Skip to content

depaul-dice/sciunit

Repository files navigation

PyPI

Sciunit

The most updated instructions on installing Sciunit can be found at: https://sciunit.run/install

sciunit is a command-line tool that answers the call for a reusable research object that containerizes and stores applications simply and efficiently, facilitates sharing and collaboration, and eases the task of executing, understanding, and building on shared work.

Installing

This section introduces the binary distributions of sciunit.

Requires: pip>=8.1.1, glibc>=2.17, python2.7 with headers and compiler

Platform: amd64-pc-linux-gnu

Ensure that ~/.local/bin is in your PATH environment variable, and then, run

pip2 install --user sciunit2

Add --upgrade to the command line arguments to upgrade an existing sciunit installation.

[ Note Attempt to install sciunit with "pip install sciunit" will get you something else; our PyPI project is registered as "sciunit2". --end note ]

Verify the installation with

sciunit --version

In the following per-distro instructions, if you see a command-line in the form of

export PATH=~/.local/bin:$PATH

, it merely means that you will need the aforementioned binary path in PATH before proceeding; normally you should adjust it according to your login shell and write it to a shell configuration such as ~/.profile.

Ubuntu 16.04

sudo apt update
sudo apt install python-dev python-pip
pip2 install --user sciunit2

Ubuntu 14.04

sudo apt update
sudo apt install python-dev python-pip
pip2 install --user -U pip
export PATH=~/.local/bin:$PATH
pip2 install --user sciunit2

Arch Linux

sudo pacman -S python2-pip
export PATH=~/.local/bin:$PATH
pip2 install --user sciunit2

Fedora 26

sudo dnf install python2-pip python2-devel gcc redhat-rpm-config
pip2 install --user sciunit2

CentOS 7

sudo yum install -y epel-release
sudo yum install python2-pip python-devel gcc
pip2 install --user sciunit2

RHEL 7

sudo yum install -y https://dl.fedoraproject.org/pub/epel/epel-release-latest-7.noarch.rpm
sudo yum install python2-pip python-devel gcc
pip2 install --user sciunit2

Debian 9

sudo apt update
sudo apt install python-dev python-pip
export PATH=~/.local/bin:$PATH
pip2 install --user sciunit2

Debian 8

sudo apt update
sudo apt install python-dev python-pip
pip2 install --user -U pip
export PATH=~/.local/bin:$PATH
pip2 install --user sciunit2

OpenSUSE 42

Note that OpenSUSE defaults user-binary path to ~/bin rather than ~/.local/bin, so you may want to symlink one to the other.

sudo zypper install python-pip python-devel gcc
pip2 install --user -U pip
export PATH=~/.local/bin:$PATH
pip2 install --user sciunit2

Build from Source

Instructions for custom build of sciunit2 is available in our Bitbucket project repository. You may want to check it out if you are running on a system that is not supported by the binary distribution.

Post-install

Execute

sciunit post-install

as a normal user to add command-line completion support for sciunit to your login shell. When running this or any other sciunit command, if your environment is not listed in the Installing section, or you missed some prerequisites before issuing pip, you may observe the following error:

ImportError: No module named _bsddb

Fortunately, there are more than one ways to fulfill this dependency. Try to find and install a package often named "python-bsddb" or "python-bsddb3" with your system package manager. In Anaconda, you will need a port called "bsddb."

To make the command-line completion work, if you are a bash users, make sure that you have the "bash-completion" package installed on your system. If you are a tcsh user, make sure that you have source ~/.complete in ~/.cshrc or ~/.tcshrc as suggested by the stock completion script.

Using

Checkout our tutorial and manpage. The manpage is available as

man sciunit

after a successful installation.

Sciunit acknowledges support from the National Science Foundation, Bloomberg Foundation, and DePaul University.