The function Realm.applyEffects has an appendGenerator argument, which seems to indicate whether it is safe to modify the current generator. It is clearly unsafe to do so when it is anticipated that the applied effects might later have to be undone. Of the 37 invocations of this function, one explicitly unsets the flag, and three explicitly set it. PR #2185 relies on this interpretation of the

System information

• OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntu 16.04
• Ray installed from (source or binary): pip install ray
• Ray version: 0.7.4
• Python version: 3.6

Describe the problem

After having successfully trained and restore an agent, one very common use case might be to make deterministic action given a state. After training, or wh

I'm proposing that we, the closure-compiler team, update our planning process to include some form of publication of those plans where they can be easily seen by the Open Source community.

I expect this increased visibility will make people much more comfortable using closure-compiler.

At the moment, I'm thinking this would include:

For each quarter:

1. New features we're actively workin

The following files seem to be of key significance to how cvxpy 1.0's reduction system works, but they have little to no documentation. It's difficult for contributors to address canonicalization issues (related to correctness, or performance) without more information on these files.

• [reductions/canonicalization.py](https://github.com/cvxgrp/cvxpy/blob/34349b5e41c124a6a1e32426e68af95b5044498c

For example ee site info should allow output in different formats like yaml,json. This should happen to maintain parity with the other commands like ee site list, ee auth list global

Example:

ee site list --format=table
ee site list --format=json
ee auth list global --format=table
ee auth list global --format=json

Now,

# This works
ee site info example.

This seems wrong to me. The optimizer seems to be aware of its uncertainty, re-evaluating the "fluke point" and discovering that it's actually not usually that great. But then it gives me the fluke point at the end! This is especially problematic when variance is non-uniform and lower near the true minimum.

Ideally, the convergence plot would be updated as well, showing the expected minimum as

When jimplifying com.reddit.frontpage-214190.apk(Reddit v. 2.26.2), the following exception is thrown:

Exception in thread "main" Unexpected type null
	at soot.jimple.toolkits.typing.integer.ClassHierarchy.typeNode(ClassHierarchy.java:152)
	at soot.jimple.toolkits.typing.integer.TypeResolver.typeVariable(TypeResolver.java:121)
	at soot.jimple.toolkits.typing.integer.ConstraintCollecto

Hey Contributor!

Thanks for checking out HyperLearn!! Super appreciate it.

Since the package is new (only started like August 27th)..., Issues are the best place to start helping out, and or check out the Projects tab. There's a whole list of stuff I envisioned to complete.

Also, if you have a NEW idea: please post an issue and label it new enhancement.

In terms of priorities, I wanted