Installation

You can install dysmalpy in a number of ways:

  1. Pip install: The quickest and simplest way to install dysmalpy is from the pypi repository using pip. This will install the latest stable version and its dependencies.

  2. Using anaconda: Installing dysmalpy using the Anaconda Python distribution.

    2.1 Using pip to install the dependencies (default)

    2.2 Using conda to install the dependencies

  3. Developement version: If you are more familiar with Python, you can install dysmalpy using the development version.

  4. MPE group installations: If you are a member of the MPE group, you can install dysmalpy using the provided .bat file or by using the Anaconda environment we have setup.

1. Pip install

To install the latest stable version of dysmalpy and its dependencies, you can use pip.

Warning

Please note that the C++ extensions are not included in the pypi package and you will need to install them separately. You can find the instructions on how to install them (and test if they can be built) in the Python Environment Setup section.

Once you have installed the extensions, you can install dysmalpy using pip:

python -m pip install dysmalpy

2. Using Anaconda

To install Anaconda and all the relevant packages and dependencies, please follow the instructions at Python Environment Setup.

After (and only after) your anaconda installation is complete. You can download the latest dysmalpy package here: tar.gz | zip (current version: 2.0.0).

From a terminal, change directories to where the package was downloaded, decompress it and navigate inside the directory with:

tar zxvf dysmalpy-N.N.N.tar.gz
cd dysmalpy-N.N.N

(Where N.N.N is the current version)

Note

You can also clone the repository from GitHub with (make sure you have git installed with e.g. conda install git):

git clone https://github.com/dysmalpy/dysmalpy.git

Now, you can install dysmalpy and its dependencies in one of the following ways:

2.1 Using pip to install the dependencies (default)

You can use pip to install dysmalpy (and its corresponding dependencies):

python -m pip install .

Alternatively, you can use the following command:

python setup.py install --single-version-externally-managed --root=/

Note

If you would like to specify your specific include and library paths you can set them by defining the C_INCLUDE_PATH and LD_LIBRARY_PATH environment variables like this:

export C_INCLUDE_PATH=/your/include/path
export LD_LIBRARY_PATH=/your/lib/path

then you can run the usual installation command.

2.2 Using conda to install the dependencies

If you have chosen to install the python dependencies with conda (following the steps at Python Environment Setup), please use the following command:

python -m pip install . --no-deps

Note

The --no-deps flag will prevent pip from installing the dependencies that you have already installed with conda in (Python Environment Setup).

3. Development version

To get the most up-to-date development version of dysmalpy, clone the repository from GitHub.

Within your desired parent directory, clone the repository from GitHub:

git clone https://github.com/dysmalpy/dysmalpy.git

Then add the path to this repository to your python path (e.g., PATH/TO/PARENT/DIRECTORY/dysmalpy to $PYTHONPATH as defined in .bashrc or .bash_profile for bash, or the equivalent for your shell).

In order to compile the C++ extensions locally please run:

python setup.py build_ext --inplace

For examples on using git fetch or git pull to get updates, or how to check out other branches, please see e.g. the tutorial here: https://git-scm.com/docs/gittutorial

Warning

This way of installing dysmalpy will give you a basic installation with most of the functionality, but your dysmalpy installation will not contain the other modules that you may need.

After the installation is complete, you should be able to run import dysmalpy within IPython or your Jupyter notebook.

Tip

Especially if working with the development version of dysmalpy, you can confirm the location of the package that is imported by checking the output of

import dysmalpy
print(dysmalpy.__file__)

4. MPE group installations

Windows bat File

A bat file, for running DysmalPy with a parameters file (e.g., fitting.params) is available for MPE-group specific architecture.

Prior to using DysmalPy with this bat file, the DysmalPy source code and dependencies will need to be installed (see Using Anaconda).

AFS Machine

If you are on an AFS machine, dysmalpy is located at /afs/mpe.mpg.de/astrosoft/dysmalpy. We have further setup an Anaconda environment the contains all of the necessary Python packages to run dysmalpy. To activate this environment as well as set environment variables, run this command in your terminal:

source /afs/mpe/astrosoft/dysmalpy/dysmalpy_setup.sh

To check whether the setup ran successfully run:

which python

This should return /afs/mpe.mpg.de/astrosoft/dysmalpy/anaconda3/bin/python. Keep in mind that using this environment will override any environment you have setup locally and only Python packages installed in the dysmalpy environment will be available. If there is a package you would like installed in the environment, please contact the maintainers at dysmalpy@mpe.mpg.de.

For those who are more familiar with Python, you can simply add /afs/mpe.mpg.de/astrosoft/dysmalpy/dysmalpy/ to your PYTHONPATH environment variable. Be sure to have all of the required packages installed.