-
-
Notifications
You must be signed in to change notification settings - Fork 1.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Decide if obstime transformations of coordinates with velocities should result in motion #6280
Comments
(not that this issue is explicitly referenced in a |
I would be in favor of not updating the position based on the velocity. The velocity vector may be that of a flow/stream rather than a point object that is moving for example. This is not to say that there couldn't be a keyword argument or a separate method to update the positions though. |
If I input Barnard's Star with its Hipparcos position and proper motion, and try to determine where it will be at a give time, then surely by default it should account for the time difference.... More generally, one thing I do very often is to register images to, say, UCAC4 or other position + proper motion catalogues. For this to work generally, one again does want to apply proper motions. I must admit I cannot see the general use case for not taking into account |
What about the case of a binary star? If I have the velocity and proper motion of a star in a binary, any linear extrapolation would be very wrong. |
We could have a required keyword argument for cases like this to avoid any ambiguity? |
Also solar system objects |
But what use would be to transform those proper motions and velocities to another frame with another |
p.s. Yes, we need something like a keyword to tell what should happen. I think the main question is about what should be the default. It is quite possible that it should just be error on anything ambiguous. |
After lengthy conversation with @mhvk and @adrn, we (@adrn, initially) are thinking to implement a I think we also came up with a clearer concept of the core issue here: there's actually two things that change when But it's much trickier to understand how to evolve this at the same time as an |
@eteq - I think for 3.0 we're just erroring if |
Agreed |
- Handled the case where the coordinate instance has multiple representations - Handled the case where coordinates instance doesn't have shape 0 - Do not transform coordinates when coordinate instance is already in required frame(i.e. in a frame parallel to ICRS and centered at attractor) - Added tests for above changes Issue: #434
I ran into the following example today. I am trying to transform the position of the ALMA observatory with velocities to be in the frame of reference of the LSR. Note that the LSRD frame/skycoord has no explicit obstime:
So I wonder if the restriction could be lifted if the target obstime is not set? |
But wouldn't the velocity of ALMA in the LSR depend on the observation time? It would seem that the code is doing the right thing here... |
@mhvk - yes good point that an error is expected but I don't think the error message is correct then - it should mention that the obstime is missing? |
Indeed, the error message is most confusing! In part this is because the |
p.s. The above seems like a bug regardless, but should be a separate issue! |
This issues is a follow-up discussion to an unresolved question from #6219 (in that PR we are punting on the question by raising an error).
The starting point for the topic is #6219 (comment) - @StuartLittlefair's position is that if you transform to a frame with a different
obstime
, a coordinate with velocities should be moved to a different position based on its velocity vector. @StuartLittlefair added more in #6219 (comment) and #6219 (comment) and @mhvk agreed in #6219 (comment)By contrast, I think this is not what we want and will be extremely confusing and inconsistent with the other coordinate machinery - see #6219 (comment). In #6219 (comment) and #6219 (comment) @adrn agreed.
This issue is hence a placeholder, because this is a somewhat significant philosophical question for
coordinates
, so we should resolve it for the next version.The text was updated successfully, but these errors were encountered: