Skip to content

Installation

Yntiglei edited this page Jan 16, 2024 · 26 revisions

Prerequisites

To use Launchcontainers, there are required modules and environment settings. We recommend first, installing the required modules. Secondly, creating a virtual environment for Launchcontainers and installing the required environment settings. We also provide a list of recommended system modules and Python packages that may help you have a better experience on CentOS.


Required

  • Singularity

     type: system module

     version: 3.5.2+

     To download Singularity follow this link: https://docs.sylabs.io/guides/3.0/user-guide/installation.html

  • DWI analysis containers:

     type: singularity images

     anatROIs

     RTP-preproc

     RTP-pipeline

  • Miniconda 3

     type: system module

     miniconda3 is a useful tool to manage Python dependencies and virtual environment

  • Python

     type: environment setting

     version: 3.9-3.11 Stable version: 3.10.9 all the development are done under 3.10.9

Recommended

  • Tree 2.1.0

     type: system module

     tree can help you visualize folder structure and check if the symbolic links have been successfully created

  • Git

     type: system module

     type: system module

  • Heudiconv

     type: system module

     To install heudiconv:

# To install heudiconv container in docker:
docker pull nipy/heudiconv:0.11.6  # version 0.11.6 (not v0.11.6)


# To install the heudiconv container in singularity:
singularity build heudiconv_0.11.6.sif docker://nipy/heudiconv:0.11.6

     Or download heudiconv following this link:https://hub.docker.com/r/nipy/heudiconv

Installation

For beginner and intermediate users

1) create a virtual environment for Launchcontainers

conda create -n yourenvname python=3.10.9

2) installing Launchcontainers using

pip install launchcontainers==0.0.6

For developers

Please refer to Poetry for more information. 'Launchcontainers' follows Poetry packaging example and also uses Poetry to publish to PyPi. In the repo, there are defined pyproject.toml, and 'poetry.lock' that you can play with.

1) create a virtual environment for Launchcontainers

conda create -n yourenvname python=3.10.9

2) git clone this repository locally.

git clone https://github.com/garikoitz/launchcontainers.git

3) use Poetry to download the dependencies you need

poetry shell