virtual memory

[jouteurestaque]

Hi guys

I have a virtual memory problem for the first time ever with crest Version 2.12,

forrtl: severe (41): insufficient virtual memory Image PC Routine Line Source
crest 000000000151947B Unknown Unknown Unknown crest 000000000048068B confcross 121 confcross.f90 crest 000000000067BF57 cross2 902 confscript2misc.f90 crest 000000000066F129 confscript2i 249 confscript2_main.f90 crest 00000000006984BE MAIN__ 276 crest_main.f90 crest 0000000000402242 Unknown Unknown Unknown crest 00000000017025D0 Unknown Unknown Unknown crest 000000000040210E Unknown Unknown Unknown

it is crashing at the

 |        Structure Crossing (GC)       |
 ========================================

input file name : crest_rotamers_4.xyz number of atoms : 52 number of points on xyz files : 25690 conformer energy window /kcal : 6.00 CN per atom difference cut-off : 0.3000 RMSD threshold : 0.2500 max. # of generated structures : 3450

if I relaunched it with --nocross it is ok but it resulted in an impressive number of conformers, pretty annoying to follow up with CENSO

any other way do managed that?

it was an xtb xtbopt_LA_rotation.xyz --opt --gfn2 -alpb water

`52

C 0.87883 0.21307 0.97715 C 0.27223 0.72110 -0.33394 C -1.23922 0.51386 -0.38987 C -1.83069 1.03293 -1.69897 C -3.34495 0.84292 -1.75550 C -3.95328 1.22590 -3.10436 C -3.86225 2.73408 -3.38598 C -4.51295 3.01947 -4.70968 O -3.96206 2.92725 -5.78713 O -5.78803 3.35415 -4.58980 C 2.35152 0.47624 1.01798 C 3.32992 -0.38321 1.25721 C 3.20950 -1.83935 1.60408 C 3.88535 -2.08774 2.92323 C 3.30082 -2.19208 4.10694 C 1.83610 -2.12820 4.40572 C 1.30908 -3.49344 4.86182 C -0.16800 -3.43790 5.24279 C -0.69062 -4.79329 5.71414 C -2.16217 -4.72921 6.10653 H 0.66454 -0.85002 1.08305 H 0.40408 0.73731 1.81242 H 0.74226 0.19782 -1.16992 H 0.49473 1.78508 -0.44394 H -1.70646 1.03410 0.44937 H -1.46355 -0.55019 -0.28639 H -1.58495 2.09134 -1.79830 H -1.36943 0.50432 -2.53641 H -3.81536 1.43178 -0.96595 H -3.57328 -0.20813 -1.56660 H -5.00531 0.93330 -3.11531 H -3.44371 0.68557 -3.90510 H -2.81867 3.03906 -3.44008 H -4.38319 3.28745 -2.60615 H -6.06336 3.48040 -3.66071 H 2.61522 1.50506 0.80274 H 4.35171 -0.02370 1.23754 H 3.72291 -2.42454 0.83543 H 2.16923 -2.15320 1.63461 H 4.96559 -2.14176 2.87027 H 3.92297 -2.36813 4.97636 H 1.67587 -1.40371 5.20877 H 1.27424 -1.78887 3.53647 H 1.89201 -3.83492 5.72056 H 1.45254 -4.21882 4.05745 H -0.75219 -3.10616 4.38129 H -0.30722 -2.70277 6.03910 H -0.55858 -5.52766 4.91631 H -0.10117 -5.12763 6.57094 H -2.76870 -4.41665 5.25961 H -2.30943 -4.01603 6.91444 H -2.51197 -5.70318 6.43964

thank you for any help`

[pprcht]

Hi, this is the same problem you will find on a couple of crossing related issues. I'm working on a fix on the 3.0dev development branch, but have to adjust it for better performance. Maybe I get to it over the weekend.

The current workaround is the --nocross that you already used. However, even fixing it won't reduce the number of conformers. On the contrary, you will end up with even more of them... I took a look at your system as well; The huge number of conformers shouldn't surprise you, it doesn't get much more flexible than such a linear molecule.

