[ENH, GRAPH] Experimental: A module for time-series graphs that relies on only networkx, and simulation and algorithms for time-series

[adam2392]

Signed-off-by: Adam Li [email protected]

Changes proposed in this pull request:

• implements a basic class for time-series graphs
• implements a simulation module for basic Gaussian weight time-series graphs

This is a WIP, that relies on dynetx for the API for time-series graphs. However, we need to create a corresponding TimeSeriesMixedEdgeGraph that is similar to MixedEdgeGraph and also has the API of dynetx ts-graphs. This is mostly experimental and a draft to see what the API would look like, and would be used in conjunction with downstream packages to test ts-causal discovery functionality.

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.

[adam2392]

@amit-sharma @bloebp just tagging you guys from today's discussion. The PR itself is still a mess/WIP, so I would restrict to just loosely looking at the time-series implementations to see if you spot any fundamental issues.

• pywhy_graphs/classes/timeseries/base.py
• pywhy_graphs/classes/timeseries/timeseries.py
• pywhy_graphs/classes/timeseries/cpdag.py

and the corresponding unit tests to see how the time-series graphs should work in terms of an API. I'm flushing out minor details in the graph API, as I implement the dodiscover time-series algos.

cc: @robertness as relevant for dodiscover

@emrekiciman

[codecov-commenter]

Codecov Report

Merging #21 (072325a) into main (5924622) will decrease coverage by 9.62%.
The diff coverage is 48.96%.

@@            Coverage Diff             @@
##             main      #21      +/-   ##
==========================================
- Coverage   71.95%   62.33%   -9.63%     
==========================================
  Files          17       28      +11     
  Lines        1159     2079     +920     
  Branches      281      483     +202     
==========================================
+ Hits          834     1296     +462     
- Misses        259      686     +427     
- Partials       66       97      +31     

Impacted Files	Coverage Δ
pywhy_graphs/classes/cpdag.py	73.46% <0.00%> (ø)
pywhy_graphs/classes/intervention.py	0.00% <0.00%> (ø)
pywhy_graphs/classes/networkxprotocol.py	0.00% <0.00%> (ø)
pywhy_graphs/simulate.py	0.00% <0.00%> (ø)
pywhy_graphs/viz/layout.py	10.71% <10.71%> (ø)
pywhy_graphs/algorithms/pag.py	74.15% <18.91%> (-9.10%)	⬇️
pywhy_graphs/config.py	38.23% <22.07%> (-49.77%)	⬇️
pywhy_graphs/algorithms/cyclic.py	77.77% <22.22%> (-14.12%)	⬇️
pywhy_graphs/viz/draw.py	36.84% <26.82%> (+1.95%)	⬆️
pywhy_graphs/classes/timeseries/mixededge.py	38.59% <38.59%> (ø)
... and 12 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[adam2392]

For documentation

Nice rendered doc demonstrating preliminary time-series graph usage: https://output.circle-artifacts.com/output/job/7dc87c94-9c95-41b6-97a8-14091b9e9c8c/artifacts/0/dev/auto_examples/visualization/plot_timeseries_graphs.html#sphx-glr-auto-examples-visualization-plot-timeseries-graphs-py