These are chat archives for astropy/astropy

1st
Aug 2017
matrixbot
@matrixbot
Aug 01 2017 16:21
`TreyHarris on Freenode` Hi--I'm hoping an expert can help out a dumb layman who just needs something for a game I'm working on. This is one of those practical "if you've done it, it's easy, if not, it's hard" things I think: anyone know how to use catalogue data to determine how far away from us a star will be X years in the future? I've read astro textbooks, and I know the method of converting RA/dec/parallax to Cartesian, then
`TreyHarris on Freenode` converting R-v and proper motion int Cartesion vectors, then applying them and reversing the whole thing to get a new RA/dec/parallax. But seeing how many astro-related libs there are, I bet there's a python module for this? I actually just need distance at year M in the future, where M < 2000. I can do this purely numerically using the method above, but it's cumbersome (particularly since several values in
`TreyHarris on Freenode` published catalogues are the wrong units or referenc poitns for this calculation so need conversion before I can even do the simple Cartesian vector work).
`TreyHarris on Freenode` I skimmed the astropy tutorial and examples, and it looks like it can probably help me, but I don't know real astronomy well enough to figure out how.
matrixbot
@matrixbot
Aug 01 2017 16:29
`TreyHarris on Freenode` (Also, throw in error values and especially error ellipses, and I'm truly lost. I don't even know whether, because I'm talking about objects close to us--specifically, ones less than M ly away for ever M years in the future, with M < 2000, that means the errors aren't that important because the time period is short and the errors are small, or whether I do need to pay attention to margins of error precisely
`TreyHarris on Freenode` because they're so close the possible sky movement is greater....)
matrixbot
@matrixbot
Aug 01 2017 16:55
`TreyHarris on Freenode` Urgh, working this brute-force, I can see that this is hard because stars may get closer to us for some period of the next 2000 years, then move farther away, so unlike a simple margin of error where I could work out an overall margin of error for the distance-to-us scalar, these error ellipses mean I end up with a complex answer with an error ellipsoid, don't I? The idea in the game--and I can precompute as
`TreyHarris on Freenode` much as necssary and store it--is, assuming a signal goes out from here at a given time, whether a given star "has seen it" by some later time or not.
`TreyHarris on Freenode` So I guess I need a two-dimensional array of answers to the question, "the minimum possible distance of the star to us between now and time M in M between now and ~4017 CE", where the number of subdivisions is arbitrary, lesser or greater depending on how much accuracy I want (if years, 2000 of them, for instance), and if given array-lookup function d, d(M) < M × c for any M <= Y, where Y is the future year
`TreyHarris on Freenode` I'm asking about, then the signal could have reached it. (The optimization of making the second column of the array strictly increase--i.e., once the star has reached periapsis to our Sun, stop calculating--would be valid if all signals in the game left at M=0, but that's not the case, so I can't make that optimization.) Since it's just a game, I could just ignore the errors entirely and pretend some
`TreyHarris on Freenode` catalog's values are exactly correct, I guess...
`TreyHarris on Freenode` (And I'm sure "periapsis" isn't the correct term when talking about the sun and a faraway star--I meant "closest approach", whatever the term is.)
matrixbot
@matrixbot
Aug 01 2017 17:06
`TreyHarris on Freenode` Hmm... is the silence an indication that this is the wrong channel for this? If so, can someone direct me somewhere more appropriate? ##astronomy seems too general... I also tried #pyastro before seeing it was a channel created for a conference that's already happened...
matrixbot
@matrixbot
Aug 01 2017 17:20
`equant on Freenode` Sometimes replies take a long time in here, but considering the length of your question, it seems like there has to be a Stack Exchange that would be good for you.
`TreyHarris on Freenode` equant: I kept adding stuff as I was hacking away at the brute-force solution and thought maybe the additional context might help. But yes, I'll got to stack exchange if IRC can't help me.
matrixbot
@matrixbot
Aug 01 2017 17:29
`TreyHarris on Freenode` equant: would the astronomy.stackexchange.com be the better place for this than a programming-related one, do you think?
matrixbot
@matrixbot
Aug 01 2017 18:22
`TreyHarris on Freenode` I'll give it a try. Thanks.