wigToBigWig error

[JohnK]

Hello,

Trying to use the script 'wigToBigWig' for the obvious and I keep receiving a floating point exception error. Can't seem to find the cause. Has anyone encountered this before?

John

[dawe]

Hello,

Trying to use the script 'wigToBigWig' for the obvious and I keep receiving a floating point exception error. Can't seem to find the cause. Has anyone encountered this before?

John

Can you post the exact command you are executing?
I've never been into wigToBigWig errors, except for chromosome sizes :-) but there's the -clip options for that

d

[JohnK]

Can you post the exact command you are executing?
I've never been into wigToBigWig errors, except for chromosome sizes :-) but there's the -clip options for that

d

#!/bin/bash

#$-cwd
#$-V
#$ -l huge_mem
#$ -l mem=24G


~/wigToBigWig ./chrchr2/j.wig ./mm9.size.bld ./chrchr2/test.bw


The error I receive is such:

/opt/gridengine/default/spool/compute-4-0/job_scripts/17246: line 9: 30495 Floating point exception~/wigToBigWig ./chrchr2/j.wig ./mm9.size.bld ./chrchr2/j.bw

[JohnK]

*the huge mem was me just trying different things.

[dawe]

#!/bin/bash

#$-cwd
#$-V
#$ -l huge_mem
#$ -l mem=24G


~/wigToBigWig ./chrchr2/j.wig ./mm9.size.bld ./chrchr2/test.bw


The error I receive is such:

/opt/gridengine/default/spool/compute-4-0/job_scripts/17246: line 9: 30495 Floating point exception~/wigToBigWig ./chrchr2/j.wig ./mm9.size.bld ./chrchr2/j.bw

Whoa... Never happened to me... Just to know:
1- which CFLAGS have been used to compile wigToBigWig? (not -O3 I hope)
2- how big is the wig? are you sure it is "kent compilant" ?

what does

qacct -j 17246

says (assuming 17246 is the $JOBID for your gridengine scheduler)?

d

[JohnK]

Whoa... Never happened to me... Just to know:
1- which CFLAGS have been used to compile wigToBigWig? (not -O3 I hope)
2- how big is the wig? are you sure it is "kent compilant" ?

what does

qacct -j 17246

says (assuming 17246 is the $JOBID for your gridengine scheduler)?

d

this is the result from qacct -j $id

ru_isrss 0
ru_minflt 697
ru_majflt 0
ru_nswap 0
ru_inblock 0
ru_oublock 0
ru_msgsnd 0
ru_msgrcv 0
ru_nsignals 0
ru_nvcsw 49
ru_nivcsw 1
cpu 0
mem 0.000
io 0.000
iow 0.000
maxvmem 0.000

i removed the track header lines, so i just have the raw .wig information like such:

3000076 -1
3000436 -1
3000437 -1
3000441 -1
3000466 -1
3000482 -1
3000545 -1
3000549 -1
3000637 -1
3000650 -1


here's an 'od -c' to show the character usage:

0000000 3 0 0 0 0 7 6 - 1 \n 3 0 0 0 4
0000020 3 6 - 1 \n 3 0 0 0 4 3 7 - 1
0000040 \n 3 0 0 0 4 4 1 - 1 \n 3 0 0 0
0000060 4 6 6 - 1 \n 3 0 0 0 4 8 2 -
0000100 1 \n 3 0 0 0 5 4 5 - 1 \n 3 0 0
0000120 0 5 4 9 - 1 \n 3 0 0 0 6 3 7
0000140 - 1 \n 3 0 0 0 6 5 0 - 1 \n 3 0
0000160 0 0 6 5 6 - 1 \n 3 0 0 0 6 7 8
0000200 - 1 \n 3 0 0 0 8 1 4 - 1 \n 3
0000220 0 0 0 8 2 4 - 1 \n 3 0 0 0 9 6
0000240 2 - 1 \n 3 0 0 1 0 0 7 - 1 \n
0000260 3 0 0 1 0 3 8 - 1 \n 3 0 0 1 2
0000300 4 1 - 1 \n 3 0 0 1 3 6 3 - 1
0000320 \n 3 0 0 1 4 8 3 - 1 \n 3 0 0 1
0000340 4 8 6 - 1 \n 3 0 0 2 2 2 6 -


just a line by line separated by a whitespace char.

[JohnK]

Whoa... Never happened to me... Just to know:
1- which CFLAGS have been used to compile wigToBigWig? (not -O3 I hope)
2- how big is the wig? are you sure it is "kent compilant" ?

what does

qacct -j 17246

says (assuming 17246 is the $JOBID for your gridengine scheduler)?

d

I thought it was already compiled from the ucsc repository here:

http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64/

the find chromosome start/stop worked for me.

the .wig file is approximately 3.7M.

[dawe]

I thought it was already compiled from the ucsc repository here:

http://hgdownload.cse.ucsc.edu/admin/exe/linux.x86_64/

the find chromosome start/stop worked for me.

the .wig file is approximately 3.7M.

I've asked because I usually download their sources and compile on my own (and had issues with optimization flags).
So, the binary is fine, gridengine doesn't report anything strange (except that the job has barely started), the wig is little and compilant (besides, I don't think you'll need to request 24 Gb to SGE to process a 4Mb wig file...)... I guess you'll have to run gdb to figure this out...

[JohnK]

I've asked because I usually download their sources and compile on my own (and had issues with optimization flags).
So, the binary is fine, gridengine doesn't report anything strange (except that the job has barely started), the wig is little and compilant (besides, I don't think you'll need to request 24 Gb to SGE to process a 4Mb wig file...)... I guess you'll have to run gdb to figure this out...

I believe I got the fix. Anyway, I downloaded the John Kent source tree package 'jksource.zip' and had it installed:

http://hgdownload.cse.ucsc.edu/admin/jksrc.zip

No more floating point exception. You were right on about the compiling issues I believe.