633 8674.39485 126 1435 - 8542.72438 - 479 78221s

310081 285886 8676.40791 142 1282 - 8542.72438 - 479 78277s

310445 286218 8680.63116 183 1057 - 8542.72438 - 478 78333s

310732 286523 8926.99500 214 693 - 8542.72438 - 479 78383s

311002 286741 8927.33834 254 408 - 8542.72438 - 479 78437s

311300 287045 infeasible 289 - 8542.72536 - 478 78529s

311471 287204 8556.81285 64 1768 - 8542.72572 - 479 78585s

311794 287495 8560.00757 96 2110 - 8542.72572 - 479 78643s

312152 287857 8568.95033 125 1814 - 8542.72572 - 479 78715s

312333 287966 8569.05357 132 1677 - 8542.72572 - 479 78792s

312607 288247 8667.62585 178 1452 - 8542.72572 - 479 78895s

312852 288472 9708.99103 194 1541 - 8542.72572 - 479 78950s

313136 288706 9725.87040 236 1824 - 8542.72572 - 479 79010s

313468 289044 infeasible 267 - 8542.72651 - 479 79093s

313634 289199 8546.59839 68 3530 - 8542.72783 - 479 79155s

313904 289440 8918.57310 92 3422 - 8542.72783 - 479 79231s

314125 289642 8938.36635 100 3580 - 8542.72783 - 479 79312s

314343 289899 8935.04287 111 3331 - 8542.72783 - 479 79370s

314606 290138 8945.34675 122 2874 - 8542.72783 - 479 79434s

314922 290430 8948.83747 136 2334 - 8542.72783 - 479 79488s

315288 290805 infeasible 166 - 8542.72860 - 479 79571s

315473 291000 8550.73227 66 3542 - 8542.72860 - 479 79631s

315818 291318 8564.50582 100 3029 - 8542.72860 - 479 79694s

316085 291591 8569.81683 114 2389 - 8542.72860 - 479 79788s

316264 291786 8598.47131 129 2145 - 8542.72860 - 479 79878s

316458 291936 8623.61139 132 1948 - 8542.72860 - 480 79943s

316758 292259 8672.37599 143 2249 - 8542.72860 - 480 80031s

317000 292483 8913.11029 180 1270 - 8542.72860 - 480 80093s

317371 292779 8917.93379 209 1168 - 8542.72860 - 480 80196s

317634 293087 8929.26360 242 966 - 8542.72860 - 480 80261s

Thanks,

Sonja Mars

unread,

Apr 4, 2017, 5:07:23 AM4/4/17

to gur...@googlegroups.com

Are you sure there is a feasible solution to your problem?

Or even better, do you know a feasible solution that you can hand over to Gurobi as MIP start?

Dejene Boru

unread,

Apr 4, 2017, 5:20:28 AM4/4/17

to Gurobi Optimization

The problem is about a network optimization when I run the model with few number of nodes e.g. 10, I am able to find a solution, well even if it is difficult to generalize I assume there is a feasible solution. I haven't tried with MIP start, but I will try that, I guess I don't need to provide solution for all decision variables.

regards,

Sonja Mars

unread,

Apr 4, 2017, 5:27:26 AM4/4/17

to gur...@googlegroups.com

Hi,

> The problem is about a network optimization when I run the model with few number of nodes e.g. 10, I am able to find a solution, well even if it is difficult to generalize I assume there is a feasible solution. I haven't tried with MIP start, but I will try that, I guess I don't need to provide solution for all decision variables.

Yes, it is fine to just hand over solution values for a subset of variables. However, if you specify not enough variables and it would take Gurobi too long to "repair" the solution, it will not accept it.

One other possibility is to use the parameters PreSOS1BigM and PreSOS2BigM to transform the SOS-constraints to linear ones, then it is sometimes easier for Gurobi to find an initial solution.

