Fix model wrapper to identify model type, remove task type from error analysis

[tongyu-microsoft]

This PR fixes an issue that blocks curated image release and other RAI repo releases.

Description

The issue is, the model wrapper created in PR #1824 has both predict function and predict_proba function. In error analysis, we check if the model has predict_proba function to identify if it is a classification model or regression model. In the case model is wrapped, it will always identify the model as classification model.

The fix is, we check if the model is classification model or regression model, and wrap the model in a separate function. So regression model won't have predict_proba function after wrapping. In this case, we don't need to specifically pass task_type into error_analysis_manager. So this PR also removes task_type from error_analysis_manager

Checklist

• I have added screenshots above for all UI changes.
• I have added e2e tests for all UI changes.
• Documentation was updated if it was needed.

[codecov-commenter]

Codecov Report

Merging #1912 (5d61dec) into main (0441923) will decrease coverage by 7.48%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##             main    #1912      +/-   ##
==========================================
- Coverage   93.32%   85.85%   -7.48%     
==========================================
  Files          93       29      -64     
  Lines        4586      523    -4063     
==========================================
- Hits         4280      449    -3831     
+ Misses        306       74     -232     

Flag	Coverage Δ
unittests	85.85% <ø> (-7.48%)	⬇️

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

Impacted Files	Coverage Δ
...i/responsibleai/managers/error_analysis_manager.py
...sibleai/responsibleai/rai_insights/rai_insights.py
...tools/error_analysis/dashboard_schemas/__init__.py
erroranalysis/erroranalysis/report/__init__.py
...ponsibleai/responsibleai/_tools/shared/versions.py
...bleai/responsibleai/_tools/causal/causal_config.py
responsibleai/responsibleai/exceptions.py
erroranalysis/erroranalysis/_internal/metrics.py
...eai/responsibleai/databalanceanalysis/constants.py
...i/responsibleai/managers/counterfactual_manager.py
... and 54 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[gaugup]

Are there any tests which validate these changes?

[github-actions]

https://responsibleai.blob.core.windows.net/pullrequest/microsoft/responsible-ai-toolbox/tongy/fixPredictProba/dashboard/index.html

[tongyu-microsoft]

Are there any tests which validate these changes?

I will add more tests in my next PR, as I test locally and I would like to merge this PR asap. Thanks!