Deprecate neighbor_iter in favor of iter_neighbors

[rht]

See #980.

[codecov]

Codecov Report

Merging #1184 (b1fab1d) into main (68b6631) will decrease coverage by 0.29%.
The diff coverage is 0.00%.

❗ Current head b1fab1d differs from pull request most recent head bcb71ed. Consider uploading reports for the commit bcb71ed to get more accurate results

@@            Coverage Diff             @@
##             main    #1184      +/-   ##
==========================================
- Coverage   89.38%   89.09%   -0.30%     
==========================================
  Files          19       19              
  Lines        1300     1302       +2     
  Branches      264      264              
==========================================
- Hits         1162     1160       -2     
- Misses        100      104       +4     
  Partials       38       38              

Impacted Files	Coverage Δ
mesa/space.py	93.88% <0.00%> (-1.09%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 68b6631...bcb71ed. Read the comment docs.

[rht]

Notice I intentionally didn't use DeprecationWarning, because when I tested this warn, it didn't show up when I used DeprecationWarning. It will show up as UserWarning.

[tpike3]

I concur this makes the methods consistent. Action (i.e. get or iter) and then target (i.e. neighbors or neighborhood).

        get_neighbors: Returns the objects surrounding a given cell.
        get_neighborhood: Returns the cells surrounding a given cell.
        iter_neighbors: Iterates over position neighbors.
        iter_neighborhood: Returns an iterator over cell coordinates that are
            in the neighborhood of a certain point.

Based on the diversity of opinions in #980 and as we are only a week away from the dev session, to ensure inclusivity we can bring up for concurrence.

Could you please squash your commits?

[rht]

Squashed.

[tpike3]

@jackiekazil Can you due to the debate on this subject can you give this pull request an additional review. I concur with @rht approach.

[rht]

I think this PR is not as controversial as #980. The iter_* vs get_* is not included in this PR. This PR only removes the neighbor_iter, which I haven't seen anyone arguing the case for it.

[rht]

I'm planning to renew #835, but this work depends on getting this PR merged, because I can't cause a merge conflict by unnecessarily annotating neighbor_iter.

[jackiekazil]

LGTM

[jackiekazil]

Thank you for your work on this @rht

[EwoutH]

Could it be that .iter_neighbors() now returns a Coordinate (Tuple(Int, Int)) instead of an Agent? If so, is that intended behaviour with the switch from .neighbor_iter() to .iter_neighbors() in Mesa 2.0?

[rht]

It still return agents:

mesa/mesa/space.py

Line 722 in 7f5a32e

return self.iter_cell_list_contents(neighborhood)

.

[EwoutH]

Ah I accidentally replaced .neighbor_iter() with .iter_neighborhood() instead of .iter_neighbors().

I also see where I went wrong, it's stated that way in the 2.0.0 release notes.

https://github.com/projectmesa/mesa/blob/7f5a32e63db93b9be2b00a705d0c28a66bc5a9dd/HISTORY.rst?plain=1#L50C13-L50C13

I opened a PR to fix this: #1807.