waffleironhead 2008-01-25 00:50:55 | I have never seen the central particle be lost before and was wondering if it is a normal thing? This is happaning quite a bit, about 4 out of every 5 sims. The only lattice i am running is Linac88MHz900MeV1. Detected 2 logical processors Muon1 initialising... Hello, [TA]waffleironhead Loading 'results.dat'... OK (17 results) New simulation Searching for auto-saved file... No file present Loading 'samplefiles\Linac88MHz900MeV1.txt'... File empty Making new genome, TrialType=Extreme... Done Interpreting lattice file 'Linac88MHz900MeV1'... Done Beamline consists of 1074 units, genome of 55 parameters Using seed 0 Adding components to simulation space Tantalum rod source data loaded. Building proximity grid 1x1x50 (50 cells)... Done Tracking central particle to synchronise RF phases... 21.7% syncrfphases_track lost the central particle after 858.28ns; falling back to syn crfphases_arclength. Done adding components Determining nearby components... Normalising beam... - Starting - t = 8.02ns (32379/53054 particles) 12.7 Mpts |
Stephen Brooks 2008-01-25 08:16:01 | With the 900MeV linac this ought to disappear as the simulation improves. It's because some of the components actually decelerate the particle to zero rather than accelerating it. Or at least that's what I'm pretty sure is happening here. I'll keep it in mind if it starts doing it for the high-yielding results too. |
waffleironhead 2008-01-25 13:54:38 | ok, sounds good. Nothing to worry about then. thanks for the reply |
Xanathorn 2008-01-25 17:36:15 | So the result pasted below is also because of this or would it have another cause? Once in a while it pops up in my results file and screws my Muon Cockpit up (Lattice is also Linac88MHz900MeV1). pd1=525;rf1v=974;ld1=156;ls1f=000;ls1l=383;linaccells=000;prf1p=294;prf1v=213;ps1f=000;ps1l=999;phaserotcells=825;db1l=999;sb1f=999;sb1r=448;sb1l=823;d9l=678;d8l=325;s9f=537;s9r=262;s9l=188;d7l=584;s8f=452;s8r=351;s8l=869;d6l=828;s7f=879;s7r=999;s7l=999;d5l=561;s6f=000;s6r=999;s6l=809;d4l=483;s5f=838;s5r=000;s5l=311;d3l=000;s4f=333;s4r=999;s4l=542;d2l=273;s2f=439;s3f=998;s2r=000;s3r=000;s2l=999;s3l=999;d1l=967;s1f=000;s1l=000;decaycells=274;tantalumrodr=000;tantalumrodz=000;rf1p=000;#gen=4; nan (23.8 Mpts) [v4.44d] <Linac88MHz900MeV1> {06F6996902D853BF7F0B5C25} |
Stephen Brooks 2008-01-25 20:31:57 | No, that's something else. I'll have to check it out on Monday and either patch Muon1 or postpone that lattice (if it's a common problem) until I do have a fix. |