Details on these parameter can be found here:
http://www.gurobi.com/documentation/7.0/refman/presos1bigm.html#parameter:PreSOS1BigM
http://www.gurobi.com/documentation/7.0/refman/presos2bigm.html#parameter:PreSOS2BigM

Best regards,
Sonja

Dejene Boru

unread,

Apr 4, 2017, 11:39:31 AM4/4/17

to Gurobi Optimization

Hi Sonja,

Thank you for your reply on my previous questions, the MIP start has helped Gurobi a lot but it is is still taking a long time to converge, do you have any recommendation?

Here is the log from Gurobi,

Changed value of parameter PreSOS2BigM to 10000.0

Prev: 0.0 Min: -1.0 Max: 10000000000.0 Default: 0.0

Changed value of parameter MIPGap to 0.05

Prev: 0.0001 Min: 0.0 Max: 1e+100 Default: 0.0001

Optimize a model with 15636 rows, 64248 columns and 241280 nonzeros

Model has 108 SOS constraints

Model has 88480 general constraints

Variable types: 32664 continuous, 31584 integer (31584 binary)

Coefficient statistics:

Matrix range [6e-09, 1e+02]

Objective range [6e-03, 5e+02]

Bounds range [1e+00, 1e+00]

RHS range [1e+00, 1e+02]

Presolve removed 37972 rows and 35128 columns (presolve time = 15s) ...

Presolve removed 38432 rows and 35956 columns (presolve time = 25s) ...

Presolve removed 38432 rows and 35956 columns

Presolve time: 25.36s

Presolved: 157692 rows, 178500 columns, 600934 nonzeros

Presolved model has 56116 SOS constraint(s)

Loaded MIP start with objective 21210.8

Variable types: 61392 continuous, 117108 integer (91056 binary)

Root simplex log...

Iteration Objective Primal Inf. Dual Inf. Time

0 6.0527909e+03 3.042750e+04 0.000000e+00 29s

34114 8.1443956e+03 3.217508e+04 0.000000e+00 30s

44645 8.4473636e+03 1.774180e+03 0.000000e+00 35s

50801 8.6175932e+03 5.373689e+01 0.000000e+00 40s

51490 8.6178861e+03 0.000000e+00 0.000000e+00 41s

Root relaxation: objective 8.617886e+03, 51490 iterations, 12.42 seconds

Nodes | Current Node | Objective Bounds | Work

Expl Unexpl | Obj Depth IntInf | Incumbent BestBd Gap | It/Node Time

0 0 8617.88615 0 2093 21210.8027 8617.88615 59.4% - 45s

0 0 8638.34732 0 3142 21210.8027 8638.34732 59.3% - 62s

0 0 8644.75698 0 4017 21210.8027 8644.75698 59.2% - 89s

0 0 8657.91399 0 5163 21210.8027 8657.91399 59.2% - 138s

0 0 8659.43507 0 5555 21210.8027 8659.43507 59.2% - 153s

0 0 8669.93472 0 6441 21210.8027 8669.93472 59.1% - 185s

0 0 8672.28902 0 7018 21210.8027 8672.28902 59.1% - 202s

......

