v1.2.0 Taylor Release

[jackiekazil]

Last release was v1.1.1 (2022-10-10) Safford
^^ This was the last release.

Prep v1.1.2 Release

• Collapse milestones into one (I think)... we have two milestone for this release. https://github.com/projectmesa/mesa/milestone/32?closed=1 & https://github.com/projectmesa/mesa/milestone/34?closed=1. Let's make the first one the one milestone to release and move the others over. #1600
• Tag with milestone - @rht, @tpike3 has everything been tagged since the last release with this milestone https://github.com/projectmesa/mesa/milestone/32
• Update history
• release

[rht]

There are lots of recent PRs that are not tagged (https://github.com/projectmesa/mesa/pulls?page=1&q=is%3Apr+no%3Amilestone+is%3Aclosed). Though I forgot the reason why all of them have to be tagged to a milestone for a release. I thought issues in milestone are reserved for only the ones that are necessary and are blocking the release itself.

[jackiekazil]

@rht I had always done it as a tracking mechanism to try to correlate with history -- it is mostly focused on PRs so they are easy to find.

[rht]

We could collect them all by the timestamps (all recent merged PR's after the release PR) and put them in the HISTORY.rst. I don't see having to find the release tag of a PR to be a common use case. But I'm fine with having to tag the recent merged PRs as usual too.

[jackiekazil]

@rht I think what you are suggesting is an automated process. We talked about that... let's revisit setting that up. I am going to do a manual tag for now before we switch to something new.

[jackiekazil]

@rht, @tpike3 can either of you do a review of everything I tagged?
I wanted to pull in the errors too since they can impact things even though they were undone.
If there is something that shouldn't be tagged, please untag.
https://github.com/projectmesa/mesa/milestone/32?closed=1.

[jackiekazil]

@rht - do you have any preference of history format?

[rht]

There are old PRs/issues that have been accidentally tagged in https://github.com/projectmesa/mesa/milestone/32?closed=1. I have untagged them.

One feature with associating PRs with a milestone of a release is that you can track which release a PR is a part of, from its milestone. But this info is not used that often. Other projects use them for blocking issues:

• https://github.com/psf/black/issues?q=is%3Aopen+is%3Aissue+milestone%3A%22Release+23.1.0%22
• https://github.com/pandas-dev/pandas/issues?q=is%3Aopen+is%3Aissue+milestone%3A1.5.4
• https://github.com/numpy/numpy/issues?q=is%3Aopen+is%3Aissue+milestone%3A%221.25.0+release%22

I have no preference for the history format (at the moment).

[jackiekazil]

Currently on a plane, but unable to push PRs. Here is a draft on history -- Feel free to this directly & comment generally on the changes. Also, what do we want to highlight in the Special notes?

---

1.2.0 (2023-03-09) Taylor
++++++++++++++++++++++++++

Special notes

New features:

• Implement radius for NetworkGrid.get_neighbors Implement radius for NetworkGrid.get_neighbors #1564

Some highlights for the perf improvements:

• Use getattr for attribute strings in model data collection Use getattr for attribute strings in model data collection #1590 this is a 2x speedup over the relevant line, but not sure if compared to the rest of the function
• Refactor iter/get_cell_list_contents methods Refactor iter/get_cell_list_contents methods #1570 at least 1.3x speedup for iter/get_cell_list_contents
• Evaluate empties set more lazily Evaluate empties set more lazily #1546 (comment) ~1.3x speedup for place_agent, remove_agent, and move_agent

Improvements

• ci: Add testing on Python 3.11 ci: Add testing on Python 3.11 #1519
• Remove auto-update GH Actions for Pipfile.lock Remove auto-update GH Actions for Pipfile.lock #1558
• ruff
  • ruff: Add isort ruff: Add isort #1594
  • ci: Replace flake8 with Ruff ci: Replace flake8 with Ruff #1587
  • ruff: Add more rules based on Zulip's config ruff: Add more rules based on Zulip's config #1596
• perf: faster is_integer function for common cases perf: faster is_integer function for common cases #1597
• Remove _reporter_decorator Remove _reporter_decorator #1591
• Change index at DataFrame creation in get_agent_vars_dataframe Change index at DataFrame creation in get_agent_vars_dataframe #1586
• Make Grid class private Make Grid class private #1575
• Make the internal grid and empties_built in Grid class private Make the internal grid and empties_built in Grid class private #1568
• Simplify code in ContinuousSpace Simplify code in ContinuousSpace #1536
• Improve docstrings of ContinuousSpace Improve docstrings of ContinuousSpace #1535
• Simplify accept_tuple_argument decorator in space.py Simplify accept_tuple_argument decorator in space.py #1531
• Enhance schedulers to support intra-step removal of agents Enhance schedulers to support intra-step removal of agents #1523
• perf: Refactor iter_cell_list_contents Performance perf: Refactor iter_cell_list_contents #1527
• Replace two loops with dictionary comprehension, list- with generator comprehension Replace two loops with dictionary comprehension, list- with generator comprehension  #1458
• Make MultiGrid.place_agent faster Make MultiGrid.place_agent faster #1508
• Update space module-level docstring summary Update space module-level docstring summary #1518
• Update NetworkGrid.init docstring Update NetworkGrid.__init__ docstring #1514
• Deprecate SingleGrid.position_agent Deprecate SingleGrid.position_agent #1512
• Make swap_pos part of Grid instead of SingleGrid Make swap_pos part of Grid instead of SingleGrid #1507
• Refactor NetworkGrid docstrings and iter/get_cell_list_contents Refactor NetworkGrid docstrings and iter/get_cell_list_contents #1498
• Hexgrid: use get_neighborhood in iter_neighbors Hexgrid: use get_neighborhood in iter_neighbors #1504
• Auto update year for copyright in docs Auto update year for copyright in docs #1503
• Refactor Grid.move_to_empty Refactor Grid.move_to_empty #1482
• Put "Mesa" instead of "it" in README Put "Mesa" instead of "it" in README #1490
• Batchrunner: Remove unnecessary dict transformation, .keys() in len() Batchrunner: Remove unnecessary dict transformation, .keys() in len() #1460
• Add Dependabot configuration for GitHub Actions update check Add Dependabot configuration for GitHub Actions update check #1480
• Use list transformation only when shuffled is True Use list transformation only when shuffled is True #1478
• Implement swap_pos Implement swap_pos #1474
• Clean up DataCollector Clean up DataCollector #1475