Zoxx 2008-01-27 15:41:28 | Just to add some additional observed info on Xanathorn's issue. Mine has completed 6 or 7 results out of 500+ done on the newer lattices and encountered this issue as well in MC, but none of mine were in Linac88MHz900MeV1 runs, all mine were in PhaseRotLinac6 or Linac88MHz900MeV6 and mine have allways been so far a trial type 0 (random) runs with a result Mpts <24.0. It does seem like it is only happening on very low -% results, the results very close to -2.0. |
Stephen Brooks 2008-01-28 12:03:18 | I can't repeat the "nan" error here. I get this result:pd1=525;rf1v=974;ld1=156;ls1f=000;ls1l=383;linaccells=000;prf1p=294;prf1v=213;ps1f=000;ps1l=999;phaserotcells=825;db1l=999;sb1f=999;sb1r=448;sb1l=823;d9l=678;d8l=325;s9f=537;s9r=262;s9l=188;d7l=584;s8f=452;s8r=351;s8l=869;d6l=828;s7f=879;s7r=999;s7l=999;d5l=561;s6f=000;s6r=999;s6l=809;d4l=483;s5f=838;s5r=000;s5l=311;d3l=000;s4f=333;s4r=999;s4l=542;d2l=273;s2f=439;s3f=998;s2r=000;s3r=000;s2l=999;s3l=999;d1l=967;s1f=000;s1l=000;decaycells=274;tantalumrodr=000;tantalumrodz=000;rf1p=000;#runs=5; |
Stephen Brooks 2008-01-28 12:06:46 | Though that did go through the queueing process whereas yours did not. Run as a single result, it got this:pd1=525;rf1v=974;ld1=156;ls1f=000;ls1l=383;linaccells=000;prf1p=294;prf1v=213;ps1f=000;ps1l=999;phaserotcells=825;db1l=999;sb1f=999;sb1r=448;sb1l=823;d9l=678;d8l=325;s9f=537;s9r=262;s9l=188;d7l=584;s8f=452;s8r=351;s8l=869;d6l=828;s7f=879;s7r=999;s7l=999;d5l=561;s6f=000;s6r=999;s6l=809;d4l=483;s5f=838;s5r=000;s5l=311;d3l=000;s4f=333;s4r=999;s4l=542;d2l=273;s2f=439;s3f=998;s2r=000;s3r=000;s2l=999;s3l=999;d1l=967;s1f=000;s1l=000;decaycells=274;tantalumrodr=000;tantalumrodz=000;rf1p=000; |
waffleironhead 2008-01-28 18:49:20 | Is it because you are running .e and we are running .d ? |
Xanathorn 2008-01-28 22:39:11 | Just from today, with v4.44d: rf1v=786;ld1=266;ls1f=000;ls1l=262;linaccells=000;prf1p=232;prf1v=179;ps1f=000;ps1l=917;pd1=386;phaserotcells=808;db1l=933;sb1f=882;sb1r=387;sb1l=679;d9l=719;d8l=454;s9f=492;s9r=127;s9l=157;d7l=456;s8f=021;s8r=297;s8l=671;d6l=981;s7f=677;s7r=863;s7l=999;d5l=539;s6f=000;s6r=999;s6l=858;d4l=467;s5f=738;s5r=000;s5l=272;d3l=000;s4f=517;s4r=999;s4l=646;d2l=274;s2f=503;s3f=995;s2r=000;s3r=246;s2l=901;s3l=999;d1l=983;s1f=000;s1l=101;decaycells=701;tantalumrodr=018;tantalumrodz=000;rf1p=000;#gen=2; nan (26.4 Mpts) [v4.44d] <Linac88MHz900MeV1> {DE6BC2F9FBD3B18D1EC9C693} rf1p=000;tantalumrodz=000;tantalumrodr=018;decaycells=422;s1l=101;s1f=000;d1l=983;s3l=999;s2l=901;s3r=441;s2r=000;s3f=995;s2f=503;d2l=274;s4l=646;s4r=999;s4f=517;d3l=000;s5l=272;s5r=000;s5f=738;d4l=467;s6l=857;s6r=998;s6f=000;d5l=539;s7l=999;s7r=863;s7f=677;d6l=981;s8l=671;s8r=297;s8f=189;d7l=456;s9l=157;s9r=127;s9f=453;d8l=453;d9l=718;sb1l=678;sb1r=387;sb1f=999;db1l=933;phaserotcells=807;ps1l=998;ps1f=000;prf1v=179;prf1p=232;linaccells=000;ls1l=394;ls1f=000;ld1=136;rf1v=786;pd1=386;#gen=1; nan (24.0 Mpts) [v4.44d] <Linac88MHz900MeV1> {96C35DE1B0F1F75E1B445332} pd1=394;rf1v=729;ld1=166;ls1f=000;ls1l=352;linaccells=002;prf1p=233;prf1v=353;ps1f=002;ps1l=995;phaserotcells=783;db1l=848;sb1f=999;sb1r=367;sb1l=805;d9l=714;d8l=436;s9f=130;s9r=000;s9l=059;d7l=418;s8f=402;s8r=310;s8l=633;d6l=994;s7f=427;s7r=358;s7l=999;d5l=507;s6f=000;s6r=999;s6l=926;d4l=496;s5f=728;s5r=000;s5l=000;d3l=000;s4f=504;s4r=999;s4l=665;d2l=210;s2f=535;s3f=996;s2r=000;s3r=475;s2l=874;s3l=999;d1l=980;s1f=002;s1l=196;decaycells=467;tantalumrodr=024;tantalumrodz=137;rf1p=000;#gen=4; nan (23.0 Mpts) [v4.44d] <Linac88MHz900MeV1> {86C29C688FD15FEC606BB4FB} rf1v=911;ld1=207;ls1f=000;ls1l=743;linaccells=000;prf1p=364;prf1v=135;ps1f=000;ps1l=999;pd1=464;phaserotcells=816;db1l=058;sb1f=990;sb1r=388;sb1l=649;d9l=617;s9f=474;s9r=260;s9l=186;d8l=410;s8f=489;s8r=385;s8l=732;d7l=671;s7f=827;s7r=971;s7l=999;d6l=792;s6f=000;s6r=999;s6l=839;d5l=996;s5f=986;s5r=000;s5l=418;d4l=360;s4f=544;s4r=986;s4l=452;d3l=994;s3f=978;s3r=011;s3l=999;d2l=253;s2f=345;s2r=000;s2l=625;d1l=885;s1f=000;s1l=026;decaycells=379;tantalumrodr=000;tantalumrodz=000;rf1p=000;#gen=5; nan (26.0 Mpts) [v4.44d] <Linac88MHz900MeV1> {79E92983468D2D005617C79A} s9r=303;s9l=000;s9f=000;tantalumrodz=200;tantalumrodr=024;decaycells=908;s1l=000;s1f=000;d1l=000;s2l=999;s2r=335;s2f=555;d2l=999;s3l=844;s3r=999;s3f=999;d3l=278;s4l=999;s4r=016;s4f=853;d4l=015;s5l=213;s5r=665;s5f=459;d5l=568;s6l=999;s6r=309;s6f=677;d6l=850;s7l=999;s7r=246;s7f=075;d7l=849;s8l=308;s8r=230;s8f=317;d8l=526;d9l=498;sb1l=904;sb1r=135;sb1f=974;db1l=323;phaserotcells=965;pd1=827;ps1l=999;ps1f=005;prf1v=999;prf1p=000;linaccells=000;ls1l=403;ls1f=000;ld1=000;rf1p=232;rf1v=557;#gen=1; nan (30.8 Mpts) [v4.44d] <Linac88MHz900MeV1> {3B82F425B7E31D6B47AAE784} pd1=000;rf1v=962;ld1=094;ls1f=000;ls1l=338;linaccells=000;prf1p=000;prf1v=181;ps1f=000;ps1l=999;phaserotcells=694;db1l=873;sb1f=997;sb1r=279;sb1l=777;d9l=154;d8l=000;s9f=853;s9r=000;s9l=999;d7l=000;s8f=000;s8r=000;s8l=678;d6l=999;s7f=477;s7r=200;s7l=999;d5l=426;s6f=000;s6r=999;s6l=820;d4l=675;s5f=703;s5r=000;s5l=371;d3l=000;s4f=737;s4r=999;s4l=509;d2l=000;s2f=458;s3f=982;s2r=000;s3r=511;s2l=822;s3l=999;d1l=999;s1f=001;s1l=685;decaycells=325;tantalumrodr=000;tantalumrodz=611;rf1p=000;#gen=4; nan (25.9 Mpts) [v4.44d] <Linac88MHz900MeV1> {0BB9E874DB6BAF64D63B0ED7} |
