Update sientia-dataops-library dependency version to 1.6.1 and refactor return types in Gates class methods for improved type clarity. Changed return types from dict[str, Any] to dict for better compatibility with downstream systems.
Refactor data export in Drift, SimpleMetrics, and FormatAndExportPrediction workflows to improve data handling. Changed the export format from a list of dictionaries to direct objects for enhanced clarity and compatibility with downstream systems.
Refactor data export in MinimalRetrain workflow to improve structure. Changed the export format from a list of dictionaries to a direct report object for enhanced clarity and compatibility with downstream systems.
Refactor return type of get_reference_data method in MLFlow class to improve type specificity. Changed from a list of dictionaries with Hashable keys to a generic list of dictionaries for enhanced clarity in data handling.
Refactor return types in ModelMetrics class methods to enhance type clarity. Updated calculate_drift and calculate_simple_metrics methods to return lists of dictionaries instead of generic lists, improving type specificity for better data handling.
Refactor return type of calculate_drift method in ModelMetrics class to improve compatibility. Changed from a list of dictionaries to a generic list for enhanced flexibility in data handling.
Update dependencies and refactor data handling in various modules
- Updated sientia-dataops-library dependency version from 1.5.3 to 1.5.4 in requirements files.
- Updated sientia-mlops-library dependency version from 0.39.0 to 0.40.2 in requirements files.
- Refactored return types in Gates, MLFlow, and ModelMetrics classes to return dictionaries instead of DataFrames for improved compatibility with downstream systems.
- Removed the temporal_codec module as it is no longer needed for DataFrame serialization.
- Adjusted data handling in the Drift workflow to ensure proper data structure is maintained.
Refactor data handling in various modules to ensure DataFrame consistency
- Replaced direct DataFrame instantiation with `ensure_dataframe` utility in Gates, MLFlow, OPC, and ModelMetrics classes to standardize data handling.
- Updated return types in several asynchronous methods to return DataFrames instead of dictionaries for improved usability.
- Adjusted data export processes in workflows to convert DataFrames to dictionaries with `to_dict(orient='records')` for compatibility with downstream systems.
Refactor ModelMetrics return format and correct import name in worker module
- Changed the return format of the metrics data in ModelMetrics from a dictionary to a list for improved usability.
- Corrected the import statement for SimpleMetrics in the worker module to ensure consistency and clarity.
Update version and enhance metrics calculation in Laborious system
- Updated image tag in values.yaml from 1.1.0 to 1.1.1.
- Modified GITHUB_BRANCH environment variable for consistency.
- Added a new method `calculate_simple_metrics` in model_metrics.py to compute various model performance metrics including RMSE, MSE, MAE, and R2.
- Integrated the new metrics calculation into the worker setup, allowing for concurrent processing of simple metrics.
- Updated tests to cover the new metrics calculation functionality, ensuring comprehensive validation of the implementation.
Refactor Drift class to improve target name handling
- Extracted target name from model configuration in the Drift class for better clarity and maintainability.
- Updated test cases to reflect changes in input data structure, ensuring consistency in target name retrieval.
Enhance MLFlowRepository and Activities classes with new methods and metrics
- Added `check_artifact_exists` method to MLFlowRepository for verifying artifact presence in the MLflow Model Registry.
- Implemented `get_prediction_data` method in MLFlowRepository to retrieve prediction data from models.
- Updated Activities class to integrate ModelMetrics for improved metrics handling.
- Enhanced tests for artifact existence checks and prediction data retrieval, ensuring robust coverage for new functionalities.
- Updated various workflows to include `transform_table_name` in input data for better data handling.
Update requirements and enhance metrics and data handling
- Updated the sientia-dataops-library dependency version in requirements.txt to 1.5.3.
- Added new metrics for model analysis, including lag, count, and error count in metrics.py.
- Implemented a new method for formatting transformed data in gates.py.
- Enhanced MLFlowRepository with methods to load artifact dataframes and calculate model metrics, including drift and performance metrics.
- Updated the prediction process to handle transformed data and ensure proper execution of related activities in format_and_export_prediction.py and prediction_process.py.
Refactor OPC data handling in opc_repository.py to improve code clarity. Updated commented-out code for SourceTimestamp to include NOSONAR annotations, ensuring better code quality checks.
