I'm working on a game engine of sorts, and one thing I did for a previous game which was under the GPL was I had a night sky with the constellations visibly connected and I'd like to do that again. I don't know where my data set originally came from but I do have a note that it was GPL, so it likely came from Stellarium, which is GPL licensed. For this I need the dataset to come from the public domain, us government, permissively licensed, etc.

If possible I'm after the coordinates in the sky, brightness, maybe class and name, but also some mechanism to know which stars would be connected in classic constellations.

I've found some data sets which contain the star data I'm after, though they would take some effort to parse and pull out the data I'm after, but I can't seem to find anywhere online, except stellarium, which provides star constellation connectivity. Any ideas?

I've gotten this list so far from what I can tell:

Are there any standards of what stars make up the primary connections in constellations? Or does it just happen to be whoever puts it together?

P.S. If I ever do succeed, I will release the database (probably 3-4k stars is all I will need with connectivity) under the MIT license to permit inclusion in any commercial or noncommercial projects.


2 Answers 2


Much to my surprise, the original constellation author for Stellarium has agreed to permit use of their constellationship maps to non-GPL applications! See the comment here: https://github.com/Stellarium/stellarium/issues/790 this means it's free to use for other people who stumble across this thread, too!

Just FYI - the mapping, found here: https://github.com/Stellarium/stellarium/blob/master/skycultures/western/constellationship.fab is based on the HIP database of stars, so you will need to parse that, but that, being published by ESA is also a permissive database.

Thank you, everyone here for all the input you gave, as it may help others doing similar tasks - some of the solutions provided may be better suited for others who are willing and interested in putting in more effort. I hope this thread helps out anyone else trying to do what I'm doing!


I have an idea how to do this, but it requires a few steps:

  1. Yes, the IAU set down which connections make up the classical constellations, but I didn't see an easy way there to get their data.
  2. Wikipedia has a list of stars by constellation. The stars there per list are many more than you need, BUT: the primary connectors of the constellation will always be made up by the brightest stars in a constellation, which will be the first entries in each list.
  3. To access each list's first entries, I would recommend to modify the simple script from this question's accepted answer, which accesses a website and then datamines it for the interesting stuff.
  4. Now not every constellation has the same amount of stars, but probably mining the 20 brightest ones each should be enough (brightness designators are the greek alphabet in order, $\alpha$ for the brightest, $\beta$ for second and so on.).
  5. Then adding the connections: I don't have a good idea for that one. Possibly you have to bite the bullet there, grab a bottle of wine in the evening, and mark the connections by hand. 88 times 5-10 pair connections. Depending on the efficiency of how you mark the connectors in your database, you could be done in 1-2 hours. Still much faster than without any preselection. Should no other answer cover this step in a more efficient way, at least you could now get back to the IAU website and get the official connectors.
  • 1
    $\begingroup$ Actually, as the IAU page you quoted notes, "In star maps it is common to mark line patterns that represent the shapes that give the name to the constellations. However, the IAU defines a constellation by its boundary (indicated by sky coordinates) and not by its pattern and the same constellation may have several variants in its representation.". So there is no "official" list of figures, just areas of the sky which constitute constellations. $\endgroup$
    – user21
    Commented Oct 8, 2019 at 15:36
  • $\begingroup$ @barrycarter: Thanks for pointing that out! $\endgroup$ Commented Oct 8, 2019 at 16:17
  • $\begingroup$ This is an interesting approach and I appreciate the detail you've given. Using Wikipedia to do the star preselection was a great idea! I feel like there's such an art to choosing how to connect the stars to make the constellations, so it makes me nervous to try to do it myself. I think it would take me much longer than 1-2 hours, but the prospect of building the first truly public constellation connectivity map is very appealing. I am nervous to say "accepted" and want to give more time to see if someone else found a proper repository. $\endgroup$ Commented Oct 8, 2019 at 20:12
  • $\begingroup$ Do you know where the constellation line data for something like this one came from: en.wikipedia.org/wiki/Lacerta#/media/File:Lacerta_IAU.svg $\endgroup$ Commented Oct 8, 2019 at 20:14
  • $\begingroup$ @CharlesLohr: When you scroll on the IAU page down to "Charts and tables" you will see detailed star charts for each constellation. That's presumably the source of those maps. For example, for Andromeda you'll see that you have pairs $\gamma-\beta$, $\beta-\pi$, $\beta-\mu$, etc. Most of the pairs are in those maps. So once you have the preselected data, you can build a function 'add_pair(star1, star2)' in whichever language you're using. Or, of course someone provides a less clunky solution. $\endgroup$ Commented Oct 8, 2019 at 20:47

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .