Gurobi only exploring 1 node in 20k seconds
進行中We have a problem with Gurobi (on a fairly large model with 80k variables and constraints) that it spends a lot of time to only explore one node. A brief version of the output is included below.
Is there any way to solve this?
Solved with dual simplex
Root relaxation: objective 6.238698e+02, 248351 iterations, 842.03 seconds
Total elapsed time = 864.44s
Total elapsed time = 870.48s
Total elapsed time = 879.21s
Total elapsed time = 880.00s
Nodes | Current Node | Objective Bounds | Work
Expl Unexpl | Obj Depth IntInf | Incumbent BestBd Gap | It/Node Time
0 0 623.86978 0 22248 - 623.86978 - - 880s
0 0 569.79107 0 20535 - 569.79107 - - 991s
0 0 569.79107 0 20264 - 569.79107 - - 1002s
0 0 545.26054 0 19852 - 545.26054 - - 1150s
0 0 544.98032 0 20102 - 544.98032 - - 1277s
0 0 544.97399 0 19867 - 544.97399 - - 1290s
0 0 540.96068 0 18388 - 540.96068 - - 1595s
0 0 538.62803 0 18347 - 538.62803 - - 1876s
0 0 538.03362 0 18751 - 538.03362 - - 1956s
0 0 538.02926 0 18572 - 538.02926 - - 1995s
0 0 526.66031 0 16899 - 526.66031 - - 2705s
0 0 510.56976 0 16030 - 510.56976 - - 3002s
0 0 492.96287 0 15414 - 492.96287 - - 3576s
0 0 492.73198 0 15324 - 492.73198 - - 3730s
0 0 492.64599 0 15472 - 492.64599 - - 3789s
0 0 492.64144 0 15708 - 492.64144 - - 3820s
0 0 477.11857 0 14722 - 477.11857 - - 4406s
0 0 475.68420 0 14774 - 475.68420 - - 4595s
0 0 475.64830 0 15137 - 475.64830 - - 4665s
0 0 457.98450 0 13285 - 457.98450 - - 5250s
0 0 456.09615 0 13441 - 456.09615 - - 5462s
0 0 455.89813 0 13555 - 455.89813 - - 5574s
0 0 455.88711 0 13478 - 455.88711 - - 5606s
0 0 440.24472 0 13042 - 440.24472 - - 6244s
0 0 438.93436 0 12982 - 438.93436 - - 6444s
0 0 438.90477 0 13128 - 438.90477 - - 6512s
0 0 430.46668 0 12469 - 430.46668 - - 7657s
0 0 429.97314 0 12361 - 429.97314 - - 7848s
0 0 429.94669 0 12463 - 429.94669 - - 7933s
0 0 421.76719 0 12138 - 421.76719 - - 8523s
0 0 420.68969 0 12558 - 420.68969 - - 8671s
0 0 420.67835 0 12589 - 420.67835 - - 8730s
0 0 415.45638 0 11892 - 415.45638 - - 9337s
0 0 415.42708 0 11963 - 415.42708 - - 9400s
0 0 410.39141 0 11508 - 410.39141 - - 9935s
0 0 410.37788 0 11613 - 410.37788 - - 10008s
0 0 409.14276 0 11336 - 409.14276 - - 10426s
0 0 409.14276 0 11362 - 409.14276 - - 10473s
0 0 407.88464 0 11449 - 407.88464 - - 10829s
0 0 407.88301 0 11492 - 407.88301 - - 10870s
0 0 407.88093 0 11450 - 407.88093 - - 11260s
H 0 0 -673.0000000 407.88093 161% - 11260s
0 0 407.88093 0 4413 -673.00000 407.88093 161% - 12542s
H 0 0 -207.0000000 407.88093 297% - 13335s
H 0 0 -186.0000000 407.88093 319% - 13379s
---- PROBLEM HERE ----
0 2 407.88093 0 3660 -186.00000 407.88093 319% - 37534s
---- PROBLEM HERE ----
Cutting planes:
Gomory: 12
Implied bound: 3
Clique: 14
MIR: 236
StrongCG: 4
GUB cover: 23
Zero half: 264
Mod-K: 12
RLT: 1734
Explored 1 nodes (2660824 simplex iterations) in 37535.90 seconds
Thread count was 6 (of 6 available processors)
Solution count 3: -186 -207 -673
Time limit reached
Best objective -1.860000000000e+02, best bound 4.070000000000e+02, gap 318.8172%"
0
-
正式なコメント
-
Gurobi Staff
This post is more than three years old. Some information may not be up to date. For current information, please check the Gurobi Documentation or Knowledge Base. If you need more help, please create a new post in the community forum. Or why not try our AI Gurobot?. -
-
-
Hi Mathias,
If there are many lines in the root node this may be an indicator that many cuts are added. This can be very good or bad. You could try experimenting with the Cuts parameter, e.g., set it to 1 or 0, and the CutPasses parameter, e.g., set it to 3. If setting the Cuts parameter to 0 helps, you could try disabling the different cut parameters such as, e.g., MIRCuts or RLTCuts, one by one to find out, which affects the performance the most.
Other than that, you could have a look at our documentation on most important parameters for MIPs.
80k variables is a lot and it is very well possible that your problem is more complex than you expect. In case that you are using penalty coefficients of bigM coefficients, you could try decreasing the order of magnitude of these particular coefficients in order to improve performance.
Best regards,
Jaromił0 -
-
Hi Jaromil,
Thank you very much for the advice. We will try to experiment with the cuts parameters. We were just looking for whether there was a quick fix or something we were missing to avoid spending days on experimenting with the parameters.
We have also begun experimenting with the MIPFocus parameter, as we are primarly looking for a feasible solution. This seems to lead to quicker branching, which might be better for performance.
But thanks again.
Best,
Mathias0 -
-
-
Hi Mathias,
Could you please also share the beginning of the log, showing the problem statistics? It is also possible that Gurobi is running some heuristics at the end of the root node, consuming quite some time. Reducing the Heuristics parameter to some small value, e.g., 0.0001 could also help. You may also be able to speed up the root by limiting some expensive heuristics via SubMIPNodes=20.
Best regards,
Jaromił0 -
-
Yes, i include the full log below:
Academic license - for non-commercial use only
Gurobi Optimizer version 9.0.1 build v9.0.1rc0 (win64)
Optimize a model with 718985 rows, 678369 columns and 3309084 nonzeros
Model fingerprint: 0xfbe528e6
Variable types: 0 continuous, 678369 integer (678369 binary)
Coefficient statistics:
Matrix range [1e+00, 1e+00]
Objective range [1e+00, 1e+00]
Bounds range [0e+00, 0e+00]
RHS range [1e+00, 2e+01]
Presolve removed 646852 rows and 597851 columns
Presolve time: 2.99s
Presolved: 72133 rows, 80518 columns, 352714 nonzeros
Variable types: 0 continuous, 80518 integer (80518 binary)
Deterministic concurrent LP optimizer: primal and dual simplex
Showing first log only...
Root simplex log...
