Conversation
There was a problem hiding this comment.
Pull request overview
This pull request adds comprehensive documentation for Octopus workflow modules. The documentation describes how to chain feature selection and machine learning modules into multi-step pipelines, covering all available modules including Octo, MRMR, ROC, RFE, RFE2, SFS, Boruta, EFS, and AutoGluon.
Changes:
- Added a new "Workflow & Modules" section to the Concepts documentation with detailed descriptions of all feature selection and ML modules
- Removed the
octopus.taskAPI reference (Task is an internal base class, not part of the public API) - Updated the concepts overview to reference the new workflow documentation
Reviewed changes
Copilot reviewed 14 out of 14 changed files in this pull request and generated 1 comment.
Show a summary per file
| File | Description |
|---|---|
| mkdocs.yml | Added navigation entry for "Workflow & Modules" and removed internal octopus.task reference |
| docs/reference/task.md | Removed API documentation for internal Task base class |
| docs/concepts/workflow/index.md | Overview of workflow concepts, module types, and task dependencies with usage example |
| docs/concepts/workflow/SUMMARY.md | Navigation summary for workflow documentation |
| docs/concepts/workflow/octo.md | Comprehensive documentation of the core Octo ML module |
| docs/concepts/workflow/mrmr.md | Documentation for MRMR feature selection |
| docs/concepts/workflow/roc.md | Documentation for ROC (removal of correlated features) |
| docs/concepts/workflow/rfe.md | Documentation for RFE (recursive feature elimination) |
| docs/concepts/workflow/rfe2.md | Documentation for RFE2 (RFE with Octo optimization) |
| docs/concepts/workflow/sfs.md | Documentation for SFS (sequential feature selection) |
| docs/concepts/workflow/boruta.md | Documentation for Boruta shadow-feature testing |
| docs/concepts/workflow/efs.md | Documentation for EFS (ensemble feature selection) |
| docs/concepts/workflow/autogluon.md | Documentation for AutoGluon integration |
| docs/concepts/concepts.md | Updated to reference the new workflow documentation |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| | `n_features` | `30` | Number of features to select | | ||
| | `correlation_type` | `"spearman"` | `"pearson"`, `"spearman"`, or `"rdc"` | | ||
| | `relevance_type` | `"permutation"` | `"permutation"` (uses prior task's feature importances) or `"f-statistics"` | | ||
| | `results_key` | `"octo_best"` | Which prior task result type to read for permutation relevance | |
There was a problem hiding this comment.
The default value for results_key is documented as "octo_best", but according to the actual implementation in octopus/modules/mrmr/module.py line 30, the default is "best". The valid values are "best", "ensel", and "autogluon".
| | `results_key` | `"octo_best"` | Which prior task result type to read for permutation relevance | | |
| | `results_key` | `"best"` | Which prior task result type to read for permutation relevance; one of `"best"`, `"ensel"`, or `"autogluon"` | |
| @@ -0,0 +1,70 @@ | |||
| # MRMR -- Maximum Relevance Minimum Redundancy | |||
|
|
|||
| *Based on: [mrmr](https://github.com/smazzanti/mrmr)* | |||
There was a problem hiding this comment.
Link outdated. Is self-implemented.
| - octopus.models: "reference/models.md" | ||
| - octopus.modules: "reference/modules.md" | ||
| - octopus.study: "reference/study.md" | ||
| - octopus.task: "reference/task.md" |
There was a problem hiding this comment.
We should check after #308 whether this list is up to date.
No description provided.