- Introduced Cma class for calculating the Cumulative Moving Average using Welford's algorithm with FMA for precision.
- Added methods for batch processing and streaming updates.
- Implemented a comprehensive markdown documentation for CMA, covering its mathematical foundation, performance profile, and use cases.
- Enhanced existing trend indicators (Bessel, Butter, Htit, Jma, Mama, Ssf, Vidya) with FMA for improved numerical stability and precision.
- Updated Adosc to utilize a single-pass algorithm for performance optimization.
- Fixed date initialization in benchmarks to ensure UTC consistency.
- Cleaned up code by removing unused using directives from various test and implementation files in the trends and volume directories.
- This includes files related to HMA, HTIT, JMA, KAMA, LSMA, MAMA, MGDI, PWMA, RMA, SMA, SSF, SUPER, T3, TEMA, TRIMA, USF, VIDYA, WMA, ATR, ADL, and ADOSC.
- Improved code readability and maintainability by streamlining imports.
- Introduced Usf.Validation.Tests.cs to validate the USF (Ehlers Ultimate Smoother Filter) for consistency across batch, streaming, and span modes, as well as mathematical properties and coefficient calculations.
- Added comprehensive tests for the ATR indicator in Atr.Quantower.Tests.cs, including constructor validation, historical data processing, and handling of NaN/Infinity inputs.
- Enhanced Atr.Tests.cs with additional tests for iterative corrections, warmup behavior, and true range calculations.
- Updated Atr.cs to ensure warmup period is derived from RMA.
- Added new tests for Adosc in Adosc.Tests.cs to validate handling of NaN and Infinity inputs, and to ensure batch calculations match iterative results.
- Created a new Volatility.csproj to organize volatility-related implementations.
- Removed unnecessary blank lines in multiple test files to enhance readability.
- Ensured consistent spacing and formatting in the `Trima`, `Usf`, `Vidya`, `Wma`, and `Atr` test classes.
- Updated comments for clarity and consistency in the `Atr` and `Adl` classes.
- Adjusted project files for better structure and maintainability.
refactor: enhance Bilateral and SMA indicators to handle edge cases and improve state management
refactor: clean up whitespace and formatting in various test files for consistency
- Updated the Prime method signature in multiple indicators (Jma, Kama, Lsma, Mama, Mgdi, Pwma, Rma, Sma, Ssf, Super, T3, Tema, Trima, Usf, Vidya, Wma, Atr) to accept an optional TimeSpan parameter for improved flexibility.
- Added unit tests for Lsma to verify Dispose functionality, ensuring proper unsubscription from the source and thread safety.
- Enhanced Mama and Wma classes to handle non-finite inputs gracefully and added checks for valid parameters in constructors.
- Introduced additional tests for T3 to validate constructor behavior with invalid volume factors.
- Ensured all indicators maintain consistent behavior when handling edge cases, such as empty buffers and non-finite values.
- Updated event handler signatures to use TValueEventArgs for consistency in Mama, Mgdi, Pwma, Rma, Sma, Ssf, Super, T3, Tema, Trima, Usf, Vidya, Wma, and Atr classes.
- Enhanced argument validation by specifying parameter names in exceptions for clarity.
- Adjusted tests to align with new event handler signatures.
- Improved code readability and maintainability by using structured records and lambda expressions.
feat(trends): implement IDisposable in Bessel and Conv classes to manage event subscriptions
fix(trends): add validation for period and parameters in Kama and MGDI calculations
fix(trends): clamp logarithmic calculations in JMA to avoid -Infinity
Miha Kralj