Dear Eric,
It isn't totally clear to me why this is failing. Also, you say that "many of them crash" - does that mean that some succeed? If so then my best guess would be that you are overloading your system and it is running out of memory, since that would cause the bad_alloc and would explain why some might succeed and others not (depending on what competition they had for memory at critical times).
I would try to run one of the failed ones on its own and see if that works.
All the best,
Mark
On 17 Apr 2013, at 05:08, Eric Walden <[log in to unmask]> wrote:
> I am running 1000 possum simulations on a cluster and many of them crash with the error below. So far, none have generated a signal file. As you can (probably) determine from below, I am using a possum folder that has already been "possumed". That is, I have already run possum from this folder and generated a signal file. Now I am trying to switch the timecourse file each time to run 1000 simulations with noise added to the timecourse, rather than just thermal noise. Each time course file is cleverly named something like timecourse1, timecourse2, etc... and the output should be named signal1, signal2, etc...
>
> Any help would be appreciated.
>
> -------------------------------------------------------------------------------
> The error file looks like this:
>
> terminate called after throwing an instance of 'std::bad_alloc'
> what(): std::bad_alloc
> /opt/gridengine/default/spool/compute-30-4/job_scripts/385621: line 10: 30265 Aborted (core dumped) possum -i BasePossum/brain.nii.gz -x BasePossum/MRpar -m BasePossum/motion -p BasePossum/pulse -o BasePossum/Signal$SGE_TASK_ID -f BasePossum/slcprof -a BasePossum/T2.nii.gz -t timecourses/timecourse$SGE_TASK_ID.txt --nproc=1
>
> -------------------------------------------------------------------------------
> The output file looks like this:
>
>
> Starting POSSUM...
>
> Setting up the coordinate system...
>
> Setting up the process ID...
> Number of processors is 1.; My ID is 0.
>
> Reading the object for ID 0...
> object:: Size = (181,217,181,3)
> object:: ROI Size = (181,217,181,3)
> object:: Dims = (1,1,1,1)
> object:: Minimum and maximum intensities are: 0 and 1
>
> Reading the tissue properties...
> T1,T2,SD,CS:
> 1.331000 0.051000 0.860000 0.000000
> 0.832000 0.044000 0.770000 0.000000
> 3.700000 0.500000 1.000000 0.000000
>
> Reading the pulse sequence...
> [SeqType,TE,TR,TRslc,Nx,Ny,dx,dy,maxG,RiseT,BW,Nvol,Nslc,SlcThk,SlcDir,Gap,zstart,FA]
> 1.000000 0.030000 3.000000 0.075000 64.000000 64.000000 0.003000 0.003000 0.035000 0.000200 143360.000000 140.000000 40.000000 0.003000 1.000000 0.000000 0.030000 90.000000 2.000000 3.000000 100.000000 1.000000
>
>
> Reading the motion file...
> Motion file is 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
> 0.100000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000
>
> Creating table for the slice profile...
> Stepsize for the table: dslcp= 0.0154915
>
> Calculating zstart, zend and level...
> Slice selection direction is 1. (-+1 is z, -+2 is y and -+3 is x.)
> Level is 1
>
> Extra slc calculation...
> Begining of the object is 28 and the end is 152 (in vox, in slice select direction).
> Motion add on - side (mm):0; Motion add on + side (mm):0; Slice profile add:0.0022412
> Extra voxels on - side:2& Extra voxels on + side (mm):2
> Extra slices on - side:1 & Extra slices on + side:1
>
> Creating activation4D volume and timecourse vector...
> 3D activation mode
> Testing if the values in the activation volumes are smaller than the T2* values
> vol= 0
> vol= 1
> vol= 2
> vol= 3
> vol= 4
> vol= 5
> vol= 6
> vol= 7
> vol= 8
> vol= 9
> vol= 10
> vol= 11
> vol= 12
> vol= 13
> vol= 14
> vol= 15
> vol= 16
> vol= 17
> vol= 18
> vol= 19
> vol= 20
> vol= 21
> vol= 22
> vol= 23
> vol= 24
> vol= 25
> vol= 26
> vol= 27
> vol= 28
> vol= 29
> vol= 30
> vol= 31
> vol= 32
> vol= 33
> vol= 34
> vol= 35
> vol= 36
> vol= 37
> vol= 38
> vol= 39
> vol= 40
> vol= 41
> vol= 42
> vol= 43
> vol= 44
> vol= 45
> vol= 46
> vol= 47
> vol= 48
> vol= 49
> vol= 50
> vol= 51
> vol= 52
> vol= 53
> vol= 54
> vol= 55
> vol= 56
> vol= 57
> vol= 58
> vol= 59
> vol= 60
> vol= 61
> vol= 62
> vol= 63
> vol= 64
> vol= 65
> vol= 66
> vol= 67
> vol= 68
> vol= 69
> vol= 70
> vol= 71
> vol= 72
> vol= 73
> vol= 74
> vol= 75
> vol= 76
> vol= 77
> vol= 78
> vol= 79
> vol= 80
> vol= 81
> vol= 82
> vol= 83
> vol= 84
> vol= 85
> vol= 86
> vol= 87
> vol= 88
> vol= 89
> vol= 90
> vol= 91
> vol= 92
> vol= 93
> vol= 94
> vol= 95
> vol= 96
> vol= 97
> vol= 98
> vol= 99
> vol= 100
> vol= 101
> vol= 102
> vol= 103
> vol= 104
> vol= 105
> vol= 106
> vol= 107
> vol= 108
> vol= 109
> vol= 110
> vol= 111
> vol= 112
> vol= 113
> vol= 114
> vol= 115
> vol= 116
> vol= 117
> vol= 118
> vol= 119
> vol= 120
> vol= 121
> vol= 122
> vol= 123
> vol= 124
> vol= 125
> vol= 126
> vol= 127
> vol= 128
> vol= 129
> vol= 130
> vol= 131
> vol= 132
> vol= 133
> vol= 134
> vol= 135
> vol= 136
> vol= 137
> vol= 138
> vol= 139
>
> Reading the RF inhomogeneity volumes...
> RFobject:: Size = (181,217,181)
> RFobject:: ROI Size = (181,217,181)
> RFobject:: Dims = (1,1,1)
> RFobject:: Minimum and maximum intensities are: 1 and 1
>
>
> Calculating the signal...
> Number of read out points is 22937600
>
> LEVEL1
> Creating 1 B0file together with 3 gradient files...
> b0:: Size = (181,217,181)
> b0:: ROI Size = (181,217,181)
> b0:: Dims = (1,1,1)
> b0:: Minimum and maximum intensities are: 0 and 0
>
> Main loop...
> Tissue type=0; zstart=28; zz=28; zend=153; Voxelnumber=0
> Table sinc calc
|