From 14a359daee3c5c6bc81e6fa3166c73c09590635b Mon Sep 17 00:00:00 2001
From: Guillaume Mollard <guillaume.mollard2@gmail.com>
Date: Thu, 27 Jun 2019 15:50:54 +0200
Subject: [PATCH] done metric
---
RLLib_training/train_experiment.py | 17 +++++++----------
1 file changed, 7 insertions(+), 10 deletions(-)
diff --git a/RLLib_training/train_experiment.py b/RLLib_training/train_experiment.py
index cd25ad0..654a100 100644
--- a/RLLib_training/train_experiment.py
+++ b/RLLib_training/train_experiment.py
@@ -50,13 +50,7 @@ def on_episode_start(info):
episode = info['episode']
map_width = info['env'].envs[0].width
map_height = info['env'].envs[0].height
- episode.horizon = map_width + map_height
-
-
-# def on_episode_step(info):
-# episode = info['episode']
-# print('#########################', episode._agent_reward_history)
-# # print(ds)
+ episode.horizon = 3*(map_width + map_height)
def on_episode_end(info):
@@ -64,17 +58,20 @@ def on_episode_end(info):
score = 0
for k, v in episode._agent_reward_history.items():
score += np.sum(v)
- score /= (len(episode._agent_reward_history) * 3 * episode.horizon)
+ score /= (len(episode._agent_reward_history) * episode.horizon)
+ done = 1
+ if len(episode._agent_reward_history) == episode.horizon:
+ done = 0
episode.custom_metrics["score"] = score
+ episode.custom_metrics["proportion_episode_solved"] = done
def train(config, reporter):
print('Init Env')
set_seed(config['seed'], config['seed'], config['seed'])
- config['map_height'] = config['map_width']
- # Example configuration to generate a random rail
+ # Environment parameters
env_config = {"width": config['map_width'],
"height": config['map_height'],
"rail_generator": config["rail_generator"],
--
GitLab