Update SDMetrics

[myyong]

Current dependencies are here.

I am creating this issue because SDMetrics has been updated significantly from 0.6 to 0.8.

The plan is:

• [x] Create new branch my/update-sdv
• [x] Delete poetry.lock,
• [x] Update pyproject.toml sdmetrics = "^0.6.0" to sdmetrics = "0.8.0"
• [x] Run poetry update

I see that the poetry.lock file has been updated and sdmetrics is now at version 0.8.

• [x] Run pytests.

Tests run fine. Simple example works.

[myyong]

I'm listing the metrics that seem useful:

• [x] Boundary Adherence (Commit here)
• [x] Range Coverage (f57079f
• [ ] Category Coverage
• [ ] Contingency Similarity
• [ ] KSComplement
• [ ] Statistic Similarity

Stuff that we do not take

• [ ] Cardinality Shape Similarity
• [ ] Categorical Cap
• [ ] Correlation Similarity() @sgreenbury Yay or Nay?
• [ ] Missing Value Similarity
• [ ] New Row Synthesis
• [ ] TV Complement

[myyong]

This is the beta list. Pick a metrics that you like, which you think shows the 'future directions for monitoring' aspect.

[sgreenbury]

• Static data: As we don't have any efficacy metrics currently and binary classification is a common task, we can aim for this ML efficacy - binary classification to start
• Time series: potentially include rough paths (esig) with application to MIMIC as a use case (some public data for testing/examples in package). Alternatively, can look to sequential detection.

[sgreenbury]

Remaining tasks:

• [x] Consider renaming hypothesis tests module to "scores"
  • Renaming on branch refactor-names tbc
• [x] Docs seem to have the "learning_machines_drift" and "learning_machines_drift package" duplicating items
• [x] Remove current logistic_detection and rename logistic_detection_custom as logistic_detection with default parameters providing the same result as current logistic_detection
• [x] Add short but complete usage example to README and refine from user not dev perspective (SG)
• [x] Check learning machines drift cases notebooks work as expected after recent changes (i.e. #39)
  • [x] Default to "predicted-labels" as series name only when labels is saved with no name
  • [x] Check case where features are not present in both registered and reference
• [x] Decide whether to include the efficacy and time series measures
  • Removed
• [x] Refine simple example (MY)
• [x] Check deps are all required (SG)
  • Updated with new dev and docs groups that are optional. Merged into my/update-sdv.
• [ ] Check the clear reference and logged paths are functioning as expected by running examples/simple_example/main.py with the clear flags set as True. (JS)
• [x] Move background into separate markdown (SG)
• [x] Check handling of depth greater than 1 path for backend

Docs tasks

• [ ] Make docstrings consistent across repo (e.g. tag should be defined in the same way across methods)
• [ ] Check formatting of variables and hyperlinks are correct in the make html build
• [ ] Check no warnings remain

[sgreenbury]

Closing with #39.