def test_video(setup):
import numpy as np
def im(x, y):
canvas = np.ones((640, 480), dtype=np.float32) * 0.001
for i in range(200):
for j in range(200):
if x - 5 < i < x + 5 and y - 5 < j < y + 5:
canvas[i, j] = 1
return canvas
frames = [im(100 + i, 80) for i in range(20)]
logger.save_video(frames, "test_video.mp4")
def test_video_gif(setup):
import numpy as np
def im(x, y):
canvas = np.zeros((200, 200))
for i in range(200):
for j in range(200):
if x - 5 < i < x + 5 and y - 5 < j < y + 5:
canvas[i, j] = 1
return canvas
frames = [im(100 + i, 80) for i in range(20)]
logger.save_video(frames, "test_video.gif")
def test_video_gif(setup):
import numpy as np
def im(x, y):
canvas = np.zeros((200, 200))
for i in range(200):
for j in range(200):
if x - 5 < i < x + 5 and y - 5 < j < y + 5:
canvas[i, j] = 1
return canvas
frames = [im(100 + i, 80) for i in range(20)]
logger.save_video(frames, "test_video.gif")
assert 'test_video.gif' in logger.glob('*.gif')
logger.save_video(frames, "/tmp/ml-logger-test/videos/test_video.gif")
assert 'ml-logger-test/videos/test_video.gif' in logger.glob(
'**/videos/*.gif', wd="/tmp")
def evaluate(env, agent, num_episodes, save_video=None):
from ml_logger import logger
episode_rewards, frames = [], []
for i in trange(num_episodes, desc="Eval"):
obs = env.reset()
done = False
episode_reward = 0
while not done:
with utils.Eval(agent):
action = agent.select_action(obs)
obs, reward, done, _ = env.step(action)
if save_video:
frames.append(env.render('rgb_array', width=64, height=64))
episode_reward += reward
if save_video:
logger.save_video(frames, key=save_video)
logger.store_metrics(episode_reward=episode_reward)
episode_rewards.append(episode_reward)
return np.mean(episode_rewards)