mahdiolfat/ssp

Statistical Digital Signal Processing and Modeling

Statistical Digital Signal Processing and Modeling

Direct impelementation of statistical digital signal processing algorithms from the book by the same name by Professor Monson H. Hayes.

The book represents algorithms as:

1. set of instructions (pseudo-code) annotated in Tables.
2. matlab code annotated in Figures.
3. described in words in the body of the text.

Chapters

1. Introduction
2. Background
3. Discrete-Time Random Processes
4. Signal Modeling
5. The Levinson Recursion
6. Lattice Filters
7. Optimum Filters
8. Spectrum Estimation
9. Adaptive Filters

TODO

• add proper reference to the book
• lint markdown readme
• finalize test cases for all codes
• python documentation generator
• mypy static type checking
• release package scripts
• stub files?
• VS code add-ons
• use matlab-equivalent psuedo-inverse ("X \ R") operation to match the book
• performance benchmarking would be cool

ssp's Issues

Random processes implementation and testing: Chapter 3 and Chapter 4

Implement and test the ARMA processes.

sanity check all inputs/outputs AND add unittests for all conditions.

Set License to BSD

remove unnecessary pip packages from the testing workflow (e.g., jupyter)

Cross-platform vs code project definition including developer extensions

devops: release, deploy, and create versioned artifacts

enforce strict static type-checking with mypy

Readme glowup

Test coverage of Chapter 7 - Wiener Processes and Optimal Filters.

Ensure all implemented functions in optimal.py have a test case with an assert for pass/fail under tests/test_optimal.py

Project directory re-structuring (cosmetic) before first release.

Move pyssp --> ssp
Move all requirements*.txt and dev files into a new dev folder.

use stubgen and stubtest to ensure consistency on stub files

          use stubgen and stubtest to ensure consistency between the code and the stub files. Focus of developer should be on the .py static type checking, type annotations, implementation, and documentation, stub files will aim to be automated to provide a clean overview of how the api evolves (grows) over time.

Originally posted by @molfat in #9 (comment)

At least one example notebook per chapter

Well documented, well explained jupyter notebooks re-creating book examples.

Use pseudo inverse to match matlab/book behaviour

Sphinx apidoc generator support.

Test coverage of Chapter 8 - Spectrum Estimation

Ensure all implemented functions in spectrum.py have a test case with an assert for pass/fail under tests/test_spectrum.py

mypy github workflow run command and pass/fail

Include pytests (tests/**) in static linting, formatting, and typecheckking

Test coverage of Chapter 4 - Modeling of Stochastic Processes / Systems

Ensure all implemented functions in modeling.py have a test case with an assert for pass/fail under tests/test_modeling.py

Test coverage of Chapter 3 - Stochastic Processes / Systems

Populate tests/test_system.py

Levinson recursion algorithm implementations and testing of Chapter 5

MKDocs api doc generator support.

Test coverage of Chapter 9 - Adaptive Filtering

Populate tests/test_adaptive.py

Create stubs and enumerate implemented vs. not implemented algorithms

Test coverage of Chapter 6 - Lattice Filters

Populate tests/test_lattice.py