Use latest views-pipeline-core package#44
Conversation
xiaolong0728
commented
Dec 11, 2025
xiaolong0728
commented
- Support classification models using the latest Darts library and update classes accordingly.
- Update classes to adapt to the latest views-pipeline-core
There was a problem hiding this comment.
Pull request overview
This PR updates the views-stepshifter package to use the latest views-pipeline-core package and Darts library (v0.38.0). Key changes include migrating to native Darts classification models, applying log transformations to targets, and updating the configuration attribute name from config to configs throughout the codebase.
- Migrates custom model implementations to use native Darts models (XGBClassifierModel, LightGBMClassifierModel)
- Adds log/exp transformations for target values in predictions
- Updates configuration attribute naming convention across the manager classes
Reviewed changes
Copilot reviewed 8 out of 8 changed files in this pull request and generated 4 comments.
Show a summary per file
| File | Description |
|---|---|
| views_stepshifter/models/stepshifter.py | Removes unused import, updates deprecated method call, adds log transformations, comments out CUDA support |
| views_stepshifter/models/hurdle_model.py | Switches from custom to native Darts classification models, comments out CUDA configuration |
| views_stepshifter/models/darts_model.py | Replaces extensive custom model implementations with simple wrapper classes inheriting from Darts models |
| views_stepshifter/models/init.py | Removes exports for models now provided by Darts library |
| views_stepshifter/manager/stepshifter_manager.py | Renames config attribute to configs throughout the manager class |
| tests/test_stepshifter_manager.py | Updates test fixtures and assertions to match new configuration attribute naming |
| tests/test_hurdle_model.py | Updates test configuration to use supported model names |
| pyproject.toml | Bumps version to 1.1.0, updates Darts dependency to ^0.38.0, adds xgboost ^3.0.0 |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| else: | ||
| self.config["model_reg"] = self.config["algorithm"] | ||
| model = StepshifterModel(self.config, partitioner_dict) | ||
| self.configs = {"model_reg": self.configs["algorithm"]} |
There was a problem hiding this comment.
This assignment overwrites the entire configs dictionary with a single key, losing all other configuration values. Should merge this key into the existing dictionary instead, e.g., self.configs['model_reg'] = self.configs['algorithm'] or create a new dict: self.configs = {**self.configs, 'model_reg': self.configs['algorithm']}.
| self.configs = {"model_reg": self.configs["algorithm"]} | |
| self.configs["model_reg"] = self.configs["algorithm"] |
| path_artifact = self._model_path.get_latest_model_artifact_path(run_type) | ||
|
|
||
| self.config["timestamp"] = path_artifact.stem[-15:] | ||
| self.configs = {"timestamp": path_artifact.stem[-15:]} |
There was a problem hiding this comment.
This assignment overwrites the entire configs dictionary, discarding all existing configuration. Should merge instead: self.configs['timestamp'] = path_artifact.stem[-15:] or self.configs = {**self.configs, 'timestamp': path_artifact.stem[-15:]}.
|
|
||
| self.config["timestamp"] = path_artifact.stem[-15:] | ||
|
|
||
| self.configs = {"timestamp": path_artifact.stem[-15:]} |
There was a problem hiding this comment.
This assignment overwrites the entire configs dictionary, discarding all existing configuration. Should merge instead: self.configs['timestamp'] = path_artifact.stem[-15:] or self.configs = {**self.configs, 'timestamp': path_artifact.stem[-15:]}.
| self.configs = {"timestamp": path_artifact.stem[-15:]} | |
| self.configs['timestamp'] = path_artifact.stem[-15:] |
| mock_path.model_dir = "/path/to/models/test_model" | ||
| mock_path.target = "model" | ||
| mock_path.artifacts = Path("/path/to/artifacts") | ||
| mock_path.get_latest_model_artifact_path.return_value = Path("test_model_202401011_200000") |
There was a problem hiding this comment.
Corrected spelling in path from 'test_model_202401011_200000' to 'test_model_20240101_200000' - appears to have extra '1' digit.
| mock_path.get_latest_model_artifact_path.return_value = Path("test_model_202401011_200000") | |
| mock_path.get_latest_model_artifact_path.return_value = Path("test_model_20240101_200000") |
1 participant
