quantling / pyndl / 364

83%

push

travis-ci-com

Adds Widrow-Hoff learning rule in different variants (#191)

* adds Widrow-Hof learning (real_to_real, real_to_binary, binary_to_real)
* tests for multiple cues and outcomes are still missing
* adds fast correlation code
* drop python 3.6 support and add python 3.9 support
* changes API from `number_of_threads / number_of_processes` to `n_jobs`
* adds scipy as an independencies
* new code only runs on linux for now as it uses openmp for parallelisation

Co-authored-by: Elnaz Shafaei <elnaz.shafaei@gmail.com>

400 of 400 new or added lines in 4 files covered. (100.0%)

1075 of 1289 relevant lines covered (83.4%)

2.5 hits per line