Prevent overwriting of dataset in erroranalysis

[gaugup]

Description

Fix overwriting of dataset in erroranalysis. This overwriting only happens in case there are only categorical features in the dataset. It doesn't trigger in case there only numerical features or mix of numerical and categorical features.

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 #2050 (b967eb6) into main (7e8fdaf) will not change coverage.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##             main    #2050   +/-   ##
=======================================
  Coverage   92.33%   92.33%           
=======================================
  Files          98       98           
  Lines        4957     4957           
=======================================
  Hits         4577     4577           
  Misses        380      380           

Flag	Coverage Δ
unittests	92.33% <100.00%> (ø)

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

Impacted Files	Coverage Δ
...is/erroranalysis/_internal/surrogate_error_tree.py	89.53% <100.00%> (ø)
...ranalysis/erroranalysis/analyzer/error_analyzer.py	95.62% <100.00%> (ø)

[github-actions]

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

[github-actions]

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

[github-actions]

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