Finter v0.1 Release Notes
Last updated
Last updated
DateConverter Utility: A new utility has been added to facilitate the retrieval of pre start dates. This utility is expected to enhance date-related operations within the system.
Command Documents with GPT Integration: The software now includes an integration with GPT (Generative Pre-trained Transformer) for generating command documents (cmdocs). This feature leverages the capabilities of GPT to improve the generation of documentation and command-related content.
(Empty - No specific bug fixes mentioned in the commit messages.)
(Empty - No known issues have been identified or listed for this release.)
Test code has been added to ensure the reliability and correctness of the new features.
Several new files and configurations have been added to support the new DateConverter and GPT integration for command documents.
AWS Integration: Added support for AWS credentials management. This includes new endpoints and documentation for handling AWS credentials.
Parquet File Support: Introduced the get_parquet_df function to support loading Sample CM data using parquet files, enhancing data handling capabilities.
Fixed a typo in the codebase that could potentially cause issues in text interpretation or execution.
CI/CD Updates: Automated the release note generation in the CI/CD pipeline to streamline the deployment process.
Testing: Added test code for AWS credentials API and response handling to ensure reliability and correctness.
Documentation: Provided new documentation for the AWS Credentials API and response format to assist users in understanding and utilizing the new features.
Tag message for 0.1.25: Load Sample CM by using parquet
Content Factory Implementation: A new Content Factory has been added for utilizing content model name aliases. This feature allows for more flexible content model management and usage within the system.
Catalog Fixes: Some issues within the catalog have been addressed and fixed to ensure smoother operation and data handling.
No known issues have been reported or documented for this release.
The release includes updates to various files and the addition of new scripts and modules to support the new features and fixes. Notably, loader.py has been added to the content_model directory, and convert.py has been added to the finter-gradio submodules, among other changes.
Implemented user event log tracking within the notebook submission process. This feature enhances the ability to monitor and audit user actions during the submission of notebooks, providing better insights and traceability.
Add data catalog (beta)
Not specified in the provided commit messages.
Not specified in the provided commit messages.
Code refactoring and cleanup have been performed, as indicated by the changes in various modules related to the framework model and settings. Additionally, the deletion of finter/utils/user.py suggests a removal of deprecated or unused user utilities.
Note: For a detailed list of bug fixes and known issues, please refer to the project's issue tracker or version control history.
Statistics Filter in Simulation: A new statistics_filter has been added to the simulation feature, allowing for more refined control over statistical data during simulations.
Quanda Bot Integration: A Quanda Bot has been introduced, offering automated Q&A support for users, enhancing the user experience by providing instant responses to common queries.
A fix has been applied to the test suite to address unspecified issues.
[Empty - To be filled in later if there are any known issues]
CodeResponse.md has been added to the documentation.
Updates have been made to SimulationRequest.md and SubmissionResponse.md to reflect the latest changes.
The test configuration has been updated, with the addition of test/simulation/test_config.py and the deletion of test/test_submission_config.py.
The sample notebook samples/finter_qc.ipynb has been modified, likely to demonstrate the new features or fixes.
Benchmark Addition (FL-632)
What's New: Introducing a benchmark feature for performance comparison.
Benefits: Enables users to compare their model's performance against a standard benchmark for better analysis.
Dashboard Improvements (FL-629): Some issues with the dashboard have been addressed to improve user experience and functionality.
No new known issues have been reported with this release.
Various minor updates and improvements have been made to the framework model and submodules, specifically within the submission and gradio apps components.
New Features
Raw Return Addition (FL-621): The software now includes functionality to handle raw returns, providing more flexibility in data processing.
Class Declaration Extraction (FL-628): Improved the structure of the code by extracting the class declaration from cell picking, which enhances maintainability and readability.
Notebook Submission Helper Warning (FL-630): A new warning has been added to alert users when declaring NotebookSubmitHelper to ensure they are aware of the SAVE operation.
The release notes do not specify any particular bug fixes. If there were any bugs addressed in this release, please provide the details here.
There are no known issues explicitly listed for this release. If there are any ongoing issues that have not been resolved, please list them here.
