==== FSL ====

[[https://fsl.fmrib.ox.ac.uk/fsl|FSL]]

FSL is a comprehensive library of analysis tools for FMRI, MRI and diffusion brain imaging data. Most of the tools can be run both from the command line and as GUIs (''point-and-click'' graphical user interfaces).

=== Job FSL 6.0.7.19 (single node, multitasks) for CPU ===

Example script ''slurm-fsl.sh'' to run ''fslmath'' on 1 node of the ''cpu'' partition using 4 cores and at most 8 GB of memory:

<code bash slurm-fsl.sh>
#!/bin/bash --login
#SBATCH --job-name=fsl
#SBATCH --output=%x.o%j
#SBATCH --error=%x.e%j
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --cpus-per-task=4
#SBATCH --time=0-02:00:00
#SBATCH --mem=8G
#SBATCH --partition=cpu
#SBATCH --qos=cpu
##SBATCH --account=<account>

test -n "$SLURM_NODELIST" || exit 1
test $SLURM_NNODES -eq 1 || exit 1

module load apptainer
module load fsl/6.0.7.19

test -n "$FSL_CONTAINER" || exit 1

apptainer exec "$FSL_CONTAINER" /bin/bash -lc '. "$FSLDIR/etc/fslconf/fsl.sh" && fslmaths -h'
</code>

Submission example:

<code>
sbatch slurm-fsl.sh
</code>

=== Interactive FSL 6.0.7.19 session ===

To run an interactive work session, you need to connect to the ''[[gui|gui.hpc.unipr.it]]'' server via RDP (Remote Desktop Protocol), change your keyboard layout (if necessary), open a terminal window. The following commands must be run in the terminal window.

To work in ''Group'' mode, run the newgrp command.

Example:

<code>
[foo@ui03 ~]$ groups
Person G_BAR
[foo@ui03 ~]$ newgrp G_BAR
</code>

In case you want to work in the ''$GROUP'' directory:

<code>
(G_BAR)[foo@ui03 ~]$ echo $GROUP
/hpc/group/G_BAR
(G_BAR)[foo@ui03 ~]$ cd $GROUP
(G_BAR)[foo@ui03 G_BAR]$ ...
</code>

In case you want to work in the ''$ARCHIVE'' directory:

<code>
(G_BAR)[foo@ui03 ~]$ echo $ARCHIVE
/hpc/archive/G_BAR
(G_BAR)[foo@ui03 ~]$ cd $ARCHIVE
(G_BAR)[foo@ui03 G_BAR]$ ...
</code>

== Interactive FSL 6.0.7.19 session (single node, multitasks) for CPU ==

Command to run ''fsl'' on 1 node of the ''cpu'' partition using 4 cores and at most 16 GB of memory:

<code>
srun \
    --job-name=fsl \
    --nodes=1 \
    --ntasks-per-node=1 \
    --cpus-per-task=4 \
    --mem=16G \
    --partition=cpu \
    --qos=cpu \
    --time=0-06:00:00 \
    --x11 \
    --pty \
    bash
</code>

Once you have access to the compute node, run these commands:

<code>
module load apptainer fsl
fsl.sh
fsleyes
</code>

Example:

<code>
[foo@wn82 ~]$ module load apptainer fsl
[foo@wn82 ~]$ fsl.sh 
(fsl-6.0.7.19.sif)[foo@wn82 ~]$ fsleyes
</code>

To close your work session, run the ''exit'' command twice. The first command exits the ''[[calcoloscientifico:userguide:apptainer|Apptainer]]'' container, while the second exits the remote node.

== Interactive FSL 6.0.7.19 session (single node, multitasks) for GPU ==

Command to run ''fsl'' on 1 node of the ''gpu'' partition using 4 cores, at most 16 GB of memory and 1 ''NVIDIA P100'' GPU:

<code>
srun \
    --job-name=fsl \
    --nodes=1 \
    --ntasks-per-node=1 \
    --cpus-per-task=4 \
    --mem=16G \
    --partition=gpu \
    --qos=gpu \
    --gres=gpu:p100:1 \
    --time=0-06:00:00 \
    --pty \
    --x11 \
    bash
</code>

Once you have access to the compute node, run these commands:

<code>
module load apptainer fsl
fsl.sh
fsleyes
</code>

Example:

<code>
[foo@wn41 ~]$ module load apptainer fsl
[foo@wn41 ~]$ fsl.sh 
(fsl-6.0.7.19.sif)[foo@wn41 ~]$ fsleyes
</code>

To close your work session, run the ''exit'' command twice. The first command exits the ''[[calcoloscientifico:userguide:apptainer|Apptainer]]'' container, while the second exits the remote node.