joyncarpio · joyncarpio · Oct 4, 2019 · Oct 8, 2019 · Oct 8, 2019 · Oct 12, 2019
diff --git a/.coveragerc b/.coveragerc
@@ -5,17 +5,21 @@ omit =
     *__init__*
     */scripts/*
     */setup.py
-    *rllab*
     */flow/utils/shflags
     *dummy_server.py
     *aimsun.py
     *aimsun/run.py
     *aimsun/api.py
     *aimsun/generate.py
     *aimsun/load.py
+    *flow/multiagent_*
+    *flow/singleagent_*
 
 exclude_lines =
     if __name__ == .__main__.:
     raise NotImplementedError
     @ray.remote
+    @abstractmethod
     def policy_mapping_fn*
+    def main(args)*
+    pragma: no cover
diff --git a/.github/ISSUE_TEMPLATE/question.md b/.github/ISSUE_TEMPLATE/question.md
@@ -8,7 +8,10 @@ about: If you have a question or run into some problem with Flow (e.g. during th
 PLEASE DO NOT ASK YOUR QUESTIONS HERE ON GITHUB!
 
 If you have a question or run into some problem (e.g. during installation or when running an example)
-with Flow, please direct your technical questions to Stack Overflow using the "flow-project" tag.
+with Flow, please direct technical questions to the Flow project Slack channel here:
 
-link: https://stackoverflow.com/questions/tagged/flow-project
-tag: flow-project
+https://join.slack.com/t/flow-users/shared_invite/enQtODQ0NDYxMTQyNDY2LTY1ZDVjZTljM2U0ODIxNTY5NTQ2MmUxMzYzNzc5NzU4ZTlmNGI2ZjFmNGU4YjVhNzE3NjcwZTBjNzIxYTg5ZmY
+
+If you have a non-technical inquiry, please send us an email:
+
+https://flow-project.github.io/contact.html
diff --git a/.travis.yml b/.travis.yml
@@ -54,6 +54,17 @@ before_install:
   - export AIMSUN_SITEPACKAGES="/home/travis/miniconda/envs/aimsun_flow"
   - export AIMSUN_NEXT_PATH="/home/user/Aimsun_Next_XXX"
 
+  # Install stable-baselines
+  - pip install stable_baselines==2.7.0
+
+  # Install h-baselines
+  - pushd $HOME
+  - git clone https://github.com/AboudyKreidieh/h-baselines.git
+  - pushd h-baselines
+  - pip install -e .
+  - popd
+  - popd
+
   - ls ../
 
 install:
@@ -62,7 +73,6 @@ install:
   - pip install -e .
   - pip install coveralls
   - pip install jupyter
-  - pip install stable_baselines==2.7.0
 
 script:
   - nose2 --with-coverage

diff --git a/CODEOWNERS b/CODEOWNERS
@@ -2,7 +2,7 @@
 # Each line is a file pattern followed by one or more owners.
 
 # These owners will be the default owners for everything in the repo.
-*       @cathywu @eugenevinitsky @AboudyKreidieh @kanaadp
+*       @eugenevinitsky @AboudyKreidieh @kjang96
 
 # Order is important. The last matching pattern has the most precedence.
 # So if a pull request only touches javascript files, only these owners

diff --git a/README.md b/README.md
@@ -21,7 +21,7 @@ See [our website](https://flow-project.github.io/) for more information on the a
 
 # Technical questions
 
-Please direct your technical questions to [Stack Overflow](https://stackoverflow.com) using the [flow-project](https://stackoverflow.com/questions/tagged/flow-project) tag.
+If you have a bug, please report it. Otherwise, join the [Flow Users group](https://join.slack.com/t/flow-users/shared_invite/enQtODQ0NDYxMTQyNDY2LTY1ZDVjZTljM2U0ODIxNTY5NTQ2MmUxMzYzNzc5NzU4ZTlmNGI2ZjFmNGU4YjVhNzE3NjcwZTBjNzIxYTg5ZmY) on Slack!  
 