Iteration Objective Primal Inf. Dual Inf. Time
20253 4.2094796e+02 5.876119e+02 6.275758e+10 5s
34907 4.2925209e+02 1.118190e+02 9.550946e+10 10s
45348 4.3762816e+02 3.595882e+01 7.746616e+10 15s
57558 4.7350959e+02 9.062670e+00 2.363636e+10 20s
66578 4.9523453e+02 1.339173e+00 1.042161e+10 25s
68799 -5.3581991e+02 0.000000e+00 5.160278e+04 27s
72111 -4.5513272e+02 0.000000e+00 7.220870e+05 30s
76088 -4.0192826e+02 0.000000e+00 1.047694e+05 35s
79804 -3.6643545e+02 0.000000e+00 2.781754e+05 40s
83099 -3.3414857e+02 0.000000e+00 2.335032e+06 45s
86002 -2.8883996e+02 0.000000e+00 2.473825e+05 50s
88951 -2.3538792e+02 0.000000e+00 3.374852e+05 55s
91747 -1.8594906e+02 0.000000e+00 3.294007e+05 60s
94297 -1.6464857e+02 0.000000e+00 8.627091e+04 65s
96936 -7.0286498e+01 0.000000e+00 3.452854e+05 70s
99719 1.7690827e+01 0.000000e+00 5.968125e+05 75s
102263 6.2367100e+01 0.000000e+00 1.247246e+06 80s
104823 7.8886104e+01 0.000000e+00 1.705640e+05 85s
107573 1.0065501e+02 0.000000e+00 3.411252e+05 90s
110163 1.2464489e+02 0.000000e+00 1.305435e+07 95s
112603 1.4042794e+02 0.000000e+00 1.304715e+06 100s
115143 1.5949461e+02 0.000000e+00 1.588078e+06 105s
117474 1.9219646e+02 0.000000e+00 1.361639e+06 110s
119817 2.1814203e+02 0.000000e+00 8.945738e+06 115s
121999 2.3952038e+02 0.000000e+00 3.407978e+05 120s
124277 2.4759869e+02 0.000000e+00 1.120571e+06 125s
126432 2.7042088e+02 0.000000e+00 2.475405e+05 130s
128556 3.2322127e+02 0.000000e+00 3.224453e+06 135s
130760 3.5821740e+02 0.000000e+00 2.882620e+05 140s
132881 3.8308035e+02 0.000000e+00 1.817436e+05 145s
135182 4.2547184e+02 0.000000e+00 1.946961e+06 150s
137252 4.3121073e+02 0.000000e+00 5.968863e+05 155s
139460 4.3840812e+02 0.000000e+00 5.336541e+05 160s
141484 4.4314967e+02 0.000000e+00 1.565474e+06 165s
143548 4.4734188e+02 0.000000e+00 4.855344e+05 170s
145598 4.5266277e+02 0.000000e+00 4.873857e+05 175s
147680 4.5721739e+02 0.000000e+00 3.687177e+06 180s
149680 4.6106889e+02 0.000000e+00 2.820338e+05 185s
151680 4.6515576e+02 0.000000e+00 2.493814e+05 190s
153660 4.6960317e+02 0.000000e+00 8.374698e+05 195s
155610 4.7363962e+02 0.000000e+00 1.070394e+06 200s
157480 4.7721936e+02 0.000000e+00 4.273164e+05 205s
159380 4.8101307e+02 0.000000e+00 1.023236e+06 210s
161310 4.8469695e+02 0.000000e+00 5.739401e+05 215s
163160 4.8761820e+02 0.000000e+00 3.184773e+05 220s
165110 4.9084183e+02 0.000000e+00 2.270792e+06 225s
167080 4.9506340e+02 0.000000e+00 4.167684e+05 230s
169030 4.9760563e+02 0.000000e+00 1.186701e+05 235s
170950 5.0004097e+02 0.000000e+00 2.779092e+06 240s
172790 5.0254101e+02 0.000000e+00 1.441614e+05 245s
174690 5.0518131e+02 0.000000e+00 1.206782e+07 250s
176598 5.0733292e+02 0.000000e+00 3.040987e+05 255s
178268 5.0923837e+02 0.000000e+00 9.867606e+04 260s
180108 5.1119148e+02 0.000000e+00 9.541611e+05 265s
181898 5.1309241e+02 0.000000e+00 2.391071e+05 270s
183568 5.1490530e+02 0.000000e+00 8.164104e+05 275s
185452 5.1661052e+02 0.000000e+00 3.127484e+05 280s
