| T O P I C R E V I E W |
| mcasp |
Posted - May 21 2008 : 3:49:36 PM
Hi All,
I'm trying to submit a script to the cluster like so
[casp@head Conte.05]$ qsub -v EXPERIMENT=Conte.05 qsub_feat_level1_6runs.sh 33214 500
qsub: Unknown option
but I keep getting this "qsub: Unknown option" and I'm not sure why this is. Has anyone run into this? |
| 7 L A T E S T R E P L I E S (Newest First) |
| tankersley |
Posted - May 29 2008 : 1:10:35 PM
Thanks everyone.
Dharol |
| syam.gadde |
Posted - May 27 2008 : 11:14:26 AM
In fact, if you explictly call the replacement strings something like REPLACETHIS_OUTPUTDIR, REPLACETHIS_C etc., you'll have no confusion when looking at the template as to what's happening. |
| petty |
Posted - May 27 2008 : 11:11:36 AM
thats because everytime sed finds these letter strings:B,G,Rr,Rnr,C,BP ... its replacing them with your variables.
if you just use more unique variables for replacement in your template.fsf everything should work. |
| josh.bizzell |
Posted - May 27 2008 : 11:09:30 AM
You need to use more specific placeholders in your template file.... for example, this line in the sed statement:
-e 's@C@'$VAR5'@g' \
will replace every capital "C" in your template fsf file with the variable in VAR5.
-Josh |
| tankersley |
Posted - May 27 2008 : 11:04:01 AM
I'm trying to write a cluster script for Level 1 feat. My sed section looks like:
for i in 'L1_template.fsf'; do
sed -e 's@OUTPUT@'$OUTPUT'@g' \
-e 's@B@'$VAR1'@g' \
-e 's@G@'$VAR2'@g' \
-e 's@Rr@'$VAR3'@g' \
-e 's@Rnr@'$VAR4'@g' \
-e 's@C@'$VAR5'@g' \
-e 's@BP@'$VAR6'@g' \
-e 's@ANAT@'$ANAT'@g' \
-e 's@DATA@'$DATA'@g' <$i> ${MAINOUTPUT}/L1_run${run}.fsf
done
The problem is that the templates it ends up creating have the info spliced in all weird:
# FEAT version number
set fmri(version) 5.92
# Are we in MELODI/mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4XBh/Control/42018/Column_Data/42018_4_C.txt?
set fmri(inmelodic) 0
# Analysis level
# 1 : First-level analysis
# 2 : Higher-level analysis
set fmri(level) 1
# Which stages to run
# 0 : No first-level analysis (registration and/or group stats only)
# 7 : Full first-level analysis
# 1 : Pre-Stats
# 3 : Pre-Stats + Stats
# 2 : Stats
# 6 : Stats + Post-stats
# 4 : Post-stats
set fmri(analysis) 6
# Use relative filenames
set fmri(relative_yn) 0
# /mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4XBh//mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4XBh/Control/42018/Column_Data/42018_4_C.txtontrol/42018//mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4XBh/Control/42018/Column_Data/42018_4_C.txtolumn_Data/42018_4_B.txtalloon help
set fmri(help_yn) 0
# Run Feat***cher
set fmri(feat***cher_yn) 0
# /mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4XBh/Control/42018/Column_Data/42018_4_C.txtleanup first-level standard-space images
set fmri(sscleanup_yn) 0
# Output directory
set fmri(outputdir) "/mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4X/mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4XBh//mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4XBh/Control/42018/Column_Data/42018_4_C.txtontrol/42018//mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4XBh/Control/42018/Column_Data/42018_4_C.txtolumn_Data/42018_4_B.txth//mnt/users.q.shares/tankersley.9qIljtNsQvhVg8Wo4XBh/Control/42018/Column_Data/42018_4_C.txtontrol/42018/Level_1/L1_run4"
. . .etc.
I assume i've got something wrong with the sed syntax but have rewritten this thing from scratch several times using the qsub templates provided on the wiki and am still getting weird splicing.
Any help appreciated.
Thanks,
Dharol
|
| dvsmith |
Posted - May 21 2008 : 5:08:33 PM
Looks like you're on the head node (e.g., "casp@head"). Were you able to submit the basic.sh script using Conte.05 (http://fourier.biac.duke.edu/wiki/suppl/intro.htm)? If not, it's probably some problem with your account.
David |
| petty |
Posted - May 21 2008 : 3:59:11 PM
are you logged into q-interact?
if so you can't submit jobs from there, you have to be at the head node. |