[jouteurestaque]

right sure about the linear molecule. I was surprised of that number cause the same one with one added double bond managed to crest it over giving 4692 conformers while this one resulted in 13549 conformers (with the nocross then)... Thank you again Philipp

[pprcht]

Hm, so the same molecule but with one less degree of freedom managed to get "only" 4692 structures, do I understand that right? Sounds plausible to me as the additional C-C torsion would add a factor of ~3 more conformers

[jouteurestaque]

yes that molecule 50 energy: -61.935060706640 gnorm: 0.000287960008 xtb: 6.4.1 (afa7bdf) C 0.31303053771271 0.12336060879078 0.46490562023108 C 0.17524145857475 -0.96637805776786 1.52597304584380 C -0.74685386296152 -0.57230889714948 2.68079902959298 C -2.19411918700789 -0.27171250478590 2.28186704184839 C -2.91184986888834 -1.40416659375279 1.54550104223409 C -3.02207093487078 -2.67946910680251 2.37731828599242 C -3.87866211568206 -3.73488554422483 1.66129983119378 C -3.93180749863264 -4.97693065804695 2.50466628003535 O -3.23002372444458 -5.95436935588729 2.34711260537426 O -4.81527419583239 -4.88706815306488 3.48577836275190 C 1.30773779554915 -0.25309301959393 -0.59578869719661 C 1.70482164624997 0.52016079632079 -1.59478220848447 C 1.25347840861395 1.92843499691692 -1.85344033135130 C 2.42280505925176 2.86429894850703 -1.75864547680825 C 2.87895488218787 3.69287596638285 -2.68530861691190 C 2.30992187377587 3.94304174297415 -4.05221746537061 C 1.87373882995525 5.37790022592659 -4.15102414326789 C 0.63346768603561 5.83850317743099 -4.09546224203708 C -0.62755892501385 5.04244688509619 -3.95964943664637 C -1.41853425480036 5.04578029173202 -5.26934857855407 H -0.65713943994417 0.30340749581595 -0.00480734875761 H 0.62180259917137 1.05502334214483 0.94711137794833 H 1.16445827559057 -1.17984152556732 1.93951188735432 H -0.18544917450328 -1.88421274225030 1.05845797746232 H -0.33887197134966 0.31841099181170 3.16538998408449 H -0.73714838622150 -1.37446954997978 3.42114751069280 H -2.21494132954414 0.61809437621273 1.65048488000643 H -2.75220955096617 -0.03618410296240 3.19118976486633 H -2.39742899453246 -1.62493645694754 0.60950485094976 H -3.91776257835436 -1.06157673749489 1.29324183241099 H -3.47541824227772 -2.44887299277933 3.34342552336797 H -2.03025436517428 -3.09758362324348 2.55675866056721 H -4.89055409633867 -3.35413127228939 1.52642504875395 H -3.43703836488151 -3.98134540762188 0.69634211022766 H -4.82582109478545 -5.67637246258468 4.06119576088667 H 1.71069214888921 -1.25463409211304 -0.51926148833756 H 2.43231046834510 0.14599724361694 -2.30371879822532 H 0.79993888921017 1.97577193017272 -2.84452819156146 H 0.50309299437866 2.22357829167773 -1.11804224438080 H 2.94282509128200 2.81199524581537 -0.80955746127138 H 3.74342968538185 4.30318043479685 -2.45293881749533 H 3.09624130603639 3.76898089227324 -4.79280282788311 H 1.48511215094566 3.26878448391510 -4.26826261076033 H 2.68810273610935 6.08594292872203 -4.24318166598488 H 0.47534994317858 6.90666095660339 -4.18442716052573 H -0.41057182407636 4.01689911787451 -3.66466700645871 H -1.23997290653165 5.49231096487131 -3.17414100923530 H -0.83430595315378 4.58660671183531 -6.06366407245794 H -1.65800931798713 6.06433432820466 -5.56631510653888 H -2.34642230766910 4.49097948046674 -5.15430530817151

[jouteurestaque]

but if I add 3 more so a total of 6 double bond compared to the 2 from the failing one I get 5206 conformers

