nrontsis / PILCO

Bayesian Reinforcement Learning in Tensorflow

MIT License

313 stars 84 forks source link

Update to GPflow v2 and include safety extensions #45

Closed nrontsis closed 4 years ago

nrontsis commented 4 years ago

ToDo list:

[x] Update README.md to say that the Safe PILCO part is an example extension and that test coverage is not counted there.
[x] Write a small readme for the safe part
[x] Move safe_pilco folder to the root folder and rename to safe_pilco_extension
[x] Remove safe_pilco from coverage count
[x] Move examples, utils in the examples folder

Nice to do:

[x] Remove comments
[ ] Put comments in some example scripts where code is hard to read.
[x] Address comments in this PR
[x] Fix Travis

nrontsis commented 4 years ago

@kyr-pol I changed most of the things I suggested in my comments.

I think the most important remaining bits in the ToDo list are:

Update READMEs
~~Move examples in the examples folder~~
Check that the safe pilco still runs after the moving of the files into different folders

codecov-io commented 4 years ago

Codecov Report

Merging #45 into master will not change coverage. The diff coverage is n/a.

@@           Coverage Diff           @@
##           master      #45   +/-   ##
=======================================
  Coverage   73.59%   73.59%           
=======================================
  Files           9        9           
  Lines         462      462           
=======================================
  Hits          340      340           
  Misses        122      122

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 5f65fbb...5f65fbb. Read the comment docs.

kyr-pol commented 4 years ago

Checked the safe pilco examples, and they work fine by 8b0a52c. Was there a reason for naming the folder safe-pilco-extension and not something like safe_pilco_extension? I renamed it to make the imports easier.