Cherry pick regen road functionality

araffin / learning-to-drive-in-5-minutes

Implementation of reinforcement learning approach to make a car learn to drive smoothly in minutes

MIT License

284 stars 88 forks source link

Closed tleyden closed 5 years ago

tleyden commented 5 years ago

Cherry pick this functionality from https://github.com/tawnkramer/sdsandbox/blob/master/src/predict_server.py#L120-L134 to send a directive to the simulator to regenerate the road.

Usage example:

    donkeyEnv = env.envs[0].env
    donkeyEnv.regen_road(rand_seed=int(time.time()))

araffin commented 5 years ago

Hello,

I added those changes in my new PR https://github.com/araffin/learning-to-drive-in-5-minutes/pull/9 (which require master version of stable-baselines, that's why I did not merge it yet) I finally made it working, in fact it came from my computer language, see https://github.com/tawnkramer/sdsandbox/pull/27

tleyden commented 5 years ago

Great! Closing this PR since it looks to be redundant.