Open Issues Need Help
View All on GitHubAI Summary: Improve the unit test coverage of the Java Fast Wavelet Transform library's core classes to at least 70%, focusing on internal, CWT, MODWT, and utility packages. This involves writing new unit tests to cover edge cases, error handling, and currently untested code paths, prioritizing high-impact classes and builders. The goal is to ensure reliable and maintainable code and remove the need for deployment workarounds.
A high-performance Fast Wavelet Transform (FWT) library for Java 21+ featuring a type-safe, extensible architecture. Supports multiple wavelet families including orthogonal (Haar, Daubechies, Symlets, Coiflets), biorthogonal, and continuous wavelets. Built with zero dependencies, comprehensive validation, and extensive benchmarking.
AI Summary: Implement proper zero-padding boundary conditions for the MODWT (Maximal Overlap Discrete Wavelet Transform) in the Java library, addressing a current fallback to circular convolution. This involves extending the input signal with zeros, adjusting convolution indices, maintaining phase alignment, and ensuring consistency with MODWT mathematical properties. The implementation should affect the `ParallelMultiLevelMODWT` and `MultiLevelMODWTTransform` classes and related convolution operations.
A high-performance Fast Wavelet Transform (FWT) library for Java 21+ featuring a type-safe, extensible architecture. Supports multiple wavelet families including orthogonal (Haar, Daubechies, Symlets, Coiflets), biorthogonal, and continuous wavelets. Built with zero dependencies, comprehensive validation, and extensive benchmarking.