Open philihp opened 2 years ago
Hello, can I work on this?
Go for it :)
Please ask questions if you get stuck on anything. I would suggest a different and separate PRs for each model, and perhaps another that renames config vars to be more in line with the paper. I would prefer multiple smaller, incremental PRs.
"EPSILON" should be "KAPPA"
src/models/
such that they match the variables defined in the PDF. This will make the code easier to maintain and perhaps help others port to other languages.sigSqToCiq
should probably be some variation ofqEta
, since it's related to ηq defined on page 287, §3.1.1