Don't act on this I will keep updating it.

[russelljjarvis]

The travis build may not show completion, but I don't think it fails on something substantial. https://travis-ci.com/russelljjarvis/neuronunit/jobs/245014372

Here is a summary of developments.

• A lighter non-docker build, supporting: the fast forward euler, and the brian2 adexponential.
• A working travis.ci build.
• Simulated observation tests, with intuitively low error.
• For passive tests only, use of the new extract-feature method in the base class defined in passive.py (edited) All the Allen Tests should still work but I have not checked them in a while.
• Optimize is now a method that inherits from the class list. Making the syntactic sugar:

  results = NU_Test_List.optimize()

  possible

[review-notebook-app[bot]]

Check out this pull request on 

You'll be able to see Jupyter notebook diff and discuss changes. Powered by ReviewNB.