187210 5.1813283e+02 0.000000e+00 1.711584e+06 285s
188952 5.2000310e+02 0.000000e+00 6.612509e+05 290s
190790 5.2144770e+02 0.000000e+00 1.173798e+06 295s
192648 5.2316312e+02 0.000000e+00 1.045847e+06 300s
194338 5.2493790e+02 0.000000e+00 9.828231e+04 305s
196124 5.2649167e+02 0.000000e+00 8.867489e+05 310s
197972 5.2782762e+02 0.000000e+00 1.854282e+05 315s
199820 5.2909180e+02 0.000000e+00 1.436088e+05 320s
201514 5.3080776e+02 0.000000e+00 1.691498e+05 325s
203268 5.3246612e+02 0.000000e+00 2.906658e+05 330s
205024 5.3370707e+02 0.000000e+00 3.707665e+05 335s
206792 5.3520820e+02 0.000000e+00 3.674470e+05 340s
208644 5.3677242e+02 0.000000e+00 6.218818e+05 345s
210490 5.3855626e+02 0.000000e+00 9.653889e+05 350s
212338 5.3996337e+02 0.000000e+00 1.722866e+05 355s
213998 5.4123823e+02 0.000000e+00 1.021819e+05 360s
215754 5.4230343e+02 0.000000e+00 1.968235e+05 365s
217598 5.4346487e+02 0.000000e+00 2.959266e+05 370s
219428 5.4467723e+02 0.000000e+00 5.812115e+05 375s
221138 5.4591410e+02 0.000000e+00 1.325242e+05 380s
222936 5.4685941e+02 0.000000e+00 2.687900e+05 385s
224714 5.4807842e+02 0.000000e+00 7.136551e+05 390s
226566 5.4911123e+02 0.000000e+00 1.594995e+07 395s
228114 5.4995328e+02 0.000000e+00 3.566287e+05 400s
229926 5.5110799e+02 0.000000e+00 1.424347e+05 405s
231650 5.5200193e+02 0.000000e+00 5.953129e+04 410s
233272 5.5282125e+02 0.000000e+00 7.732538e+04 415s
235082 5.5397125e+02 0.000000e+00 1.942008e+05 420s
236802 5.5503602e+02 0.000000e+00 1.418577e+06 425s
238542 5.5599961e+02 0.000000e+00 4.913357e+05 430s
240302 5.5691773e+02 0.000000e+00 8.641043e+05 435s
242062 5.5771143e+02 0.000000e+00 4.389408e+05 440s
243862 5.5857961e+02 0.000000e+00 2.677979e+06 445s
245698 5.5942849e+02 0.000000e+00 7.474124e+04 450s
247458 5.6021666e+02 0.000000e+00 5.302288e+05 455s
249208 5.6105894e+02 0.000000e+00 1.346723e+05 460s
251014 5.6182250e+02 0.000000e+00 9.507893e+05 465s
252704 5.6245657e+02 0.000000e+00 1.776386e+05 470s
254334 5.6321195e+02 0.000000e+00 6.665055e+05 475s
256138 5.6400859e+02 0.000000e+00 3.955814e+06 480s
257820 5.6469430e+02 0.000000e+00 1.000901e+06 485s
259650 5.6531577e+02 0.000000e+00 4.742576e+05 490s
261268 5.6613276e+02 0.000000e+00 7.139943e+05 495s
263016 5.6688144e+02 0.000000e+00 1.276941e+05 500s
264872 5.6751422e+02 0.000000e+00 1.213123e+05 505s
266572 5.6805800e+02 0.000000e+00 4.344407e+05 510s
268318 5.6871944e+02 0.000000e+00 1.527038e+05 515s
270038 5.6940287e+02 0.000000e+00 6.272162e+04 520s
271666 5.6995613e+02 0.000000e+00 5.079223e+04 525s
273486 5.7067581e+02 0.000000e+00 3.276376e+05 530s
275284 5.7133913e+02 0.000000e+00 6.427237e+05 535s
276958 5.7194245e+02 0.000000e+00 1.424871e+05 540s
278700 5.7259341e+02 0.000000e+00 2.780658e+05 545s
280484 5.7320425e+02 0.000000e+00 3.303295e+05 550s
282314 5.7393379e+02 0.000000e+00 1.343835e+05 555s
284164 5.7450397e+02 0.000000e+00 1.379941e+05 560s
285844 5.7518889e+02 0.000000e+00 3.739063e+05 565s
