ENH add StreamDecisionForest class for streaming trees

[PSSF23]

Related to #107

Changes proposed in this pull request:

• Add sktree.experimental.StreamDecisionForest for streaming tree tests
• Optimize internal imports

Before submitting

• I've read and followed all steps in the Making a pull request
  section of the CONTRIBUTING docs.
• I've updated or added any relevant docstrings following the syntax described in the
  Writing docstrings section of the CONTRIBUTING docs.
• If this PR fixes a bug, I've added a test that will fail without my fix.
• If this PR adds a new feature, I've added tests that sufficiently cover my new functionality.

After submitting

• All GitHub Actions jobs for my pull request have passed.

[codecov]

Codecov Report

Patch coverage: 89.28% and project coverage change: +0.02% 🎉

Comparison is base (b582895) 87.68% compared to head (5c867df) 87.71%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #116      +/-   ##
==========================================
+ Coverage   87.68%   87.71%   +0.02%     
==========================================
  Files          28       30       +2     
  Lines        2323     2426     +103     
==========================================
+ Hits         2037     2128      +91     
- Misses        286      298      +12     

Files Changed	Coverage Δ
sktree/experimental/sdf.py	69.23% <69.23%> (ø)
sktree/ensemble/_honest_forest.py	91.46% <100.00%> (ø)
sktree/ensemble/_supervised_forest.py	100.00% <100.00%> (ø)
sktree/ensemble/_unsupervised_forest.py	70.63% <100.00%> (ø)
sktree/experimental/__init__.py	100.00% <100.00%> (ø)
sktree/experimental/tests/test_sdf.py	100.00% <100.00%> (ø)
sktree/tree/_honest_tree.py	99.40% <100.00%> (ø)

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

[PSSF23]

WIP, will add some unit tests.

[PSSF23]

Seems good after the fix in fork. Can we merge this? @adam2392

[adam2392]

Codecov states that there are certain LOC not tested. Specifically around the swapping trees logic. Might be useful for code (and your paper) to think of a few unit-tests. A naive one that comes to mind is:

• the performance should improve as you stream iris data in w/ swaps, relative to non-swapping.

Can you also add a changelog entry?

Otw LGTM

[PSSF23]

@adam2392 I'm actually considering removing the tree swapping because it barely has any performance influences. I think I will remove it from this time of paper submission.

[adam2392]

Thanks @PSSF23 !