Installation

Depending modules

The irfpy.ica module depends on third party modules. These must be installed independently, preferably before:

  • h5py

  • hdf5storage

  • deepdish

  • plumbum

Please install them manually, e.g. if you are using anaconda by:

conda install anaconda
conda install conda-forge::hdf5storage conda-forge::deepdish
conda install conda-forge::plumbum

On Windows this maybe followed (in an anadonda shell) by:

conda install git

Access to the data server data.irf.se

  1. Create folder and key, use empty passphrase for this key

For advanced users you may also use a existing key and copy it to the correct place.

mkdir ~/.ssh
ssh-keygen -f ~/.ssh/icapipelinekey

  1. Send the public key to IRF contact person to enable access to the ICA data repository.

Example (show content of public key file): Copy the key

cat ~/.ssh/icapipelinekey.pub

The text may look like this:

ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILvLKVvY/uxpFt8atxuuohz+V5ciPkQrbwa9F94IUWeI leif@computer

After you have received confirmation of access to IRF ICA repository you can test access.

ssh -p 2200 -i ~/.ssh/icapipelinekey icapy@data.irf.se rsync --list-only /

Output should look similar to this

drwxr-xr-x             17 2024/10/28 12:43:52 .
drwxr-xr-x              4 2016/10/16 18:08:41 aux
drwxr-xr-x              5 2016/09/22 09:45:07 bin
drwxr-xr-x              2 2017/03/06 15:43:52 cops
drwxr-xr-x            725 2020/08/28 11:17:23 dailydata
drwxr-xr-x              6 2019/02/15 12:16:15 lap
drwxrwxr-x              4 2015/01/27 09:42:44 level0
drwxrwxr-x              6 2016/10/17 05:20:37 level1
drwxr-xr-x              6 2017/05/22 19:05:59 level2
drwxr-xr-x              2 2016/03/03 08:42:22 level3
drwxr-xr-x              4 2025/01/13 17:53:42 level4
drwxr-xr-x              8 2016/09/22 09:53:18 lib
drwxr-xr-x              3 2015/09/30 10:14:35 lib64
drwxr-xr-x              7 2020/02/17 22:15:35 mag
drwxrwxr-x              9 2022/10/28 13:39:01 special
drwxr-xr-x              5 2015/08/06 11:07:45 spice

Numpy, Scipy, Matplotlib and sympy

A commonly used numerical and graphic libraries should be installed.

  • numpy

  • scipy

  • matplotlib

  • sympy

The build of the above four libraries are sometimes not simple. I recommend to install them from an OS-level packaging system.

irfpy.ica installation

The latest stable version of irfpy.ica can be installed using:

pip install --no-index --find-links=https://irfpy.irf.se/sdist irfpy.ica

If you want to work with the development version you can install it like this as the module is setuptools compatible :

git clone "git@gitlab.irf.se:irfpy/ica.git"
cd ica
git checkout develop
git pull
pip install -e .

On Windows, the last step of the installation works better with:

pip install .

If the development version changed with or without version number increase you can do:

git pull
pip install . --force

This will get the new version installed.

Author

Main author is Martin Wieser (Swedish Institute of Space Physics).