google-deepmind / google-deepmind/pysc2
RGB pixels are all zero?
- Dominant language
- Python
- Stars
- 8.3k
- Forks
- 1.2k
- PR merge metrics
- No merged PRs in 30d
Description
I'm trying to get the RGB screen and minimap values. The exact code that I'm running is as such (I'm on a mac):
```
from pysc2.agents import base_agent, scripted_agent
from pysc2.env import sc2_env
from pysc2.lib import actions, features, units
from absl import app
def main(unused_argv):
agent = scripted_agent.MoveToBeacon()
try:
while True:
with sc2_env.SC2Env(map_name='MoveToBeacon',
players=[sc2_env.Agent(sc2_env.Race.zerg)],
agent_interface_format=features.AgentInterfaceFormat(
rgb_dimensions=features.Dimensions(screen=84, minimap=64),
),
step_mul=16,
game_steps_per_episode=0,
visualize=True) as env:
agent.setup(env.observation_spec(), env.action_spec())
timesteps = env.reset()
agent.reset()
while True:
step_actions = [agent.step(timesteps[0])]
if timesteps[0].last():
break
timesteps = env.step(step_actions)
except KeyboardInterrupt:
pass
if __name__ == '__main__':
app.run(main)
```
Basically, I'm just running the MoveToBeacon mini game agent.
However, when I check `timesteps[0].observation['rgb_screen']`, I am always getting (480, 276, 3) numpy array with all zeros!
* Getting all zeros for RGB values doesn't seem right. How do I fix this?
* 480, 276 dimension values look arbitrary since I set the dimension of screen as 84. How are the dimensions calculated for RGB screen and minimap values?
Contributor guide
Assessment
This issue has not been assessed yet.