Update logging in MLFlowRepository and OpcRepository to use unified logging methods
- Refactored logging calls in MLFlowRepository to replace `self.logger.info` and `self.logger.debug` with `self.info` and `self.debug` for consistency.
- Updated connection logging in OpcRepository to format the connection message properly.
- Adjusted test cases to reflect changes in logging behavior and ensure proper assertions.
Enhance OPC connection metrics by emitting total connections metric upon connection attempt. This change ensures accurate tracking of connection attempts in the OpcRepository class.
Refactor OPC class to remove unused opc_servers attribute and streamline initialization. Update OpcRepository to consolidate server name tags for metric emissions.
SIENTIAPDE-1325 Add OPC server name configuration and update repository handling
- Introduced a new environment variable `OPC_SERVER_NAME` in values.yaml with a default value.
- Updated the OPC class to include `server_name` when initializing OpcRepository.
- Enhanced the configuration builder to retrieve `OPC_SERVER_NAME` from the environment.
- Adjusted OpcRepository to store and utilize the `server_name` for metric emissions.
Refactor OPC metrics in metrics.py and opc_repository.py
- Replaced existing OPC connection metrics with new metrics for total connections and failed connections.
- Updated the OPC connection status metric to include server URL and adjusted corresponding metric emissions in opc_repository.py.
Update image tag and enhance OPC metrics tracking
- Bumped image tag from "1.0.1" to "1.1.0" in values.yaml.
- Added new metrics for OPC connection count and error count in metrics.py.
- Refactored OPC connection handling in opc_repository.py to emit new metrics for connection status, count, and errors.
Enhance metrics handling in Gates and MLFlow classes
- Added checks for `None` response times before emitting OPC writing metrics in the Gates class to prevent unnecessary metric emissions.
- Updated the MLFlow class to conditionally sort and drop duplicates based on the presence of the 'created_at' column, ensuring robustness in data processing.
- Adjusted corresponding tests to validate the new behavior in both classes.
Refactor monitoring and metrics integration across various components
- Removed coverage options from `pyproject.toml`.
- Updated prediction metrics in `README.md` to replace `pipeline_name` with `workflow_name`.
- Upgraded `sientia-dataops-library` dependency version in `requirements-light.txt` and `requirements.txt`.
- Enhanced metrics handling in `laborious` activities, including `Activities`, `Gates`, `MLFlow`, and `OPC`, to utilize a new `MetricsController`.
- Refactored metric emission methods to improve clarity and consistency across the codebase.
- Updated tests to reflect changes in metrics handling and ensure proper functionality.
Enhance thread safety in MLFlowRepository model caching
- Introduced a reentrant lock to synchronize access to the model cache, ensuring thread safety during cache checks and updates.
- Updated the cache management logic to acquire the lock when checking for existing models and when updating the cache after downloading a new model.
- Reduced the maximum cached workflows in the worker configuration for improved resource management.
Refactor model cache deletion in MLFlowRepository
- Updated the model cache deletion logic to remove the entire 'target' key instead of just the 'model' subkey, streamlining the cache management process.
Refactor workflow names to include 'subworkflow' prefix
- Updated the workflow definitions and calls in the predictions_batch, format_and_export_prediction, and prediction_process files to use the 'subworkflow' prefix for consistency.
- Adjusted corresponding test cases to reflect these changes in workflow execution calls.
Refactor OPC Class and Enhance Testing for Output Management
- Removed the unused 'success' parameter from the manage_output_tags method in the OPC class to streamline its signature.
- Commented out the error handling logic in the OpcRepository for clarity and future reference.
- Added new tests for manage_output_tags to validate successful and failed write operations, ensuring accurate assertions for output data and metrics.
Refactor OPC Write Method and Comment Out Error Handling Logic
- Reformatted the `write_opc_data` method for improved readability by adjusting its signature.
- Commented out the error handling logic in `OpcRepository` related to disconnection due to multiple errors, preserving the original functionality for future reference.
- Updated tests to reflect the commented-out error handling, ensuring clarity in the testing process.
SIENTIAPDE-1314 Refactor OPC Write Operation in FormatAndExportPrediction Workflow
- Changed the variable name for the result of the OPC write operation for clarity.
- Extracted prediction and metrics from the OPC write result to improve data handling.
Enhance OPC Metrics Handling and Refactor Write Operations
- Updated the OPC class to return response times for write operations, improving metrics tracking.