The release includes updates to various files such as helper_notebook.py, helper_simulation.py, notebook.py, settings.py, and some sample and test notebooks, indicating improvements and optimizations in these areas.
Bucket Name in SM API: The SM API has been updated to include bucket_name as part of its functionality, allowing for more granular control and identification of storage resources.
User Tracking: A new User Tracking feature has been implemented, which can be utilized through the User API. This feature is designed to enhance user management and activity monitoring.
Finter Import Optimization: Clarified instructions on optimizing finter imports to improve performance and efficiency in code execution.
Gallery Component: Addressed an issue where the gallery component would not load images under certain conditions. This fix ensures a smoother user experience when browsing image galleries.
No known issues have been reported for this release.
Added UserApi.md to the documentation to provide guidance on the new User Tracking feature.
Updated SmApi.md to reflect the changes made to the SM API, including the addition of bucket_name.
Introduced test_user_api.py to ensure the User API functions as expected and to validate the new user tracking feature.
Implemented conditional functionality for the 'halo work' feature, enhancing the application's responsiveness and user experience.
Dashboard enhancements for better credential management.
Fine-tuning of application performance for a smoother user interface.
Added loading status indicators during API calls to inform users about ongoing processes.
Improved validation processes to ensure data integrity and user input accuracy.
[List any specific bugs that were fixed in this release]
[List any known issues that are still present in this release]
Cookbook Integration: A new cookbook.py file has been added, which likely provides users with a set of predefined recipes or patterns for common tasks or problems.
Gradio Dashboard: The introduction of a Gradio dashboard, which is an easy way to create ML demos, suggests that users can now interact with the models through a web interface. This includes several Gradio apps for different functionalities such as GPT-simple, PNL, and sector analysis, each with its own Dockerfile and requirements, indicating a significant expansion in the user interface capabilities.
File Exclusion: The update includes a fix to exclude files with a certain pattern (.~~), which may help in avoiding unnecessary files in the project or during deployment.
[Empty - To be filled later if any known issues are present]
The addition of Docker support files like docker-compose.yml, run.sh, and run_dev.sh suggests improvements in containerization and deployment processes.
New thumbnails for chat, pnl, and sector have been added to the Gradio apps, likely for a better user experience.
The helper_github.py file has been modified, which might include improvements or fixes related to GitHub integration or operations.
GitHub Sync: A new feature has been introduced that allows synchronization with GitHub. This feature likely includes new helper modules for GitHub operations and configurations, as well as updates to existing submission-related functionalities to support GitHub integration.
No specific bug fixes have been mentioned in the commit messages provided.
No known issues have been provided.
Added .env-template which might be used for setting up environment variables necessary for the new GitHub sync feature.
Updated finter/framework_model/submission/notebook.py which could include enhancements or modifications related to the submission process in the context of the new GitHub feature.
Updated samples/finter_qc.ipynb, possibly to demonstrate the new features or to provide examples of how to use the updated submission process.
Added test/test_submission_github.py which suggests that new tests have been written to ensure the GitHub sync feature works as expected.
GPT Server Publishing: The GPT server has been published, allowing for easier access and integration.
URL Change: The URL for the GPT server has been updated to reflect the new publishing status.
Initialization Check: Added an initialization check test to ensure that the system is correctly set up before use.
No known issues have been reported for this release.
Improvements to internal scripts and testing mechanisms have been made to enhance the development workflow.
No new features have been listed for this release.
Fixed an issue with missing .__init__.py file which could potentially cause problems with package discovery and imports in the finter/utils/testing directory.
No known issues have been reported for this release.
No additional notes or updates have been provided for this section.
Calendar API Integration: A new Calendar API has been integrated into the system, allowing users to access calendar-related features directly from the README documentation.
Trading Day Information: Users can now retrieve trading day information through the newly added Calendar API, which includes a dedicated response model for trading days.
Function Name Cleanup: Cleaned up function names for better readability and consistency across the codebase.
Unified Response Structure: Standardized the response structure for the Calendar API to ensure consistency in the data returned to the users.
No known issues have been reported or documented for this release.