769 616 9842.46610 161 6130 11650.9353 8944.66431 23.2% 2188 8884s

770 616 9171.44502 47 6137 11650.9353 8945.26542 23.2% 2186 8894s

771 617 9557.53826 123 6151 11650.9353 8946.66203 23.2% 2183 8909s

772 618 9012.39474 21 6263 11650.9353 8947.62216 23.2% 2180 8920s

773 618 9139.15992 13 6032 11650.9353 8950.82777 23.2% 2177 8937s

774 619 9920.00386 198 6207 11650.9353 8951.23616 23.2% 2174 8948s

775 620 9002.31431 22 6310 11650.9353 8952.54318 23.2% 2171 8965s

776 620 9468.31056 73 6155 11650.9353 8952.80287 23.2% 2169 8974s

777 621 9363.09186 44 6103 11650.9353 8953.24148 23.2% 2166 8985s

778 622 8953.39706 16 6144 11650.9353 8953.39706 23.2% 2163 8994s

779 622 9177.61410 46 6188 11650.9353 8953.54835 23.2% 2160 9025s

780 623 9339.18452 36 6188 11650.9353 8953.54835 23.2% 2158 9032s

781 624 9482.97951 104 6188 11650.9353 8953.54835 23.2% 2155 9064s

782 627 8954.36439 13 7739 11650.9353 8953.55809 23.2% 2345 9133s

784 629 8954.45053 14 7727 11650.9353 8954.36439 23.1% 2343 9179s

786 630 8959.00757 14 7336 11650.9353 8954.45053 23.1% 2346 9229s

788 631 8955.36143 15 7806 11650.9353 8954.45053 23.1% 2349 9261s

790 633 8956.48332 15 7812 11650.9353 8955.36143 23.1% 2349 9293s

792 634 8956.12769 16 7790 11650.9353 8955.36143 23.1% 2346 9308s

794 635 8956.25384 16 7739 11650.9353 8956.13022 23.1% 2342 9327s

796 633 8964.32030 17 7367 11650.9353 8956.13022 23.1% 2344 9352s

798 638 8957.51201 17 7170 11650.9353 8956.33181 23.1% 2345 9369s

800 637 8959.27510 18 7107 11650.9353 8956.33181 23.1% 2342 9456s

804 640 8959.29922 19 7095 11650.9353 8956.33181 23.1% 2366 9479s

808 641 8962.76149 20 7132 11650.9353 8956.33181 23.1% 2364 9513s

810 646 8961.24526 20 6838 11650.9353 8956.33181 23.1% 2363 9529s

812 643 8992.31507 21 7256 11650.9353 8956.33181 23.1% 2368 9590s

814 648 8968.19358 21 6993 11650.9353 8956.33181 23.1% 2381 9606s

817 649 9043.39661 22 7215 11650.9353 8956.33181 23.1% 2389 9665s

819 650 8968.52216 22 6599 11650.9353 8956.33181 23.1% 2396 9723s

823 653 8978.20358 23 6678 11650.9353 8956.33181 23.1% 2403 9761s

827 654 8979.07050 24 6653 11650.9353 8956.33181 23.1% 2403 9802s

832 659 8980.56342 26 6731 11650.9353 8956.33181 23.1% 2402 9834s

838 665 9293.20364 27 6270 11650.9353 8956.33181 23.1% 2420 9979s

840 664 9004.07802 28 6448 11650.9353 8956.33181 23.1% 2417 10024s

844 667 9004.57735 29 6387 11650.9353 8956.33181 23.1% 2423 10142s

848 666 9015.97190 30 6443 11650.9353 8956.33181 23.1% 2445 10227s

851 669 9016.23761 31 6337 11650.9353 8956.33181 23.1% 2442 10324s

855 675 9019.94073 32 6012 11650.9353 8956.33181 23.1% 2442 10362s

862 675 9029.92573 35 5934 11650.9353 8956.33181 23.1% 2436 10507s

872 683 9026.40403 37 5365 11650.9353 8956.33181 23.1% 2421 10554s

880 690 9024.10380 39 6042 11650.9353 8956.33181 23.1% 2412 10609s

891 696 9028.22060 42 5631 11650.9353 8956.33181 23.1% 2401 10677s

901 704 9040.48208 44 5425 11650.9353 8956.33181 23.1% 2397 10757s

906 705 9040.44816 46 5459 11650.9353 8956.33181 23.1% 2395 10809s

913 711 9040.61767 47 5744 11650.9353 8956.33181 23.1% 2395 10857s

921 717 9046.41437 49 5494 11650.9353 8956.33181 23.1% 2392 10917s

936 724 9047.64890 53 5649 11650.9353 8956.33181 23.1% 2372 11042s

938 728 9041.57167 53 5675 11650.9353 8956.33181 23.1% 2368 11102s

