Jmafoster1/remove data collector

[jmafoster1]

• DataCollector classes have been removed
• ObservationalDataCollector has been replaced by directly passing dataframes
• ExperimentalDataCollector has been replaced by ExperimentalEstimator class to directly run models
• Removed test generation and execution functionality from MetamorphicRelation since this can be done via the JSON front end
• Moved MetamorphicRelation classes into the testing package instead of specification
• Changed estimators (except IPCWEstimator) to take a BaseTestCase instead of a treatment and outcome variable

[github-actions]

🦙 MegaLinter status: ✅ SUCCESS

Descriptor	Linter	Files	Fixed	Errors	Elapsed time
✅ PYTHON	black	29		0	0.87s
✅ PYTHON	pylint	29		0	5.36s

See detailed report in MegaLinter reports

MegaLinter is graciously provided by

[codecov]

Codecov Report

Attention: Patch coverage is 98.93048% with 2 lines in your changes missing coverage. Please review.

Project coverage is 94.64%. Comparing base (48e636d) to head (15fa8ee).
Report is 49 commits behind head on main.

Files with missing lines	Patch %	Lines
causal_testing/testing/causal_test_result.py	50.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #308      +/-   ##
==========================================
- Coverage   96.87%   94.64%   -2.23%     
==========================================
  Files          29       25       -4     
  Lines        1856     1345     -511     
==========================================
- Hits         1798     1273     -525     
- Misses         58       72      +14     

Files with missing lines	Coverage Δ
causal_testing/__init__.py	100.00% <ø> (ø)
causal_testing/estimation/abstract_estimator.py	100.00% <100.00%> (ø)
...esting/estimation/abstract_regression_estimator.py	96.00% <100.00%> (-1.96%)	⬇️
...ausal_testing/estimation/cubic_spline_estimator.py	96.55% <100.00%> (+0.12%)	⬆️
...ausal_testing/estimation/experimental_estimator.py	100.00% <100.00%> (ø)
...ting/estimation/instrumental_variable_estimator.py	100.00% <100.00%> (ø)
causal_testing/estimation/ipcw_estimator.py	99.25% <100.00%> (+<0.01%)	⬆️
..._testing/estimation/linear_regression_estimator.py	100.00% <100.00%> (ø)
...esting/estimation/logistic_regression_estimator.py	100.00% <100.00%> (ø)
causal_testing/specification/causal_dag.py	99.49% <100.00%> (+0.53%)	⬆️
... and 8 more

... and 1 file with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 1240348...15fa8ee. Read the comment docs.

[jmafoster1]

This PR looks to drop the overall percentage code coverage, but that's because I've ripped out a bunch of code, so the proportions have changed slightly.

[jmafoster1]

Resolved merging conflicts from main

[f-allian]

@jmafoster1 Thanks Michael, this all looks good. However, we should also update our documentation to reflect these changes too. If you don't have the time let me know and I'll take it on 👍🏼

[jmafoster1]

Good point. Done

[jmafoster1]

I think this is ready for review now. Absolute monster of a PR!

[f-allian]

@jmafoster1 This looks fine now. There are still some mentions of the data collector / JSON class in our examples/ directory, but it makes sense to leave that for another PR once we've had our discussions. 👍🏼

[jmafoster1]

Thanks Farhad. I had a quick look through and removed all mentions of data collection and JSON from docs and examples by running grep for "collect" and "json" and removing all the problematic content. I agree that we can get this merged in now for the JOSS paper and give the docs a proper overhaul in the coming months, but thought it was important to lose anything that could be confusing or misleading!

[jmafoster1]

No need to re-review. Once this batch of tests pass, I'll get this PR merged in and do the JOSS release