Open mdavis-xyz opened 5 years ago
@mlda065 policy_gradient.py is already there in a repo. clone again and run.
Ok, so I'm trying again and it's still not straightforward.
The README says to install gym
. So I run
pip install gym
It also says to run
pip install -e '.[box2d]'
But that command fails. I've never seen that notation before. pip install box2d
also fails. What is that command supposed to do?
So I can't run the lunar_lander example.
I try python run_cartpole.py
. That fails because I don't have tensorflow
installed.
The README should say that you need to install tensorflow as well as gym.
pip install tensorflow
python run_cartpole.py
Now it fails because I don't have matplotlib installed. The README should say that you need to install matplotlib as well as gym and tensorflow.
pip install matplotlib
python run_cartpole.py
Now it fails because I don't have tkinter installed.
pip install tkinter
pip install tk
pip install ttk
Those commands all fail. I can't figure out how to install this dependency.
This is the reason why a requirements.txt
file would be useful.
Once I figure out how to install the dependencies get both demos working, I'll submit a pull request to fix the README and add a requirements.txt file.
So how can I get the demos working?
@mlda065 apologies for not including a requirements.txt. Could you try following the tutorial/readme here? https://github.com/openai/gym#what-s-new and here https://gym.openai.com/docs/
Let me know if you are able to run those demos
There are 2 solutions to get it working:
import matplotlib
matplotlib.use('agg')
import matplotlib.pyplot as plt
2.
apt-get install python3-tk
I don't know what the 'agg'
thing really is. But it seems like a nicer solution than an apt
dependency.
Shall I submit a pull request with option 1, or option 2?
In addition to having requirements.txt
gym
tensorflow
matplotlib
Also, there's a typo in the README, it says "carpole" not "cartpole".
And the README should be modified to state that the funny pip install command for box2d should be run inside the cloned gym repo.
Can you please add a
requirements.txt
or similar?I'm trying to run this and python throws an error because it doesn't know the library
policy_gradient
. I can't find any similarly named library on pypi.What library is that contained in?