waffleironhead 2008-01-29 00:47:33 | I reran your first result Xanathorn and this is what I got. rf1v=786;ld1=266;ls1f=000;ls1l=262;linaccells=000;prf1p=232;prf1v=179;ps1f=000;ps1l=917;pd1=386;phaserotcells=808;db1l=933;sb1f=882;sb1r=387;sb1l=679;d9l=719;d8l=454;s9f=492;s9r=127;s9l=157;d7l=456;s8f=021;s8r=297;s8l=671;d6l=981;s7f=677;s7r=863;s7l=999;d5l=539;s6f=000;s6r=999;s6l=858;d4l=467;s5f=738;s5r=000;s5l=272;d3l=000;s4f=517;s4r=999;s4l=646;d2l=274;s2f=503;s3f=995;s2r=000;s3r=246;s2l=901;s3l=999;d1l=983;s1f=000;s1l=101;decaycells=701;tantalumrodr=018;tantalumrodz=000;rf1p=000; nan (26.5 Mpts) [v4.44d] <Linac88MHz900MeV1> {09C944236A86B64428AF0857} |
Zoxx 2008-01-29 12:53:50 | I just picked one of Xanathorn's at random as well, they do seem to repeat on client 4.44D so whatever changes you have made in that newer test client your using Stephen (that 4.44E you be testing out) it does appear to cure this issue. pd1=394;rf1v=729;ld1=166;ls1f=000;ls1l=352;linaccells=002;prf1p=233;prf1v=353;ps1f=002;ps1l=995;phaserotcells=783;db1l=848;sb1f=999;sb1r=367;sb1l=805;d9l=714;d8l=436;s9f=130;s9r=000;s9l=059;d7l=418;s8f=402;s8r=310;s8l=633;d6l=994;s7f=427;s7r=358;s7l=999;d5l=507;s6f=000;s6r=999;s6l=926;d4l=496;s5f=728;s5r=000;s5l=000;d3l=000;s4f=504;s4r=999;s4l=665;d2l=210;s2f=535;s3f=996;s2r=000;s3r=475;s2l=874;s3l=999;d1l=980;s1f=002;s1l=196;decaycells=467;tantalumrodr=024;tantalumrodz=137;rf1p=000;#gen=4; nan (23.1 Mpts) [v4.44d] <Linac88MHz900MeV1> {9A025C688FD46FEC327BA4FA} |
Stephen Brooks 2008-01-29 16:48:43 | 4.44e has hardly any changes that I can think of over 4.44d, I just changed the version number so I'd know if I'd accidentally uploaded any results using it. Are those consecutive results in Xanathorn's file? It's possible some of your data files (pion input etc.) have become corrupted, so try downloading the archive and extracting them again. |
Xanathorn 2008-01-29 22:40:33 | No it happens once every few hundred simulations, I just put those 'faulty' results away in a separate file and deletd the lines from my results.dat. I'll try that and see if those results come up again. I noticed though now my simulations average 0.1x % muons it occurs less often. |
Stephen Brooks 2008-01-30 19:46:47 | Do these "nan" results cause Muon1 to crash, or just Muon monitor? There's been a drop in calculation speed just this last 3-4 days since you noticed the problem. |
Xanathorn 2008-01-30 22:53:27 | It just crashes muon monitor, the client keeps running fine (no output drop). The "nan" results kept coming after a clean install downloaded from this website this morning by the way. |
Stephen Brooks 2008-02-05 11:06:26 | I've just tried all 6 of the results in Xanathorn's post and got:rf1v=786;ld1=266;ls1f=000;ls1l=262;linaccells=000;prf1p=232;prf1v=179;ps1f=000;ps1l=917;pd1=386;phaserotcells=808;db1l=933;sb1f=882;sb1r=387;sb1l=679;d9l=719;d8l=454;s9f=492;s9r=127;s9l=157;d7l=456;s8f=021;s8r=297;s8l=671;d6l=981;s7f=677;s7r=863;s7l=999;d5l=539;s6f=000;s6r=999;s6l=858;d4l=467;s5f=738;s5r=000;s5l=272;d3l=000;s4f=517;s4r=999;s4l=646;d2l=274;s2f=503;s3f=995;s2r=000;s3r=246;s2l=901;s3l=999;d1l=983;s1f=000;s1l=101;decaycells=701;tantalumrodr=018;tantalumrodz=000;rf1p=000; ...so half of them seem to produce 'nan' results, which is a bit strange. I'll try them again to see if this is a deterministic bug. |