287554 5.7596520e+02 0.000000e+00 1.070088e+06 570s
289214 5.7653365e+02 0.000000e+00 4.831411e+05 575s
290874 5.7707883e+02 0.000000e+00 8.054500e+05 580s
292684 5.7770460e+02 0.000000e+00 2.291337e+05 585s
294330 5.7828715e+02 0.000000e+00 1.335929e+05 590s
296106 5.7900764e+02 0.000000e+00 3.999204e+04 595s
297876 5.7952941e+02 0.000000e+00 1.245554e+07 600s
299612 5.8004635e+02 0.000000e+00 3.938890e+04 605s
301312 5.8049769e+02 0.000000e+00 2.924931e+06 610s
303022 5.8109036e+02 0.000000e+00 5.416931e+04 615s
304686 5.8152568e+02 0.000000e+00 7.458792e+05 620s
306374 5.8210978e+02 0.000000e+00 9.121043e+04 625s
308066 5.8254425e+02 0.000000e+00 9.092113e+04 630s
309870 5.8294681e+02 0.000000e+00 7.389353e+05 635s
311492 5.8336267e+02 0.000000e+00 1.049034e+05 640s
313186 5.8383100e+02 0.000000e+00 3.767129e+05 645s
314880 5.8435441e+02 0.000000e+00 1.659852e+05 650s
316480 5.8480651e+02 0.000000e+00 9.803162e+05 655s
318160 5.8520898e+02 0.000000e+00 1.396710e+06 660s
319900 5.8562086e+02 0.000000e+00 1.610926e+05 665s
321500 5.8613457e+02 0.000000e+00 1.560482e+05 670s
323250 5.8659509e+02 0.000000e+00 5.178975e+04 675s
324720 5.8702568e+02 0.000000e+00 2.856569e+05 680s
326300 5.8749793e+02 0.000000e+00 1.660520e+05 685s
327940 5.8790128e+02 0.000000e+00 3.922887e+05 690s
329678 5.8834979e+02 0.000000e+00 2.922348e+06 695s
331272 5.8880169e+02 0.000000e+00 7.246080e+04 700s
332908 5.8918782e+02 0.000000e+00 3.714649e+05 705s
334520 5.8956585e+02 0.000000e+00 1.642664e+05 710s
335974 5.8986190e+02 0.000000e+00 3.547859e+05 715s
337636 5.9019392e+02 0.000000e+00 3.392394e+05 720s
339286 5.9055663e+02 0.000000e+00 3.019631e+05 725s
340906 5.9092274e+02 0.000000e+00 1.028169e+06 730s
342536 5.9130568e+02 0.000000e+00 7.419633e+04 735s
344156 5.9165672e+02 0.000000e+00 8.681471e+05 740s
345676 5.9193423e+02 0.000000e+00 3.267727e+05 745s
347306 5.9227191e+02 0.000000e+00 1.309041e+05 750s
348796 5.9252109e+02 0.000000e+00 3.063238e+05 755s
350446 5.9286809e+02 0.000000e+00 7.848257e+04 760s
351926 5.9320860e+02 0.000000e+00 3.718607e+05 765s
353526 5.9350118e+02 0.000000e+00 4.147381e+04 770s
355016 5.9380097e+02 0.000000e+00 8.878677e+04 775s
356634 5.9405427e+02 0.000000e+00 1.488779e+05 780s
358094 5.9438621e+02 0.000000e+00 1.006930e+05 785s
359708 5.9467309e+02 0.000000e+00 5.166960e+04 790s
361254 5.9498572e+02 0.000000e+00 2.308388e+05 795s
362744 5.9525594e+02 0.000000e+00 1.257563e+06 800s
364198 5.9557208e+02 0.000000e+00 6.236926e+05 805s
365752 5.9584671e+02 0.000000e+00 7.443501e+04 810s
367140 5.9611422e+02 0.000000e+00 9.361074e+04 815s
368742 5.9639446e+02 0.000000e+00 6.577148e+04 820s
370182 5.9658212e+02 0.000000e+00 7.407510e+04 825s
371752 5.9677151e+02 0.000000e+00 5.091189e+05 830s
373188 5.9698326e+02 0.000000e+00 1.671590e+06 835s
374734 5.9725841e+02 0.000000e+00 1.835916e+05 840s
376328 5.9755223e+02 0.000000e+00 1.827007e+05 846s
Concurrent spin time: 0.00s
Solved with dual simplex