Fixes

• Fix accident from a #1488 Fix accident from https://github.com/projectmesa/mesa/pull/1488 #1489
• pre-commit autoupdate [pre-commit.ci] pre-commit autoupdate #1598, [pre-commit.ci] pre-commit autoupdate #1576, [pre-commit.ci] pre-commit autoupdate #1548, [pre-commit.ci] pre-commit autoupdate #1494
• Fix docstring of DataCollector Fix docstring of DataCollector #1592
• Update Pipfile.lock (dependencies) Update Pipfile.lock (dependencies) #1495 Update Pipfile.lock (dependencies) #1487
• build(deps):
  • build(deps): bump codecov/codecov-action from 2 to 3 dependencies Pull requests that update a dependency file build(deps): bump codecov/codecov-action from 2 to 3 #1486
  • build(deps): bump actions/upload-artifact from 2 to 3 dependencies Pull requests that update a dependency file build(deps): bump actions/upload-artifact from 2 to 3 #1485
  • build(deps): bump peter-evans/create-pull-request from 3 to 4 dependencies Pull requests that update a dependency file build(deps): bump peter-evans/create-pull-request from 3 to 4 #1484
  • build(deps): bump actions/setup-python from 3 to 4 dependencies Pull requests that update a dependency file build(deps): bump actions/setup-python from 3 to 4 #1483
• Establish reproducibility for NetworkGrid.get_neighbors when radius > 1 Establish reproducibility for NetworkGrid.get_neighbors when radius > 1 #1569
• Format js code Format js code #1554
• Add some missing const declarations Add some missing const declarations #1549
• fix tutorial url in examples fix tutorial url in examples #1538
• Update cookiecutter to flat import style. Update cookiecutter to flat import style. #1525
• Fix bug in Grid.get_neighborhood Fix bug in Grid.get_neighborhood #1517
• Revert changes of Use list transformation only when shuffled is True #1478 and Remove unnecessary list operations #1456 Revert changes of #1478 and #1456 #1516
• Fix return types of some NetworkGrid methods Fix return types of some NetworkGrid methods #1505
• Update year for copyright Update year for copyright #1501
• Add default_value function to NetworkGrid Add default_value function to NetworkGrid #1497
• Remove extraneous spaces from docstrings in modules 2 Remove extraneous spaces from docstrings in modules 2 #1496
• Remove extraneous spaces from docstrings in modules Remove extraneous spaces from docstrings in modules #1493
• SingleGrid: Remove extraneous attribute declaration (empties) SingleGrid: Remove extraneous attribute declaration (empties) #1491

[jackiekazil]

@tpike3 ^^ Can you make sure any updates you pushed over the weekend are included?

[rht]

examples: Initialize Sugarscape ({G1}, {M, T}) #1290

This one was merged in the examples repo.

[rht]

New features:

• Implement radius for NetworkGrid.get_neighbors #1564

Some highlights for the perf improvements:

• Use getattr for attribute strings in model data collection #1590 this is a 2x speedup over the relevant line, but not sure if compared to the rest of the function
• Refactor iter/get_cell_list_contents methods #1570 at least 1.3x speedup for iter/get_cell_list_contents
• Evaluate empties set more lazily #1546 (comment) ~1.3x speedup for place_agent, remove_agent, and move_agent

Those are harder to find from the PR list because the PR titles don't explicitly have "perf:" prefix.
Overall, seems to be mainly perf, refactors, and bug fixes.

[tpike3]

@rht Thanks! @jackiekazil I have nothing else to add. That is an impressive list.

[jackiekazil]

@rht I believe I made your changes. Can you review the PR?

RE: perf:, etc -- do we want to try to enforce this to for automate history writes in the future?

@tpike3 It was a lot more than I was expecting.

[rht]

Yes, I will make sure that perf commits will have perf: prefix in the future.

[jackiekazil]

https://pypi.org/project/Mesa/
^^ 1.2 is published. Closing this ticket. Please reopen if you notice any issues.