Changelog¶
All notable changes to chronos-lab will be documented in this file.
The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
Unreleased¶
0.2.1 - 2026-02-11¶
Added¶
- Interactive Brokers Integration: Complete market data support with
IBMarketDataAPI for real-time ticks and historical bars, plus high-levelohlcv_from_ibandohlcv_from_ib_asyncfunctions for seamless OHLCV data retrieval TimeSeriesCollectionclass for multi-symbol data orchestration (experimental, undocumented, subject to change)
Documentation¶
- New Interactive Brokers market data tutorial and API reference
0.2.0 – 2026-02-02¶
Added¶
- Configurable OHLCV data sources in DAGs:
Analysis calculations can now pull data directly from multiple sources (yfinance,intrinio, orarcticdb) using a unified configuration interface, making it easy to switch between a variable input and external providers. - Source-aware anomaly detection:
AnalysisDriver.detect_anomaliesnow supports both preloaded OHLCV DataFrames and automatic retrieval from configured data sources. - Dynamic dataset and ArcticDB outputs:
Analysis results can be flexibly routed to different datasets/backends, enabling easier experimentation and storage control.
Changed¶
AnalysisDriverimprovements:
Streamlined configuration, clearer parameters, and a more modular design for building composable analysis workflows.- OHLCV standardization:
standardize_ohlcvupdated to work consistently across dynamically selected data sources. - Validation and robustness:
Added parameter validation and safer defaults for more predictable behavior. - Documentation refresh:
Expanded guides and examples reflecting the new source configuration andAnalysisDriver-based workflows.
Removed¶
- Legacy
MCPdependencies and related modules. - The deprecated
analysis.calculationmodule and outdated APIs. - Plot module is no longer a part of the official documented API.
Deprecated¶
- Remaining legacy anomaly detection functions and parameters. Use
AnalysisDrivergoing forward.
0.1.8 – 2026-01-29¶
Added¶
- Configurable ArcticDB backends:
ohlcv_from_arcticdb()andohlcv_to_arcticdb()now support explicit selection of the ArcticDB backend (LMDB,S3, orMEM), allowing the same code to run against local, in-memory, or cloud-backed stores. AnalysisDriverAPI:
A new interface intended to serve as the foundation for multiple analysis calculations, with caching and Hamilton-based DAG execution.- Updated documentation and examples covering backend configuration and
AnalysisDriverAPI.
Changed¶
- Anomaly detection interface:
detect_ohlcv_anomaliesis deprecated in favor ofAnalysisDriver-based analysis, aligning anomaly detection with the new unified analysis API. - Plot rendering behavior:
Plotting logic was refined to ensure figures render correctly in notebooks and scripts without accumulating open figures. - Logging defaults:
The global log level now defaults toWARNINGfor a quieter out-of-the-box experience.
Deprecated¶
- The
analysis.calculationmodule and related anomaly detection helpers. Migration guidance is available in the documentation.
0.1.7 - 2026-01-27¶
Added¶
- New Tutorial: Getting Started with Chronos Lab
- Anomaly executor configuration
Changed¶
- Dependency updates
- Documentation
0.1.6 - 2026-01-26¶
Changed¶
- Documentation cleanup and enhancements
0.1.5 - 2026-01-25¶
Added¶
- Anomaly Detection System: Complete ML-powered pipeline for detecting anomalies in OHLCV data using Isolation Forest, with Hamilton DAG-based architecture for scalable symbol-level processing
- Visualization: Anomaly plots with
mplfinanceintegration, featuring customizable styling, human-readable axis formatting (1K, 1M, 1B), and flexible date range filtering - Dataset Export: Export anomaly results to DynamoDB or local storage with configurable TTL support
- File Storage: Save plots and data locally or to S3 with the new
to_storeutility - Interactive Documentation: Jupyter notebook support in documentation via
mkdocs-jupyter, with comprehensive tutorials and API reference sections
Changed¶
- Enhanced plotting with modular
plot_ohlcv_anomaliesfunction for reusability - Improved error handling for edge cases in anomaly collection
Dependencies¶
- Added
scikit-learnandsf-hamilton[visualization]for ML and DAG execution - Added
mplfinance,matplotlib, and visualization support packages - Added
mkdocs-jupyterfor interactive documentation
0.1.4 - 2026-01-18¶
Added¶
- AWS Integration: Comprehensive utilities for SSM parameters, Secrets Manager, S3, and DynamoDB operations
- Dataset Management: Store and retrieve structured datasets locally or in DynamoDB with flexible mapping and serialization
- Enhanced ArcticDB setup with shared AWS session for S3 backend
Documentation¶
- New documentation modules:
aws.mdfor AWS utilities anddataset.mdfor dataset handling - Expanded
sources.mdandstorage.mdwith dataset examples
0.1.3 - 2026-01-18¶
Added¶
to_datasetfunction for saving structured datasets with DynamoDB or local storage support
0.1.2 - 2026-01-17¶
Added¶
- Automated Documentation: GitHub Actions workflow for deploying documentation on version tags
- MkDocs site with Material theme
- Comprehensive docstrings and usage examples for core modules
0.1.1 - 2026-01-17¶
Changed¶
- Updated Python requirement to
>= 3.12with improved cross-platform support
0.1.0 - 2026-01-17¶
Added¶
- Core Data Access: Functions for reading and writing OHLCV data to ArcticDB (
ohlcv_from_arcticdb,ohlcv_to_arcticdb) - Multi-Source Support: Fetch data from Intrinio (
ohlcv_from_intrinio) and Yahoo Finance (ohlcv_from_yfinance) - Securities Discovery: Retrieve securities lists from Intrinio via
securities_from_intrinio - Flexible Storage: Support for local and S3-backed ArcticDB instances
- Auto-Configuration: Automatic
.envfile generation on first import - GitHub Actions workflow for PyPI publishing
Documentation¶
- Complete README with installation, configuration, and usage examples
0.0.1 - 2026-01-14¶
Added¶
- Initial project structure with Intrinio SDK integration
- ArcticDB for time-series data storage
- MCP server implementation
- Docker support and environment configuration