56 energy: -71.437156524748 gnorm: 0.000551576388 xtb: 6.4.1 (afa7bdf) C 1.59008703040016 -1.16360898994732 0.40191370726537 C 1.99351884898124 -2.05635522785042 -0.48913032540192 C 2.30537865516419 -1.80179325567461 -1.93764984077622 C 1.08175181912029 -2.08357345666749 -2.76519251472222 C 0.80172568318169 -3.21765962226259 -3.38872440787507 C 1.66737259328823 -4.44675559402789 -3.42963728666667 C 2.47482800734144 -4.44924995376852 -4.69640579917892 C 3.79370253370741 -4.40672741856302 -4.80941502369999 C 4.80615077442151 -4.37034118163921 -3.70721753699531 C 5.92084651896434 -5.35171133844415 -3.93672336212892 C 7.07910532186555 -5.33004965895733 -3.29472467563591 C 7.45140916243740 -4.29849253754861 -2.26484468154802 C 7.86502080800739 -3.02202190448258 -2.94037481898189 C 7.43916522221973 -1.79497784381079 -2.67907634283448 C 6.48100928104523 -1.37263336256193 -1.60093909344339 C 7.27543265444558 -0.79139747627855 -0.46556052584208 C 7.40278982690120 0.50861806988920 -0.23324028617096 C 8.23620646748802 1.10629992436418 0.85615122722866 C 7.41876161786450 2.04422628947352 1.74503342081777 C 1.39338882025947 0.30421888165617 0.17579376162768 C 2.48205738951083 1.11671884881294 0.90108866854494 C 3.82045150893803 0.79916769834310 0.28863374815371 O 4.14734252031907 1.10126706938107 -0.83981396639236 O 4.61059068840500 0.12750466611662 1.10768104025223 H 1.40098118261620 -1.48813308047596 1.41720582651321 H 2.07564944382355 -3.09219350728635 -0.18639471528156 H 3.12721395047533 -2.44413255418504 -2.25121245148783 H 2.61330107961074 -0.76619538239329 -2.08874434372061 H 0.36451292953062 -1.27383651600114 -2.80083270523974 H -0.11400693801614 -3.28625227868482 -3.96176119721495 H 1.01987037731797 -5.32807993027858 -3.42737765450671 H 2.30877132801198 -4.49330863203641 -2.55279970847149 H 1.87544114509404 -4.46196122537244 -5.59871045927148 H 4.22710417732443 -4.39258763032444 -5.80247166628130 H 5.22926537847035 -3.36124689917485 -3.65220527493448 H 4.33970024279952 -4.58576651164686 -2.74223548790605 H 5.72652561314742 -6.12375729390598 -4.66893048294705 H 7.83215298935825 -6.07029334351710 -3.52677107455651 H 8.29342501791738 -4.66749452736152 -1.67404539729488 H 6.61419450304052 -4.12365502169384 -1.58912256937566 H 8.58366365318126 -3.16240688248805 -3.73849663142144 H 7.84435352286073 -0.97017129613718 -3.25162657573724 H 5.79909254317989 -0.61646780426051 -1.99788869840610 H 5.89605069618161 -2.22190854160231 -1.24524324790597 H 7.81452043865140 -1.50938541534878 0.13934606886916 H 6.87612753858836 1.22035460118216 -0.85865445617246 H 8.68247597747697 0.31477607744552 1.46067667053781 H 9.04773063021230 1.67773336003842 0.39617577318863 H 6.64014060882367 1.49577434118908 2.27046109079915 H 6.94691153644396 2.81973256510844 1.14479397155002 H 8.06041936764820 2.52162487762391 2.48115334582401 H 0.42092043751321 0.59977797867731 0.57287288696600 H 1.42065015257786 0.54692308750030 -0.88594056259969 H 2.50945836060203 0.86330074201498 1.95968088062159 H 2.28424348107376 2.18096900100412 0.77304769200568 H 5.48095488019523 -0.08912498317959 0.69643606824297

[pprcht]

This could happen if the PES becomes more flat and more minima end up in the 6 kcal energy window. With such flexible molecules it is sometimes hard to say. But the amounts of conformers certainly sound plausible.

[jouteurestaque]

thank you.

[moabe84]

Hi Philipp, I'm also facing this issue. Have you been able to fix it? Thanks.

Mostafa

[pprcht]

If it is also in the GC part for you, you may try the 3.0dev version, as I suggested above. The GC routines were revised from scratch in there, which hopefully avoids this error entirely.

[moabe84]

Yes, it happens in the GC part. Is this the 3.0dev version: https://github.com/crest-lab/crest/releases/download/latest/crest-latest.tar.xz

[pprcht]

Hm no, there the continuous release is build from the main branch. You can either build it yourself from the respective branch, or use this: crest.zip Please keep in mind this is a preliminary version while using it.

[moabe84]

Great. I'll try it and let you know about the results. Many thanks.

[moabe84]

Hi Philipp, I'd like to update you that the problem has been resolved using the 3.0dev version. Thank you very much.

[pprcht]

Thanks for letting me know!