feat: update to_Vector3D to pass new coordinate values

[Naman-Priyadarshi]

Description

Kindly take a look at CONTRIBUTING.md.

Please describe the purpose of this pull request. Reference and link to any relevant issues or pull requests.

Checklist

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't any other open Pull Request for the required change ?
• Does your submission pass pre-commit? ($ pre-commit run --all-files or $ nox -s lint)
• Does your submission pass tests? ($ pytest or $ nox -s tests)
• Does the documentation build with your changes? ($ cd docs; make clean; make html or $ nox -s docs)
• Does your submission pass doctests? ($ xdoctest ./src/vector or $ nox -s doctests)

Before Merging

• Summarize commit messages into a brief review of the Pull request.

[Naman-Priyadarshi]

Making changes to test_conversions.py now

[codecov-commenter]

Codecov Report

Base: 82.75% // Head: 82.72% // Decreases project coverage by -0.04% ⚠️

Coverage data is based on head (cf850b1) compared to base (8f25cbc).
Patch coverage: 53.84% of modified lines in pull request are covered.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #278      +/-   ##
==========================================
- Coverage   82.75%   82.72%   -0.04%     
==========================================
  Files          96       96              
  Lines       11354    11366      +12     
==========================================
+ Hits         9396     9402       +6     
- Misses       1958     1964       +6     

Impacted Files	Coverage Δ
src/vector/_methods.py	74.15% <53.84%> (-0.17%)	⬇️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[Saransh-cpp]

This is how it works right now -

In [1]: import vector

In [2]: v = vector.obj(x=1, y=2)

In [3]: v.to_Vector3D(z=1)
Out[3]: VectorObject3D(x=1, y=2, z=1)

In [4]: v = vector.obj(px=1, y=2)

In [5]: v.to_Vector3D(z=1)
Out[5]: vector.MomentumObject3D(px=1, py=2, pz=1)

Should we also have the option to enter the momentum counterparts of coordinates? Or should we leave it as it is?

[jpivarski]

This looks like a good thing to add, except that if it's on Vector2D.to_Vector3D, users will expect it to be on 2D → 4D and 3D → 4D as well. @Saransh-cpp, how much effort do you think it would be to add them?

Also, a new feature like this should have tests. Like, one example for each of z, theta, and eta.

[henryiii]

You can set this to something like l_type: type[Longitudinal] = LongitudinalZ or something like that, IIRC, to avoid the type ignore.

[Saransh-cpp]

I did try this (in fact, IIRC, this was added by you in one of the commits above, and I removed it to make the CI pass), but mypy still required the type ignore statements :(

Oh wait, that worked, maybe I was trying something else previously.

[henryiii]

users will expect it to be on 2D → 4D and 3D → 4D as well

I was thinking that as well. Though maybe asking people to do 2D -> 3D -> 4D would be easier than adding it to the 2D -> 4D one. But 3D -> 4D does clearly make sense.

[jpivarski]

2D → 4D could be implemented in terms of 2D → 3D → 4D.

Also, does this only apply to the single-object backend (vector.obj), or does it generalize to all backends? Setting coord_value to a scalar 0 might work on array-based backends if it gets broadcasted.

Also, the distinction between 0 and 0.0 is relevant: coordinates are allowed to be any numeric type, including integers (though that's less often desired).

[Saransh-cpp]

Thanks for the review, @jpivarski and @henryiii! Could you also please take a look at #317 and let me know if this should go in v1.0.0?

---

users will expect it to be on 2D → 4D and 3D → 4D as well. @Saransh-cpp, how much effort do you think it would be to add them?

Also, a new feature like this should have tests. Like, one example for each of z, theta, and eta.

Should be simple to add. I can maybe merge this PR and add the tests + make it available for 2D → 4D (2D → 3D → 4D) and 3D → 4D in another PR.

Also, does this only apply to the single-object backend (vector.obj), or does it generalize to all backends? Setting coord_value to a scalar 0 might work on array-based backends if it gets broadcasted.

Generalizes to all the backends!

In [1]: import vector

In [2]: v = vector.array(
   ...:     [(1.1, 2.1), (1.2, 2.2), (1.3, 2.3), (1.4, 2.4), (1.5, 2.5)],
   ...:     dtype=[("x", float), ("y", float)],
   ...: )

In [3]: v.to_Vector3D(z=1)
Out[3]: 
VectorNumpy3D([(1.1, 2.1, 1.), (1.2, 2.2, 1.), (1.3, 2.3, 1.), (1.4, 2.4, 1.),
               (1.5, 2.5, 1.)], dtype=[('x', '<f8'), ('y', '<f8'), ('z', '<f8')])

In [4]: v = vector.Array(
   ...:     [
   ...:         [{"x": 1, "y": 1.1}, {"x": 2, "y": 2.2}],
   ...:         [],
   ...:         [{"x": 3, "y": 3.3}]
   ...:     ]
   ...: )

In [5]: v.to_Vector3D(z=1)
Out[5]: <VectorArray3D [[{x: 1, y: 1.1, z: 1, ... z: 1}]] type='3 * var * Vector3D["x": ...'>

[henryiii]

I think this would be nice for a v1. Merge then add things is fine.

How much work is it to keep Awkward v1? If not much, then I'd say we don't need to rush to drop it, it will make adoption of Vector v1 easier. But fine to drop if it is.

[Saransh-cpp]

Thanks! Will merge this, create a new PR, and set up stuff for v1!

[Saransh-cpp]

Thanks for taking this up, @Naman-Priyadarshi!