From 47593785adf7146ae3aa877e1c22514a912090d2 Mon Sep 17 00:00:00 2001
From: Whi Kwon <30768596+whikwon@users.noreply.github.com>
Date: Thu, 30 May 2019 22:01:43 +0900
Subject: [PATCH 1/4] Fix success_count not reset issue (#59)

---
 scripts/envs/open_manipulator/ros_interface.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/envs/open_manipulator/ros_interface.py b/scripts/envs/open_manipulator/ros_interface.py
index 1c681fc..e47c385 100755
--- a/scripts/envs/open_manipulator/ros_interface.py
+++ b/scripts/envs/open_manipulator/ros_interface.py
@@ -306,7 +306,6 @@ class OpenManipulatorRosBaseInterface(object):
             self.success_count += 1
             if self.success_count == self.cfg["SUCCESS_COUNT"]:
                 print ("Current episode succeeded")
-                self.success_count = 0
                 return True
             else:
                 return False
@@ -388,6 +387,7 @@ class OpenManipulatorRosGazeboInterface(OpenManipulatorRosBaseInterface):
 
         # self.delete_target_block()
         self.init_robot_pose()
+        self.success_count = 0
         time.sleep(0.5)
 
         self.set_target_block(block_pose)

From 09b1f2e9a8e57d5062e37abd0b002c29178d452a Mon Sep 17 00:00:00 2001
From: Whi Kwon <30768596+whikwon@users.noreply.github.com>
Date: Thu, 30 May 2019 22:39:44 +0900
Subject: [PATCH 2/4] Fix episode unreset bug (#62)

---
 .../envs/open_manipulator/open_manipulator_reacher_env.py    | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/scripts/envs/open_manipulator/open_manipulator_reacher_env.py b/scripts/envs/open_manipulator/open_manipulator_reacher_env.py
index 1b2c730..7315b2f 100755
--- a/scripts/envs/open_manipulator/open_manipulator_reacher_env.py
+++ b/scripts/envs/open_manipulator/open_manipulator_reacher_env.py
@@ -76,10 +76,11 @@ class OpenManipulatorReacherEnv(gym.Env):
         if self.env_mode == "sim":
             self.reward = self.compute_reward()
             # TODO: Add termination condition
-            #            if self.ros_interface.check_for_termination():
-            #                self.done = True
+            # if self.ros_interface.check_for_termination():
+            #     self.done = True
             if self.ros_interface.check_for_success():
                 self.done = True
+                self.episode_steps = 0
 
         obs = self.ros_interface.get_observation()
 

From 9264c03f76ddfaf7b5406776dae1892e12091267 Mon Sep 17 00:00:00 2001
From: Whi Kwon <30768596+whikwon@users.noreply.github.com>
Date: Fri, 31 May 2019 15:23:59 +0900
Subject: [PATCH 3/4] Add installation guide for openmanipulator local training
 (#63)

---
 README.md | 26 +++++++++++++++++++++++++-
 1 file changed, 25 insertions(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 35ed2fd..c92b750 100644
--- a/README.md
+++ b/README.md
@@ -97,8 +97,11 @@ wandb login
 
 #### OpenManipulator
 
+Follow the ROS installation commands in [Dockerfile](https://github.com/kairproject/kair_algorithms_draft/blob/master/Dockerfile) to train.
+
 ```
-python run_open_manipulator_reacher_v0.py --algo [algo] --off-render --log
+roslaunch kair_algorithms open_manipulator_env.launch gui:=false &
+rosrun run_open_manipulator_reacher_v0.py --algo [algo] --off-render --log
 ```
 
 #### LunarLanderContinuous-v2
@@ -113,6 +116,27 @@ python run_lunarlander_continuous.py --algo [algo] --off-render --log
 python run_reacher_v1.py --algo [algo] --off-render --log
 ```
 
+## How to Test
+
+#### OpenManipulator
+
+```
+roslaunch kair_algorithms open_manipulator_env.launch gui:=false &
+rosrun python run_open_manipulator_reacher_v0.py --algo [algo] --off-render --test --load-from [trained_weight_path]
+```
+
+#### LunarLanderContinuous-v2
+
+```
+python run_lunarlander_continuous.py --algo [algo] --off-render --test --load-from [trained_weight_path]
+```
+
+#### Reacher-v1
+
+```
+python run_reacher_v1.py --algo [algo] --off-render --test --load-from [trained_weight_path]
+```
+
 ## How to Cite
 
 We are currently writing a white paper to summarize the results. We will add a BibTeX entry below once the paper is finalized.

From 9c2475498cd0d4a9c49b45a90208f947a01b9bb2 Mon Sep 17 00:00:00 2001
From: Whi Kwon <30768596+whikwon@users.noreply.github.com>
Date: Wed, 26 Jun 2019 23:33:38 +0900
Subject: [PATCH 4/4] Modify terminal done state (#65)

---
 scripts/envs/open_manipulator/open_manipulator_reacher_env.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/scripts/envs/open_manipulator/open_manipulator_reacher_env.py b/scripts/envs/open_manipulator/open_manipulator_reacher_env.py
index 7315b2f..3525ac3 100755
--- a/scripts/envs/open_manipulator/open_manipulator_reacher_env.py
+++ b/scripts/envs/open_manipulator/open_manipulator_reacher_env.py
@@ -85,7 +85,7 @@ class OpenManipulatorReacherEnv(gym.Env):
         obs = self.ros_interface.get_observation()
 
         if self.episode_steps == self._max_episode_steps:
-            self.done = True
+            self.done = False
             self.episode_steps = 0
 
         return obs, self.reward_rescale_ratio * self.reward, self.done, None