ndcube requires Python 3.5+, SunPy 0.9+, astropy and matplotlib.

Installing the Stable Version

There are two options for installing the stable version of ndcube. The first is via the anaconda distribution using the conda-forge channel. For more information on installing the anaconda distribution, see the anaconda website.

$ conda install --channel conda-forge ndcube

To update ndcube do:

$ conda update ndcube

The second option for installing the stable version of ndcube is via pip.

$ pip install ndcube

Then to update ndcube do:

$ pip install ndcube --upgrade

Installing the Development Version

The stable version of ndcube will be reliable. However, if you value getting the latest updates over reliablility, or want to contribute to the development of ndcube, you will need to install the development version via GitHub. Let’s step through how to do this using anaconda. For information on installing the anaconda distribution, see the anaconda website.

First, create a conda environment on your local machine to hold the ndcube bleeding edge version. Using a new environment allows you to keep your root environment for stable package releases. Let’s call the new conda environment ndcube-dev. Type the following into a terminal:

$ conda config --append channels conda-forge
$ conda create -n ndcube-dev sunpy hypothesis pytest-mock pip sphinx coverage ipython jupyter

Be sure to activate the environment, i.e. switch into it. In Linux or MacOS, type:

$ source activate ndcube-dev

In Windows, type:

> activate ndcube-dev

Next clone the ndcube repo from GitHub to a new directory. Let’s call it ndcude-git.

$ git clone ndcube-git

To install, change into the new directory and run the install script.

$ cd ndcube-git
$ pip install -e .

Voila! The ndcube development version is now installed! Be sure you get the latest updates by regularly doing:

$ git pull origin master