Skip to content

Latest commit

 

History

History
170 lines (125 loc) · 6.38 KB

README.md

File metadata and controls

170 lines (125 loc) · 6.38 KB
Raw

Integration Testing: CircleCI

Note: This demo is intended to run on OS X and GNU/Linux environments, but you can use the Docker container to run on any system you can run Docker or Singularity on.

Note: This is the GitHub Repo that accompanies: Ghosh SS, Poline JB, Keator DB et al. A very simple, re-executable neuroimaging publication [version 2; referees: 1 approved, 3 approved with reservations]. F1000Research 2017, 6:124 (doi: 10.12688/f1000research.10783.2)

Notice for commercial use

The following non-free Debian packages are part of the Docker container:

non-free/science fsl-5.0-core non-free/science fsl-atlases, fsl-first-data

If you are considering commercial use of the container please consult the relevant license.

Information queried from NITRC-IR and stored in a google spreadsheet

https://docs.google.com/spreadsheets/d/11an55u9t2TAf0EV2pHN0vOd8Ww2Gie-tHp9xGULh_dA

1. To execute demo with Docker

First download the image

docker pull repronim/simple_workflow:latest

ERRATA (9 Feb 2022): The docker image contains an older set of SSL certificate authority certificates. This results in an error when running the command. We can provide the following environmental variable to make container work without rebuilding it. -e CURL_CA_BUNDLE=''

Now you can run the image as follows to test on 1 participant's data:

docker run -it --rm -v $PWD/output:/opt/repronim/simple_workflow/scripts/output \
   -e CURL_CA_BUNDLE='' repronim/simple_workflow:latest run_demo_workflow.py \
   --key 11an55u9t2TAf0EV2pHN0vOd8Ww2Gie-tHp9xGULh_dA -n 1 
docker run -it --rm -v $PWD/output:/opt/repronim/simple_workflow/scripts/output \
   -e CURL_CA_BUNDLE='' repronim/simple_workflow:latest check_output.py --ignoremissing

You can extend this to all participants as follows.

docker run -it --rm -v $PWD/output:/opt/repronim/simple_workflow/scripts/output \
   -e CURL_CA_BUNDLE='' repronim/simple_workflow:latest run_demo_workflow.py \
   --key 11an55u9t2TAf0EV2pHN0vOd8Ww2Gie-tHp9xGULh_dA
docker run -it --rm -v $PWD/output:/opt/repronim/simple_workflow/scripts/output \
   -e CURL_CA_BUNDLE='' repronim/simple_workflow:latest check_output.py

2. To execute demo within your current environment

The script will check for availability of the following:

  1. FSL
  2. curl
  3. git or unzip

Download the script

curl -Ok https://raw.githubusercontent.com/ReproNim/simple_workflow/3a64f3ffdf174e7c8970de620aa62a9167e3134f/Simple_Prep.sh

Setup the environment.

This will setup a complete environment within a directory called simple_workflow. It will not add anything to your existing environment.

Note: Because of some limitations with the underlying packages, it may determine that your current working directory path is too long and will ask you to move to a shorter directory path.

bash Simple_Prep.sh

Execute a single subject test.

bash Simple_Prep.sh test

Run the full demo on all subjects

Option 1:

bash Simple_Prep.sh replay

Option 2: To use the newly created environment

cd simple_workflow/scripts
export OLDPATH=$PATH
export PATH=$PWD/../miniconda/envs/bh_demo/bin:$PATH

Now you can run the demo script

python run_demo_workflow.py --key 11an55u9t2TAf0EV2pHN0vOd8Ww2Gie-tHp9xGULh_dA
python check_output.py

To run on one subject you can do:

python run_demo_workflow.py --key 11an55u9t2TAf0EV2pHN0vOd8Ww2Gie-tHp9xGULh_dA -n 1

To restore old environment do:

export PATH=$OLDPATH
unset OLDPATH

3. To build docker image with custom environment

Using containerization solutions, such as Docker, allows to create multiple complete computation environments while varying versions of any analysis pipeline components or inputs. You could use Simple_Prep_docker script to generate environments based on previous Debian or Ubuntu releases for which NeuroDebian builds of FSL were available in the past.

N.B. ATM NeuroDebian snapshots repository is not widely open yet, so if you would like to browse it, please "knock" first by running curl -s http://neuro.debian.net/_files/knock-snapshots command in your shell.

Generate an environment

For the current test data, we generated an environment based on Debian jessie release with FSL 5.0.9 that was available in April 2017:

/Simple_Prep_docker jessie 20170410T000000Z

which will generate a local docker image repronim:simple_prep_${USER}_jessie_20170410T000000Z (${USER} will correspond to your user name), with all necessary for computation components installed.

You can then run the demo using the docker image run commands above by replacing repronim/simple_workflow:latest with repronim:simple_prep_${USER}_jessie_20170410T000000Z

Changing the FSL version

You can also generate an environment based on Debian jessie release with FSL 5.0.8-3 as it was available in March of 2015 by simply changing the timestamp:

./Simple_Prep_docker jessie 20150306T060524Z

which will generate a local docker image repronim:simple_prep_${USER}_jessie_20150306T060524Z (${USER} will correspond to your user name), with all necessary for computation components installed.

You can then run the demo using the docker image run commands above by replacing repronim/simple_workflow:latest with repronim:simple_prep_${USER}_jessie_20150306T060524Z

4. Other containers

You can also use Singularity to run the docker image from DockerHub. The following commands require Singularity v2.4.3 or later.

singularity build repronim.img docker://repronim/simple_workflow:latest
singularity run -B $PWD/output:/opt/repronim/simple_workflow/scripts/output -c \
  --pwd /opt/repronim/simple_workflow/scripts/ --env CURL_CA_BUNDLE='' repronim.img run_demo_workflow.py \
  --key 11an55u9t2TAf0EV2pHN0vOd8Ww2Gie-tHp9xGULh_dA -n 1
singularity run -B $PWD/output:/opt/repronim/simple_workflow/scripts/output -c \
  --pwd /opt/repronim/simple_workflow/scripts/ --env CURL_CA_BUNDLE='' repronim.img check_output.py --ignoremissing