949 735 9048.07200 55 5030 11650.9353 8956.33181 23.1% 2363 11165s

957 741 9045.47787 58 5423 11650.9353 8956.33181 23.1% 2366 11228s

972 748 9043.27301 61 5614 11650.9353 8956.33181 23.1% 2347 11296s

988 762 9049.68360 65 5147 11650.9353 8956.33181 23.1% 2332 11362s

1001 768 9044.20286 68 5143 11650.9353 8956.33181 23.1% 2321 11448s

1014 777 9044.89228 71 5069 11650.9353 8956.33181 23.1% 2317 11525s

1029 786 9046.69965 75 4965 11650.9353 8956.33181 23.1% 2310 11606s

1047 797 9055.72570 80 4493 11650.9353 8956.33181 23.1% 2298 11739s

1062 805 9061.11622 85 4288 11650.9353 8956.33181 23.1% 2292 11833s

1083 818 9068.93643 90 3900 11650.9353 8956.33181 23.1% 2281 11921s

1101 827 9092.18335 97 4215 11650.9353 8956.33181 23.1% 2278 12005s

1122 839 9101.83457 102 4063 11650.9353 8956.33181 23.1% 2269 12108s

1153 861 9109.39760 111 3342 11650.9353 8956.33181 23.1% 2241 12239s

1181 879 9112.95640 117 3162 11650.9353 8956.33181 23.1% 2218 12351s

1213 902 9126.78771 125 3073 11650.9353 8956.33181 23.1% 2196 12484s

1246 922 9143.14411 136 3125 11650.9353 8956.33181 23.1% 2175 12621s

1289 948 9166.13274 152 3282 11650.9353 8956.33181 23.1% 2140 12862s

1336 966 9211.70434 163 2933 11650.9353 8956.33181 23.1% 2102 13043s

1375 988 9268.27884 173 2741 11650.9353 8956.33181 23.1% 2089 13209s

1394 987 9273.21943 179 2549 11650.9353 8956.33181 23.1% 2107 13377s

1424 1000 9281.38970 193 2483 11650.9353 8956.33181 23.1% 2087 13596s

Thanks,

Sonja Mars

unread,

Apr 4, 2017, 1:29:23 PM4/4/17

to gur...@googlegroups.com

Hi,

This looks much better. You might want to take a look at this section of the Gurobi documentation:
http://www.gurobi.com/documentation/7.0/refman/parameter_guidelines.html#sec:ParameterGuidelines

One thing to note is this:

> Coefficient statistics:
> Matrix range [6e-09, 1e+02]
> Objective range [6e-03, 5e+02]
> Bounds range [1e+00, 1e+00]
> RHS range [1e+00, 1e+02]

This is a pretty large range in the matrix coefficients. This can cause a lot of trouble inside the solver (http://www.gurobi.com/documentation/7.0/refman/tolerances_and_ill_conditi.html). You should try to reduce this range in your matrix coefficients. What are these small number (6e-9) standing for? Are they really needed? Can you rescale your model?

Best regards,

J. Friedman

unread,

Apr 4, 2017, 9:06:42 PM4/4/17

to Gurobi Optimization

Try tuning a small instance first.

Dejene Boru

unread,

Apr 5, 2017, 4:36:05 AM4/5/17

to Gurobi Optimization

Thank you Sonja, Friedman,

Your suggestions were very useful.

Regards,

On Tuesday, April 4, 2017 at 10:04:47 AM UTC+2, Dejene Boru wrote:

Siddharth Swaminathan

unread,

Feb 16, 2018, 3:52:36 PM2/16/18

to Gurobi Optimization

Dear Sonja,

I am having a similar problem and was hoping you would explain a little more on how exactly i could reduce the range in the matrix coefficients? Are there some tricks that i can use to find this, in general? Can you provide an example on how to rescale the matrix?

Thanks very much,

Sid

Sonja Mars

unread,

Feb 19, 2018, 6:42:54 AM2/19/18

to 'Renan Garcia' via Gurobi Optimization

Hi Sid,

You might find this here interesting:
http://files.gurobi.com/Numerics.pdf

Siddharth Swaminathan

unread,

Feb 19, 2018, 11:58:25 AM2/19/18

to Gurobi Optimization

Thank you very much Sonja

Reply all

Reply to author

Forward