Skip to content

Installation

lightcurver requires several components to function:

  • STARRED: it will be installed automatically with pip install lightcurver, but if you want to use it with a GPU there might be some more setup to do. See the installation instructions of the package itself.
  • The dependencies handled by your python package manager, such as astropy, shapely, astroquery, pandas, pyyaml ...these will be installed automatically by pip install lightcurver.
  • (optional) Astrometry.net: their installation instructions should get you started. Alternatively, you can get an API key from their nova service. I would recommend against using it in production, as to not overload their servers.

So, I would suggest creating a python (3.9+, ideally 3.11) environment, say lightcurver_env, and install the present package in it.

The quick version

Chances are this will work: 

    conda activate lightcurver_env # if using conda
    source lightcurver_env/bin/activate # if using python's venv
    pip install lightcurver

Or for the git version (includes some minimal test data): 

    git clone git@github.com:duxfrederic/lightcurver.git
    cd lightcurver
    conda activate lightcurver_env
    pip install .

If the quick version fails: list of dependencies

Should the above fail, there might be a dependency problem requiring the manual handling of the different packages. Here is the list of dependencies that need be installed:

  1. numpy < 2.00 - as of June 2024, sep is not compatible with numpy >= 2.00
  2. scipy
  3. matplotlib
  4. pandas
  5. astropy
  6. astroquery - for querying Gaia and VizieR
  7. h5py - for storing cutouts and PSF models
  8. photutils - for aperture photometry used as initial guess
  9. astroalign - for finding transformations between frames
  10. shapely - for computing footprints of frames
  11. ephem - for calculating airmass, moon distance, etc.
  12. pytest - for executing the automatic tests
  13. sep - for source and background extraction
  14. astroscrappy - for cleaning the cosmics
  15. pyyaml - for reading the config file
  16. starred-astro - assume the latest version, will install its own dependencies.
  17. widefield_plate_solver - an astrometry.net wrapper
  18. ccdproc - for identifying bad columns of pixels

Testing your installation

You can test your installation by following the tutorial. The automated tests also include the processing of a subset of the dataset given in the tutorial, you can thus run them instead to check functionality (should take 1-2 minutes). 

cd /your/clone/of/lightcurver
pytest .

If you are going to use a local installation of Astrometry.net, do not forget to download their index files as well! The combination of 4100 and 5200 should do the trick.