Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Model explainability interface for H2O models and AutoML objects in R/Python #7846

Closed
exalate-issue-sync bot opened this issue May 11, 2023 · 1 comment
Assignees

Comments

@exalate-issue-sync
Copy link

In R, we have h2o.explain() and h2o.explain_row() and Python we have the .explain() and .explain_row() methods, however there are a number of utility plotting functions that are also added (and
used within these functions).

The explain functions return classes of the following types: {{h2o.explain._explain.H2OExplanation}} in Python and {{H2OExplanation}} in R.

There are a number of new functions that come along with the main {{.explain()}} interface:

h3. Python

  • New methods added to H2OAutoML and at the same time functions accepting list of models in {{h2o}} module, e.g., {{aml.partial_dependences(test_frame, column)}}/ {{h2o.partial_dependences(list_of_models, test_frame, column)}}
    ** {{pd_multi_plot}}
    ** {{varimp_heatmap}}
    ** {{model_correlation_heatmap}}
    ** {{explain}}
    ** {{explain_row}}
  • Methods usable only by tree-based models (i.e., models that implement {{predict_contributions}})
    ** {{shap_explain_row_plot}}
    ** {{shap_summary_plot}}
  • Methods implemented for every model:
    ** {{ice_plot}}
    ** {{pd_plot}}
    ** {{explain}}
    ** {{explain_row}}
  • Method implemented only for regression:
    ** {{residual_analysis_plot}}

h3. R

  • Functions applicable on H2OAutoML or list of models:
    ** {{h2o.explain}}
    ** {{h2o.explain_row}}
    ** {{h2o.model_correlation_heatmap}}
    ** {{h2o.pd_multi_plot}}
    ** {{h2o.varimp_heatmap}}
  • Functions applicable to single model:
    ** {{h2o.explain}}
    ** {{h2o.explain_row}}
    ** {{h2o.ice_plot}}
    ** {{h2o.pd_plot}} - the only reason for not using {{h2o.partialPlot}} here is that it is using base plotting which would make it impossible to use as the rest of the plots here - return an object that can later be plotted (also it could not be rendered by {{plotly}} to make it interactive)
    ** {{h2o.residual_analysis_plot}} - only for regression problems
    ** only for tree-based models (i.e., models that implement {{predict_contributions}})
    *** {{h2o.shap_explain_row_plot}}
    *** {{h2o.shap_summary_plot}}
@h2o-ops
Copy link
Collaborator

h2o-ops commented May 14, 2023

JIRA Issue Migration Info

Jira Issue: PUBDEV-7796
Assignee: Tomas Fryda
Reporter: Erin LeDell
State: Resolved
Fix Version: 3.32.0.1
Attachments: N/A
Development PRs: Available

Linked PRs from JIRA

#4932

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

2 participants