I am running bedpost on a cluster using the following code:
############################################################
#!/bin/bash
##script written in Bash. Comments at the beginning should be prefaced with two number signs.
## output files from the job will go into the output_files folder. error files go into the error_files folder
#PBS -e /home/nsanchez/Desktop/error_files/error.$PBS_JOBID
#PBS -o /home/nsanchez/Desktop/output_files/output.$PBS_JOBID
# we don't have a Sun Grid Environment cluster (SGE), so don't use it
SGE_ROOT=""
export SGE_ROOT
FSLDIR=/opt/fsl # directory where the fsl program is located. This should not change.
EXAMPLEDIR=/home/nsanchez/Desktop # directory where the script and data folder are located
DATADIR=$EXAMPLEDIR/NT_DTI # directory where data is stored
RESULTSDIR=$EXAMPLEDIR/resultsNT # directory where results will be stored
EXECUTEDIR=$RESULTSDIR/NT_DTI # directory that data is copied to, which will be analyzed - last folder should be the same as the data directory
rm -rf $RESULTSDIR # remove previous results if necessary
mkdir $RESULTSDIR # make directory to store results from this analysis
echo "Starting FDT (bedpost) at $(date)"
# Copy data directory into the results directory. Results directory will be split into a data dir and a .bedpostx dir, which contains the results of the analysis.
cp -r $DATADIR $RESULTSDIR
# run bedpostx
unset FSLMACHINELIST
bedpostx $EXECUTEDIR -n 2
# finish up
echo "end time = $(date)"
################################################################
I uploaded the requested documents
Thanks a lot for your help!
|