Stephen Brooks 2008-02-05 11:16:20 | Second time around,rf1v=786;ld1=266;ls1f=000;ls1l=262;linaccells=000;prf1p=232;prf1v=179;ps1f=000;ps1l=917;pd1=386;phaserotcells=808;db1l=933;sb1f=882;sb1r=387;sb1l=679;d9l=719;d8l=454;s9f=492;s9r=127;s9l=157;d7l=456;s8f=021;s8r=297;s8l=671;d6l=981;s7f=677;s7r=863;s7l=999;d5l=539;s6f=000;s6r=999;s6l=858;d4l=467;s5f=738;s5r=000;s5l=272;d3l=000;s4f=517;s4r=999;s4l=646;d2l=274;s2f=503;s3f=995;s2r=000;s3r=246;s2l=901;s3l=999;d1l=983;s1f=000;s1l=101;decaycells=701;tantalumrodr=018;tantalumrodz=000;rf1p=000; |
Stephen Brooks 2008-02-05 11:19:57 | The 'nan' are the same but the other results are very slightly different in score. I'm suspecting something in the "-0.x" scoring algorithm right now, maybe it's reading a particle over the end of the buffer or something. |
Stephen Brooks 2008-02-07 12:54:08 | I noticed the "optimise" compiler switch which I determined not to use previously had been set on the "4.44e" development code. Switching it off gives merf1v=786;ld1=266;ls1f=000;ls1l=262;linaccells=000;prf1p=232;prf1v=179;ps1f=000;ps1l=917;pd1=386;phaserotcells=808;db1l=933;sb1f=882;sb1r=387;sb1l=679;d9l=719;d8l=454;s9f=492;s9r=127;s9l=157;d7l=456;s8f=021;s8r=297;s8l=671;d6l=981;s7f=677;s7r=863;s7l=999;d5l=539;s6f=000;s6r=999;s6l=858;d4l=467;s5f=738;s5r=000;s5l=272;d3l=000;s4f=517;s4r=999;s4l=646;d2l=274;s2f=503;s3f=995;s2r=000;s3r=246;s2l=901;s3l=999;d1l=983;s1f=000;s1l=101;decaycells=701;tantalumrodr=018;tantalumrodz=000;rf1p=000;#runs=5; |
Stephen Brooks 2008-02-07 12:57:05 | However, running it with the v4.44d as released I gotrf1v=786;ld1=266;ls1f=000;ls1l=262;linaccells=000;prf1p=232;prf1v=179;ps1f=000;ps1l=917;pd1=386;phaserotcells=808;db1l=933;sb1f=882;sb1r=387;sb1l=679;d9l=719;d8l=454;s9f=492;s9r=127;s9l=157;d7l=456;s8f=021;s8r=297;s8l=671;d6l=981;s7f=677;s7r=863;s7l=999;d5l=539;s6f=000;s6r=999;s6l=858;d4l=467;s5f=738;s5r=000;s5l=272;d3l=000;s4f=517;s4r=999;s4l=646;d2l=274;s2f=503;s3f=995;s2r=000;s3r=246;s2l=901;s3l=999;d1l=983;s1f=000;s1l=101;decaycells=701;tantalumrodr=018;tantalumrodz=000;rf1p=000; So maybe I left the bad compiler switch on when I compiled 4.44d? It's possible. Anyway, working towards version 4.45 now, which I guess should be rid of that problem. |
Stephen Brooks 2008-02-07 12:59:05 | Oh 4.44e with the bad switch does give the same three 'nan' results as 4.44d I think. |