Docs/extras (#270)

* update installation instructions

* update agent instansiation in basic concepts

* update watch instructions in README

* fix np variance issue

Co-authored-by: Nota, Christopher <cnota@irobot.com>

README.md

@@ -46,24 +46,24 @@ It also contains implementations of the following "vanilla" agents, which provid
 ## Installation
 
 First, you will need a new version of [PyTorch](https://pytorch.org) (>1.3), as well as [Tensorboard](https://pypi.org/project/tensorboard/).
-Then, you can install the `autonomous-learning-library` through PyPi:
+Then, you can install the core `autonomous-learning-library` through PyPi:
 
 ```
 pip install autonomous-learning-library
 ```
 
-Alternately, you can install directly from this repository:
+You can also install all of the extras (such as Gym environments) using:
 
 ```
-git clone https://github.com/cpnota/autonomous-learning-library.git
-cd autonomous-learning-library
-pip install -e .
+pip install autonomous-learning-library[all]
 ```
 
-You can also install the prerequisites using:
+Finally, you can install directly from this repository including the dev dependencies using:
 
 ```
-pip install autonomous-learning-library[pytorch]
+git clone https://github.com/cpnota/autonomous-learning-library.git
+cd autonomous-learning-library
+pip install -e .[dev]
 ```
 
 ## Running the Presets
@@ -81,7 +81,7 @@ tensorboard --logdir runs
 ```
 
 and opening your browser to http://localhost:6006.
-Once the model is trained to your satisfaction, you can watch the trained model play using:
+Once the model is fully trained, you can watch the trained model play using:
 
 ```
 all-watch-atari Breakout "runs/a2c_[id]/preset.pt"

all/experiments/experiment.py

@@ -75,7 +75,7 @@ def _log_test_episode(self, episode, returns):
     def _log_test(self, returns):
         if not self._quiet:
             mean = np.mean(returns)
-            sem = np.variance(returns) / np.sqrt(len(returns))
+            sem = np.var(returns) / np.sqrt(len(returns))
             print('test returns (mean ± sem): {} ± {}'.format(mean, sem))
         self._writer.add_summary('returns-test', np.mean(returns), np.std(returns))
 

docs/source/guide/basic_concepts.rst

@@ -290,12 +290,12 @@ Here is a quick example:
     from all.environments import AtariEnvironment
 
     agents = [
-        atari.dqn(),
-        atari.ddqn(),
-        atari.c51(),
-        atari.rainbow(),
-        atari.a2c(),
-        atari.ppo(),
+        atari.dqn,
+        atari.ddqn,
+        atari.c51,
+        atari.rainbow,
+        atari.a2c,
+        atari.ppo,
     ]
 
     envs = [AtariEnvironment(env, device='cuda') for env in ['BeamRider', 'Breakout', 'Pong', 'Qbert', 'SpaceInvaders']]