From 08489b46faffe3bfdeaa1287fa108260e74e194d Mon Sep 17 00:00:00 2001
From: Reggie McLean <reginald.mclean@ryerson.ca>
Date: Mon, 29 Jul 2024 10:26:27 -0400
Subject: [PATCH] expert actions page

---
 docs/index.md             |  2 +-
 docs/usage/basic_usage.md | 36 ++++++++++++++++++++++++++++++++++++
 2 files changed, 37 insertions(+), 1 deletion(-)
 create mode 100644 docs/usage/basic_usage.md

diff --git a/docs/index.md b/docs/index.md
index d236a704b..330d76293 100644
--- a/docs/index.md
+++ b/docs/index.md
@@ -43,7 +43,7 @@ obs, reward, terminate, truncate, info = env.step(a)
 introduction/basic_usage
 installation/installation
 rendering/rendering
-
+usage/basic_usage
 ```
 
 
diff --git a/docs/usage/basic_usage.md b/docs/usage/basic_usage.md
new file mode 100644
index 000000000..cc2443ff9
--- /dev/null
+++ b/docs/usage/basic_usage.md
@@ -0,0 +1,36 @@
+---
+layout: "contents"
+title: Generate data with expert policies
+firstpage:
+---
+
+# Generate data with expert policies
+
+## Expert Policies
+For each individual environment in Meta-World (i.e. reach, basketball, sweep) there are expert policies that solve the task. These policies can be used to generate expert data for imitation learning tasks.
+
+## Using Expert Policies
+The below example provides sample code for the reach environment. This code can be extended to the ML10/ML45/MT10/MT50 sets if a list of policies is maintained.
+
+
+```python
+from metaworld import MT1
+
+from metaworld.policies.sawyer_reach_v2_policy import SawyerReachV2Policy as p
+
+mt1 = MT1('reach-v2', seed=42)
+env = mt1.train_classes['reach-v2']()
+env.set_task(mt1.train_tasks[0])
+obs, info = env.reset()
+
+policy = p()
+
+done = False
+
+while not done:
+    a = policy.get_action(obs)
+    obs, _, _, _, info = env.step(a)
+    done = int(info['success']) == 1
+
+
+```