Merge pull request #516 from prabhatnagarajan/policy_docs

Adds policies to the documentation
chainer · Aug 20, 2019 · b0f003e · b0f003e
2 parents 8bee9da + c01ae56
commit b0f003e
Show file tree

Hide file tree

Showing 3 changed files with 39 additions and 0 deletions.
diff --git a/chainerrl/policies/gaussian_policy.py b/chainerrl/policies/gaussian_policy.py
@@ -31,8 +31,10 @@ class FCGaussianPolicy(chainer.ChainList, Policy):
         Let y as the output of the mean layer.
         If bound_mean=False:
             mean = y (if bound_mean=False)
+
         If bound_mean=True:
             mean = min_action + tanh(y) * (max_action - min_action) / 2
+
     The variance of the Gaussian is computed as follows:
         Let y as the output of the variance layer.
         variance = softplus(y) + min_var

diff --git a/docs/policies.rst b/docs/policies.rst
@@ -0,0 +1,36 @@
+========
+Policies
+========
+
+Policy interfaces
+=================
+
+.. autoclass:: chainerrl.policy.Policy
+   :members:
+
+   .. automethod:: __call__
+
+Policy implementations
+======================
+
+.. autoclass:: chainerrl.policies.ContinuousDeterministicPolicy
+
+.. autoclass:: chainerrl.policies.FCDeterministicPolicy
+
+.. autoclass:: chainerrl.policies.FCBNDeterministicPolicy
+
+.. autoclass:: chainerrl.policies.FCLSTMDeterministicPolicy
+
+.. autoclass:: chainerrl.policies.FCGaussianPolicy
+
+.. autoclass:: chainerrl.policies.FCGaussianPolicyWithStateIndependentCovariance
+
+.. autoclass:: chainerrl.policies.FCGaussianPolicyWithFixedCovariance
+
+.. autoclass:: chainerrl.policies.GaussianHeadWithStateIndependentCovariance
+
+.. autoclass:: chainerrl.policies.MellowmaxPolicy
+
+.. autoclass:: chainerrl.policies.SoftmaxPolicy
+
+.. autoclass:: chainerrl.policies.FCSoftmaxPolicy
diff --git a/docs/reference.rst b/docs/reference.rst
@@ -11,5 +11,6 @@ API Reference
    experiments
    explorers
    links
+   policies
    recurrent
    replay_buffers