💚:bug::snake: Fix bug for 0 passing predictions

[evanroyrees]

💚⬆️ Pin scikit-learn to >=1.3
🐍🎨 Replace conda commands with mamba in Makefile
🔥⬇️ Remove hdbscan (now available in scikit-learn >= 1.3)

HDBSCAN has been failing in pytests and this seems to be an artifact of hdbscan being integrated into scikit-learn.

Some of these issues have cropped up in the past few weeks (scikit-learn-contrib/hdbscan#600).

scikit-learn is now pinned to >=1.3 as this is the earliest release of HDBSCAN being available via sklearn.cluster.HDBSCAN

The run_hdscan function has now been updated according to the API that was adopted in the sklearn library.

PR checklist

• This comment contains a description of changes (with reason).
• If you've fixed a bug or added code that should be tested, add tests!
• Have you followed the pipeline conventions in the contribution docs

[codecov]

Codecov Report

Patch coverage: 30.66% and project coverage change: -0.29% ⚠️

Comparison is base (6637426) 27.59% compared to head (bc34fae) 27.31%.
Report is 46 commits behind head on dev.

Additional details and impacted files

@@            Coverage Diff             @@
##              dev     #341      +/-   ##
==========================================
- Coverage   27.59%   27.31%   -0.29%     
==========================================
  Files          45       50       +5     
  Lines        5349     5730     +381     
==========================================
+ Hits         1476     1565      +89     
- Misses       3873     4165     +292     

Flag	Coverage Δ
unittests	27.31% <30.66%> (-0.29%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files Changed	Coverage Δ
autometa/binning/large_data_mode.py	0.00% <0.00%> (ø)
autometa/config/databases.py	0.00% <0.00%> (ø)
autometa/config/utilities.py	25.00% <ø> (ø)
autometa/validation/benchmark.py	0.00% <0.00%> (ø)
autometa/validation/cami.py	0.00% <0.00%> (ø)
autometa/taxonomy/majority_vote.py	13.53% <11.53%> (+0.83%)	⬆️
autometa/binning/summary.py	41.48% <18.18%> (-0.38%)	⬇️
autometa/taxonomy/lca.py	9.37% <19.04%> (+0.57%)	⬆️
autometa/taxonomy/gtdb.py	21.01% <21.01%> (ø)
autometa/common/utilities.py	23.35% <37.50%> (+0.65%)	⬆️
... and 11 more

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[chasemc]

It seems like the test coverage for the code is good so assuming the tests are good then the changes are fine and seem to fix the builds. Should get another pair of eyes to look but seems good

[chasemc]

Code is fine. Going to need to search and replace mamba/conda in the Docs though

[chasemc]

May need to look at hdbscan in tests
e.g.

Autometa/tests/environment.yml

Line 14 in 8ebd7a7

- hdbscan

[Sidduppal]

Looks good, minor comments:

• What happens if the user has conda but not mamba? I created an environment without mamba and the installation was using (PYTHON_INTERPRETER) -m pip install -q virtualenv virtualenvwrapper. So maybe specify in the docs or somewhere to install mamba if the user needs the advantage conda offers.
• On the environment with mamba the installation worked well 👍🏼
• As @chasemc mentioned, update the docs telling user to use mamba and change scikit-learn in tests/environment.yml
• I didn't try running clustering using hdbscan but I assume it should work as the tests are passing.