Test Code Added: Added test code for the Calendar API to ensure its functionality and reliability.
Documentation Updates: Updated the README and documentation to reflect the new Calendar API features and usage.
Codebase Cleanup: Removed unnecessary generate.sh scripts that were no longer needed in the project. Tag Message:
FL-370 add calendar on readme - The Calendar API has been documented in the README for easier access and understanding by the users.
Quanda Bot Integration: A new feature quanda_bot has been added to the SDK, which likely provides an interface for interacting with a question-and-answer style bot within the software ecosystem.
ModelData Load Functionality: The ModelData.load function has been introduced, providing a mechanism for loading model data. This might be related to the legacy C2DF.load functionality, offering an improved or updated way to handle data loading processes.
DataFrame Index Type Fix: An issue with the DataFrame index type has been resolved, ensuring that the index behaves as expected within the software's data handling components.
[Empty - To be filled in later if applicable]
Documentation updates and additional tests have been added to support the new features, including QuandaBotApi.md and test files for ModelData and quanda_bot_api.
Simulation & Validation in NotebookSubmissionHelper: Added functionality for one-pass research, allowing users to simulate and validate within the NotebookSubmissionHelper.
ModelStat for Simulation Results: A new feature to present simulation results in a more readable and attractive format.
Timer in NotebookSubmissionHelper: Introduced a timer to track the duration of processes within the NotebookSubmissionHelper.
Pretty NotebookSubmissionHelper Log: Enhanced the logging within NotebookSubmissionHelper for better readability.
Automatic Validation and Position Extraction: Improved the submission process by handling edge cases for automatic validation and position extraction.
Configuration Error Fix: Resolved an issue where a configuration error occurred when importing finter.
[Empty - To be filled later if applicable]
Codebase Enhancements: The codebase has seen improvements in error readability and model configuration.
Testing and Pipeline Optimization: Added pytest and optimized the CI/CD pipeline to ensure code quality and reliability.
We are excited to announce the release of Finter 0.1.0, a comprehensive library that combines finance and quantitative analysis for a variety of purposes. This release includes a host of new features, improvements, and the initial setup for many of our core components.
CM (Data) Initialization: Set up your data environment with ease. Initially focusing on Korean stock market data, with plans to include U.S. stock market data in the future.
AM (Alpha Model) Initialization: Start building your alpha models right away.
PM (Portfolio Model) Initialization: Create and manage your portfolio models efficiently.
Model Validation Initialization: Validate your models to ensure their accuracy and reliability.
Model Submission Initialization: Submit your models for review or deployment.
Simulation Initialization: Test your strategies with our simulation tools.
Dashboard Initialization: Monitor your models and portfolios with a comprehensive dashboard.
Finter Handbook: Access a detailed guide to using Finter effectively.
Playground Initialization: Explore Finter's capabilities with ongoing additions of usage examples.
Model Production Initialization: Live production is now available, allowing you to deploy your models in real-time.
Off Limit for Date Range (CM): No more restrictions on the date range for your data.
.env Settings: Keep your API keys secure and out of your notebooks with environment variable settings.
Submission Helper: Simplify the submission process with notebook cell extraction, submission settings, and the ability to run .py files and submit directly.
Change Log: Keep track of updates and changes with a detailed change log.
Chat with GPT (General API): Engage in conversations with GPT for general inquiries and assistance.
Log In: Secure and personalized access to the Finter platform.
More Convenience Features: Look forward to additional features designed to enhance your user experience.
(Empty - To be filled later if applicable)
(Empty - To be filled later if applicable)
(Empty - To be filled later if applicable)
We are starting with a strong focus on the Korean stock market, ensuring that our users have comprehensive and detailed data at their fingertips. This is just the beginning, as we have plans to expand our data offerings to include the U.S. stock market, providing even more versatility and options for our users.
We hope you enjoy the new features and improvements in Finter 0.1.0. Stay tuned for more updates as we continue to enhance the platform. Thank you for your support!
: An 'easy simulation' class has been introduced to simplify the process of setting up and running simulations.
: Select your benchmark during model setup. Options include using the default, specifying a custom benchmark, or opting out.