Jack Tang

@blankjul I'd like to fix it.

@CompRhys I suggest to design pymoo own `RandGenerator` class and keep the state rather than using variable. The simplest implementation may looks like: ``` import numpy as np class RandGenerator:...

> Seems like a much bigger refactor to be instantiating new PRNGs everywhere and then potentially runs the risk of users using the same seed in everything and ending up...

Hello @CompRhys , for multi-thread local random generator in pymoo, I didn't spend enough time on looking deep into it this week. And here is the document of multi threading...

> * Global Instance (like in this PR currently). I don't think this is thread-safe. > * Singleton Pattern (very similar but more flexible e.g. return a PRG for different...

Hello @blankjul, please consider thread-safe singleton: ``` class Singleton(object): _lock = threading.Lock() _instance = None @classmethod def instance(cls): if not cls._instance: with cls._lock: if not cls._instance: cls._instance = cls() return...

Well, I've tested both in macos and ubuntu, and above bug only occured in macos, I paste detailed error ``` Traceback (most recent call last): File "", line 1, in...

Hi jlouis, - I agreed to adding some high level API extensions and leave low level APIs in driver itself. - As to test cases I will add some later...

I have the same issue

Thanks for the reponse :) I switched to agent-trainer-docker and xvfb was installed, when I run `make train-new`, it printed: ``` root@3c44b3f8aeb0:/home/agent-trainer# make train-new python -m agent train-new 20161012|05:29:30|INFO: Session...