Installation
Clone the repo
To download the DAP software and associated data, clone the mangadap GitHub repo by executing:
git clone https://github.com/sdss/mangadap.git
This will create a mangadap
directory in the current directory.
Although we try to keep the master
branch of the repository stable,
we recommend using the most recent tag. You can do so by executing:
cd mangadap ./checkout_current_tag
Warning
There is a distribution of the DAP that can be installed via pip, but the installation will be unsuccessful. We’re still working out the bugs…
Install Python 3
The DAP is supported for Python 3 only. To install Python, you can do so along with a full package manager, like Anaconda, or you can install python 3 directly from python.org.
Python environment setup
You are strongly encouraged to build an isolated python environment specifically for the DAP. This can be done using virtualenv:
virtualenv dap
source dap/bin/activate
or conda:
conda create -n dap python=3.9
conda activate dap
See the Virtualenv documentation and/or Managing Environments with Conda for more details. See also virtualenvwrapper as an option for more easily managing virtualenv environments.
Install the DAP code
Once you have cloned the repository, the preferred method to install the DAP and
ensure its dependencies are met is to perform a local pip install. From the
top-level mangadap
directory, run:
pip install -e .
This approach is preferred because it eases uninstalling the code:
pip uninstall sdss-mangadap
If you plan to develop the code, you should install the development dependencies as follows:
pip install -e ".[dev]"
Note
The use of the quotations above is shell dependent; e.g., you need them for zshell, but not for bash. Also beware the exact characters used in the html above may not be the quotation characters you need for the command line (i.e., copy-pasting the line above may throw an error).
Test your installation
To test the installation, make sure you have pytest installed and then
cd mangadap/tests
pytest . -W ignore
Some tests require a set of “remote” data that are not located in the repo for space considerations. You can download these data by running the following script:
python download_test_data.py
And then executing the tests with the same commands above.
Note
If you have SDSS Collaboration Access and you have a ~/.netrc file (e.g., from using collaboration access with Marvin), this script will pull the data from the (still public) mangawork directory on the SAS.
Problems?
If you have problems, particularly those that you think may be a more general problem, please Submit an issue.