Skip to contents

Installation

installations.Rmd

This page walks through the installation of ModelArray package and necessary dependent packages/libraries. If you have difficulty installing necessary packages (e.g., on HPC clusters), you also have an option to use the container image of ModelArray + ConFixel. Please refer to vignette("container") for more.

We’ll directly install ModelArray from GitHub. Please notice that, this is only fully tested and supported on Linux Ubuntu system and macOS system. If you are a Windows user, you will face errors when running with more than one CPU cores (n_cores = 2 or above). You may install a Linux subsystem or virtual machine on your Windows computer; or using the container image of ModelArray and running on High Performance Computing (HPC) clusters (see vignette("container") for more).

We will first set up the conda environment, then install some dependent packages and libraries, and finally install ModelArray.

Set up a conda environment

We first create a conda environment modelarray for installing the companion software ConFixel etc. We’ll install python version 3.9:

foo@bar:~$ conda create --name modelarray python=3.9
foo@bar:~$ conda activate modelarray

Install MRtrix (Only required for fixel-wise data)

When converting fixel-wise data’s format (.mif), ConFixel uses function mrconvert from MRtrix, so please make sure MRtrix has been installed. It can either be installed via conda in this conda environment we just created, or be compiled from source. See MRtrix’s webpage for more. Type mrview in the terminal to check whether MRtrix installation is successful.

If your input data is voxel-wise data, you can skip this step.

Install HDF5 libraries in the system

Because ModelArray works with the Hierarchical Data Format 5 (HDF5) file format, we need to make sure necessary libraries of HDF5 are installed in the system.

On a Linux Ubuntu system

If you’re on Linux Ubuntu system: First, please check if libhdf5-dev has been installed in the system:

foo@bar:~$ ldconfig -p | grep libhdf5*

If you got more than one line of outputs, congrats, you have libhdf5-dev installed. Otherwise, please install it via:

foo@bar:~$ sudo apt-get update -y
foo@bar:~$ sudo apt-get install -y libhdf5-dev

On a macOS system

Use Homebrew to install:

foo@bar:~$ brew install hdf5

For details you may refer to the webpage here

Install ConFixel python package from GitHub

ConFixel provides file format conversion for both fixel-wise data (.mif) and voxel-wise data (NIfTI). Follow the commands below to install it from GitHub:

# We first activate the conda environment we just created:
foo@bar:~$ conda activate modelarray  

# Then install ConFixel:
foo@bar:~$ cd ~/myProject
foo@bar:myProject$ git clone https://github.com/PennLINC/ConFixel.git
foo@bar:myProject$ cd ConFixel
foo@bar:myProject$ pip install .

# You may remove the original source code if you are an end user instead of a developer:
foo@bar:myProject$ cd ..
foo@bar:myProject$ rm -r ConFixel

Install R

ModelArray requires R >=4.1.2, and we currently only tests ModelArray on R 4.1.2. If you have already installed it, you can skip this step. If you don’t, you may download it from CRAN.

(Optional) Install RStudio

RStudio provides a good IDE for using R. However it’s optional to install RStudio. If you haven’t got one but want to install it, you may download it from here.

Install ModelArray R package from GitHub

ModelArray R package’s source code is available on GitHub. To install it in R:

# First, load library "devtools":
library(devtools)

# if you got error, it means you don't have devtools installed, please install it by: install.packages("devtools")

# Then, install ModelArray:
devtools::install_github("PennLINC/ModelArray")

Now, ModelArray is ready to use:

(Optional) Other potential packages only needed for downloading demo data in vignette("walkthrough") page

There are several potential packages only needed for downloading demo data in vignette("walkthrough") page. You may skip this step if you will not use the demo data.

  • wget
  • tar

If you don’t have them, please install them first. For macOS system, you may try out brew to install them. For Linux Ubuntu system, you may try out sudo apt-get to install them.