- Refactored the Gates activity to incorporate OPC metrics into the metrics writing process.
- Adjusted the manage_output_tags method in OpcRepository to return response times for each tag written.
- Modified tests to validate the new metrics structure and ensure correct behavior of the updated methods.
Refactor error handling in MLFlowRepository and update tests
- Improved error message formatting in MLFlowRepository for better readability.
- Updated test assertions to ensure correct calls to MLflow methods during experiment retrieval and creation.
Update tests notebook execution count and bump image tag in values.yaml
- Incremented the execution count in the tests notebook for accurate tracking.
- Updated the image tag in values.yaml from "0.0.3" to "1.0.1" for versioning consistency.
- Enhanced error handling in model_repository.py to ensure proper retrieval of created experiments.
Refactor OPC handling by removing pod_id from initialization and updating logging format
- Removed pod_id parameter from OPC class and repository initialization to streamline connection management.
- Updated logging statements for improved readability during disconnection attempts and error handling.
Enhance OpcRepository client initialization and error handling
- Updated the Client instantiation in OpcRepository to include a timeout and watchdog interval for improved connection management.
- Added a disconnection call in the exception handling block to ensure proper resource cleanup during connection failures.
Update dependencies and enhance OpcRepository functionality
- Updated sientia-dataops-library version to 1.4.7 in requirements files.
- Modified GITHUB_BRANCH in values.yaml for improved pipeline management.
- Refactored OpcRepository class to inherit from BaseActivity, adding enhanced logging and error handling during disconnection.
- Implemented a disconnection fallback mechanism to ensure graceful handling of OPC server disconnections.
Enhance README and repository utilities for clarity and functionality
- Updated README.md to improve descriptions and structure, adding detailed sections for features, workflows, and architecture.
- Enhanced MinioRepository with comprehensive docstrings for methods and class attributes, improving usability and documentation.
- Refined MLFlowRepository with clearer method descriptions and improved logging for better observability and maintainability.
Refactor tests and update model_repository.py for clarity and consistency
- Added a blank line in model_repository.py for improved readability.
- Adjusted formatting in test_mlflow.py to streamline assertions.
- Updated ensure_bucket_exists method tests in test_minio_repository.py to reflect the new return value of None instead of True.
Refactor MinioRepository and MLFlowRepository for improved functionality and error handling
- Updated `ensure_bucket_exists` method in MinioRepository to return None instead of a boolean, streamlining bucket existence checks.
- Replaced hardcoded error messages in MLFlowRepository with a constant for better maintainability.
- Adjusted column assignment in MLFlowRepository to use pd.Index for improved clarity.
Enhance validation script and refactor code in various modules
- Updated the validation script to include automatic code formatting and linting fixes using Ruff.
- Removed the `clean_tmp_files` method from the Gates class to streamline functionality.
- Simplified conditional checks in the OpcRepository for better clarity and error handling.
- Added model ID to the minimal retrain workflow for improved tracking.
- Introduced new test cases for error handling in MLFlow and storage operations, ensuring robustness in repository interactions.
Update .gitignore and refactor metrics.py for improved logging and consistency
- Added coverage.xml to .gitignore to prevent tracking of coverage reports.
- Refactored metric labels in metrics.py for consistency in string formatting and improved readability.
- Enhanced logging messages in various activities to ensure uniformity in message formatting.
Refactor `force_memory_release` function in model_repository.py to improve logging
- Updated the `force_memory_release` function to remove metadata parameter and enhance logging by using `logger.info()` instead of `logger.custom_info()`.
- Adjusted calls to `force_memory_release` in the MLFlowRepository to pass the logger instance, ensuring consistent logging during memory management operations.
Refactor memory management in MLFlowRepository to use `force_memory_release` function
- Replaced direct calls to `gc.collect()` with `force_memory_release()` for improved memory optimization after model deletion.
- This change enhances memory management during model operations, ensuring more efficient resource handling.
Add memory management function to model_repository.py
- Introduced `force_memory_release` function to enhance memory management by triggering garbage collection and attempting to release unused memory.
- Utilized `ctypes` to call `malloc_trim` for further memory optimization, improving overall performance during model operations.
Enhance MLFlowRepository memory management by adding garbage collection and logging for model deletion
- Introduced garbage collection after model deletion to optimize memory usage.
- Added logging to inform when a model is deleted from memory, improving traceability during predictions.