Root relaxation: objective 6.238698e+02, 248351 iterations, 842.03 seconds
Total elapsed time = 864.44s
Total elapsed time = 870.48s
Total elapsed time = 879.21s
Total elapsed time = 880.00s
Nodes | Current Node | Objective Bounds | Work
Expl Unexpl | Obj Depth IntInf | Incumbent BestBd Gap | It/Node Time
0 0 623.86978 0 22248 - 623.86978 - - 880s
0 0 569.79107 0 20535 - 569.79107 - - 991s
0 0 569.79107 0 20264 - 569.79107 - - 1002s
0 0 545.26054 0 19852 - 545.26054 - - 1150s
0 0 544.98032 0 20102 - 544.98032 - - 1277s
0 0 544.97399 0 19867 - 544.97399 - - 1290s
0 0 540.96068 0 18388 - 540.96068 - - 1595s
0 0 538.62803 0 18347 - 538.62803 - - 1876s
0 0 538.03362 0 18751 - 538.03362 - - 1956s
0 0 538.02926 0 18572 - 538.02926 - - 1995s
0 0 526.66031 0 16899 - 526.66031 - - 2705s
0 0 510.56976 0 16030 - 510.56976 - - 3002s
0 0 492.96287 0 15414 - 492.96287 - - 3576s
0 0 492.73198 0 15324 - 492.73198 - - 3730s
0 0 492.64599 0 15472 - 492.64599 - - 3789s
0 0 492.64144 0 15708 - 492.64144 - - 3820s
0 0 477.11857 0 14722 - 477.11857 - - 4406s
0 0 475.68420 0 14774 - 475.68420 - - 4595s
0 0 475.64830 0 15137 - 475.64830 - - 4665s
0 0 457.98450 0 13285 - 457.98450 - - 5250s
0 0 456.09615 0 13441 - 456.09615 - - 5462s
0 0 455.89813 0 13555 - 455.89813 - - 5574s
0 0 455.88711 0 13478 - 455.88711 - - 5606s
0 0 440.24472 0 13042 - 440.24472 - - 6244s
0 0 438.93436 0 12982 - 438.93436 - - 6444s
0 0 438.90477 0 13128 - 438.90477 - - 6512s
0 0 430.46668 0 12469 - 430.46668 - - 7657s
0 0 429.97314 0 12361 - 429.97314 - - 7848s
0 0 429.94669 0 12463 - 429.94669 - - 7933s
0 0 421.76719 0 12138 - 421.76719 - - 8523s
0 0 420.68969 0 12558 - 420.68969 - - 8671s
0 0 420.67835 0 12589 - 420.67835 - - 8730s
0 0 415.45638 0 11892 - 415.45638 - - 9337s
0 0 415.42708 0 11963 - 415.42708 - - 9400s
0 0 410.39141 0 11508 - 410.39141 - - 9935s
0 0 410.37788 0 11613 - 410.37788 - - 10008s
0 0 409.14276 0 11336 - 409.14276 - - 10426s
0 0 409.14276 0 11362 - 409.14276 - - 10473s
0 0 407.88464 0 11449 - 407.88464 - - 10829s
0 0 407.88301 0 11492 - 407.88301 - - 10870s
0 0 407.88093 0 11450 - 407.88093 - - 11260s
H 0 0 -673.0000000 407.88093 161% - 11260s
0 0 407.88093 0 4413 -673.00000 407.88093 161% - 12542s
H 0 0 -207.0000000 407.88093 297% - 13335s
H 0 0 -186.0000000 407.88093 319% - 13379s
0 2 407.88093 0 3660 -186.00000 407.88093 319% - 37534s
Cutting planes:
Gomory: 12
Implied bound: 3
Clique: 14
MIR: 236
StrongCG: 4
GUB cover: 23
Zero half: 264
Mod-K: 12
RLT: 1734
Explored 1 nodes (2660824 simplex iterations) in 37535.90 seconds
Thread count was 6 (of 6 available processors)
Solution count 3: -186 -207 -673
Time limit reached
Best objective -1.860000000000e+02, best bound 4.070000000000e+02, gap 318.8172%0 -
-
Hi Mathias
Barrier=2 probably helps to give you a quicker root node solution. But this will not do anything to accelerate the search. For that you could see whether "NoRelHeurTime=3600" (or so) helps?
Best regards
Simon0 -
投稿コメントは受け付けていません。
コメント
6件のコメント