From conda packages (recommended if possible)

The easiest and often best way to install the Orekit Python Wrapper module in Python is to use pre-made packages for the conda package manager. Mambaforge. The Anaconda distribution is also compatible with these packages but that distribution may have license restrictions.

There are no pip packages for Orekit, which is due to the need of a well-controlled build and execute environment.

Orekit packages is being continuously built in the conda-forge system and can then be installed using conda:

conda install -c conda-forge orekit

This will install all dependencies (including openjdk) and install scripts that sets the necessary path variables (JCC_JDK, PATH) at activation.

Activation of the environment in conda

Note that it is needed to active the conda environment to be used, this activation triggers a script that sets up the right environment variables (JCC_JDK and in some cases PATH). If accessing the python executable directly, for example through an IDE such as PyCharm, the tool conda-wrappers can be used to create executable that runs the activation automatically when starting the python executable. The wrapped python executable is then located under Scrips/wrappers/conda/python.bat for windows. Point your IDE to this file as python interpreter:

Screenshot from PyCharm configured with conda-wrappers to automatically activate the environment

Note: In recent versions of anaconda on Windows it seems like there are some issues if the conda is not "init'ed" on the machine. To do this, run "conda init" in the command prompt.

Physical Data

The physical data is not included in the orekit library and needs to be provided. There are three ways to load the Orekit data:

Download Orekit data Zip

A starting package of data is available from https://gitlab.orekit.org/orekit/orekit-data and can be downloaded through the helper function download_orekit_data_curdir().

The python helper function setup_orekit_curdir() loads such file, with the default filename argument "orekit-data.zip" and sets up orekit to use it (you need to remove the "-master" from the filename if the repo download is used).

Download Orekit data as a library using pip

Alternatively, the Orekit data repo is now installable as a Python library using pip. You can do the following:

• In a terminal with the conda environment activated, type:

pip install git+https://gitlab.orekit.org/orekit/orekit-data.git

• In Python, load data from the orekitdata library using:

setup_orekit_curdir(from_pip_library=True)

Manage your own Orekit data folder

If you want to manage your own Orekit data folder, you can just pass its path to the setup_orekit_curdir method:

setup_orekit_curdir(filename=<PATH TO YOUR OREKIT DATA FOLDER>)

From Source

Installation from source requires to build a jar of the Orekit library including the python wrapper classes. This jar is then fed to the JCC tool, which builds the actual python wrapper together with some helper files. The process is described in the INSTALL.txt file in the repository.

The git repository for orekit is located at:

https://gitlab.orekit.org/orekit-labs/python-wrapper.git

The modified orekit source code with the python additions are available at: https://github.com/petrushy/Orekit

The orekit conda package is automatically built and can be of interest if needed to compile it manually:

https://github.com/conda-forge/orekit-feedstock