 # Getting involved
 

diff --git a/docs/libsumo_mac.md b/docs/libsumo_mac.md
@@ -0,0 +1,23 @@
+# How to install Libsumo for Mac OS
+
+This is adapted from an email exchange with the SUMO staff.
+
+
+
+To install libsumo requires re-building and installing SUMO from source.
+
+## Steps
+
+- **Install swig:** ```brew install swig```
+- **Clone the repo:** ```git clone https://github.com/eclipse/sumo.git```
+- **Create a “cmake-build” directory inside sumo/build/ and navigate to it:** ```mkdir build/cmake-build && cd build/cmake-build```
+
+**The next 3 steps are inside that directory**
+
+- ```cmake ../..```
+- ```make```
+- ```make install```
+
+## Additional Notes
+- You can test if libsumo has been built looking at (./testlibsumo) inside the sumo/bin/ directory.
+- Bear in mind to use libsumo with the same Python version with which CMake built SUMO.
diff --git a/docs/source/conf.py b/docs/source/conf.py
@@ -32,9 +32,6 @@ def __getattr__(cls, name):
 
 
 MOCK_MODULES = [
-    'rllab',
-    "rllab.core.serializable.Serializable",
-    'rllab.sampler.utils.rollout',
     "AAPI",
     'PyANGKernel',
     'AAPI.GKGUISystem',

diff --git a/docs/source/examples.rst b/docs/source/examples.rst
@@ -9,7 +9,7 @@ A few points of clarification:
 
 * Occasionally we describe environments as *fully observed*. By this we mean that all relevant pieces of information (speed, relative positions, traffic light states) of the system are available to the controller. *Partially observed* refers to only a subset being available. This is the default configuration of the environment and the set of observations can be customized.
 
-* Each environment has customizable environment/network parameters that can be used to configure it beyond what is described here. Particularly pertinent parameters are described. Additional parameters can be found by examining the documentation in the relevant scenario and environment files.
+* Each environment has customizable environment/network parameters that can be used to configure it beyond what is described here. Particularly pertinent parameters are described. Additional parameters can be found by examining the documentation in the relevant network and environment files.
 
 * In the figures below, the following key is used for vehicle colors, where AV stands for autonomous vehicle.
 
@@ -22,12 +22,12 @@ Figure Eight
 The figure-eight is a closed-ring version of an intersection. The goal is to maximize
 the system-wide velocity for fourteen vehicles,
 which necessitates spacing the vehicles so that they don't
-run into conflicts at the merging points. The scenario is fully observed: all vehicles
+run into conflicts at the merging points. The network is fully observed: all vehicles
 speeds and positions are visible to the controller.
-This scenario is also a benchmark, and has been
+This network is also a benchmark, and has been
 extensively tested at three penetration rates: 1 AV 13 humans, 7 AVs 7 humans, 14 AVs.
 
-The scenario, pictured below,
+The network, pictured below,
 is relatively light-weight and can be trained the quickest. It can serve both as a test
 that the training process is working correctly and as a study of the difficulty of controlling
 many vehicles simultaneously.
@@ -43,15 +43,14 @@ In this scenario, human drivers placed on a ring develop a travelling wave that
 decelerations and lowers the average velocity of the system. The goal is to train a single
 autonomous vehicle to eliminate the shockwave.
 
-The pre-built run script for running this in human-only mode is `examples/sumo/sugiyama.py`.
-The run scripts in rllib and aimsun are `examples/rllib/stabilizing_the_ring.py` and
-`examples/aimsun/stabilizing_the_ring.py`. These control environments are partially observed:
-the autonomous vehicle only observes its own distance to the leading vehicle, its speed,
-and the speed of the leading vehicle.
+The experiment config for this in human-only mode (non-RL) is `examples/exp_configs/non_rl/ring.py`.
+The experiment config for RL-based is `examples/exp_configs/rl/singleagent/singleagent_ring.py`. 
+These control environments are partially observed: the autonomous vehicle only observes its own 
+distance to the leading vehicle, its speed, and the speed of the leading vehicle.
 
 To make this task more difficult, the environment has a configurable parameter, `ring_length`, which
 can be set to a list containing the minimum and maximum ring-size. The autonomous vehicle must
-learn to distinguish these scenarios from each other and pick the appropriate driving behavior.
+learn to distinguish these networks from each other and pick the appropriate driving behavior.
 
 .. image:: ../img/stabilizing_the_ring.png
    :width: 400