johanneswiesner / workshop_cimh Goto Github PK

View Code? Open in Web Editor NEW

1.0 1.0 1.0 51.35 MB

Neuroimaging Workshop

Home Page: https://johanneswiesner.github.io/workshop_cimh/

License: BSD 3-Clause "New" or "Revised" License

Python 2.36% Jupyter Notebook 93.18% Shell 2.83% Dockerfile 1.63%

workshop_cimh's Introduction

"What's in this repository?"

This repository contains the skeleton for the CIMH neuroimaging workshop.

Credits

This repository was created using the G0RELLA template

"I have some questions..."

Open an issue on this repository and someone will try and get back to you as soon as possible!

workshop_cimh's People

Contributors

Stargazers

Watchers

Forkers

caggursoy

workshop_cimh's Issues

Need to use winpty for docker in Git Bash under Windows

if [[ "$OSTYPE" == "linux-gnu"* ]]; then
	singularity run -B /zi-flstorage/group_csp/analyses/johannes.wiesner/projects/hcp_project/code/:/home/neuro/workdir \
	-B /zi-flstorage/group_csp/open_access_datasets/humanconnectome/raw_data/:/data \
	-B /zi-flstorage/group_csp/open_access_datasets/humanconnectome/derivatives/:/output \
	/csp-tools/singularity_images/csp_neurodocker/csp.sif jupyter-lab --notebook-dir=/home/neuro/workdir
elif [[ "$OSTYPE" == "msys" ]]; then
	winpty docker run -it --rm -v /C:/Users/Johannes.Wiesner/Documents/projects/hcp_project/code/:/home/neuro/workdir \
	-v /C:/Users/Johannes.Wiesner/Documents/projects/hcp_project/data/humanconnectome/raw_data/:/data \
	-v /C:/Users/Johannes.Wiesner/Documents/projects/hcp_project/data/humanconnectome/derivatives/:/output \
	-p 8888:8888 johanneswiesner/csp:1.0.0 jupyter lab
fi

Don't use python but bash in check_install.sh

Don't use python here, we had problems with a Windows machine that couldn't execute this code in Git Bash because Python wasn't installed.

Do we need WSL+Datalad?

Workshop won't have an introduction to datalad so I guess we don't need users to install WSL+datalad?

Dockerfile: Switch back to root user

Run root-branch from neurodocker to generate right folder structure. Combine this with #5

download_courses for Git Bash

Windows users should be able to use Git Bash to download course material. For this, we have to tweak the bash script:

https://stackoverflow.com/a/65571522

Add check_install.sh script

https://github.com/ReproNim/DGPA_workshop_2022/blob/main/check_install.sh

Add ANTS + FSL

Add ANTS + FSL so we can run a minimal-preprocessing pipeline for the anatomical workflow

Do we need VSCode?

VSCode + Docker installation is nice, but if we only stick to jupyter-notebooks a docker image with jupyter-lab included would be sufficient?

Introduce docker run with real life example

Use this script https://github.com/JohannesWiesner/workshop_cimh/blob/master/run_docker.sh so users already use the right folder paths for their analysis

Add OSO Video

https://gist.github.com/bitsurgeon/acd769ae3236c48a433e0acb11784c2e

Find another way to download the SPM auditory dataset that is equal to `nilearn.datasets.fetch_spm_auditory` but without using nilearn

Instead of adding a /data folder to the github repository, one should be able to use

https://nilearn.github.io/dev/modules/generated/nilearn.datasets.fetch_spm_auditory.html

to download the auditory spm dataset. The problem is that we need to find a way to retrieve this dataset without nilearn as we can't rely on people having python + nilearn installed before the workshop

Include items of pre-workshop and post-workshop questionnaires

This offers other users templates for these questionnaires (and also makes it possible to use other providers than redcap).

Possible to download all the data before hand?

Currently, we use both

from nilearn.datasets import fetch_oasis_vbm and fetch_spm_auditor from nilearn to download data (functional and anatomical).

1.) That takes time during the workshop (would be nice if participants could do this beforehand)
2.) We cannot be sure that people end up with firewall-issues in the workshop (smarter to tell participants to download before workshop so we can debug stuff in case something goes wrong).

Then use setfacl to set full permission (read+write+execute) on all the folders (/home/csp/data, /home/csp/output, etc.)

https://wiki.ubuntuusers.de/ACL/#Setzen-und-Loeschen-von-ACLs