thecedl / homework2 Goto Github PK

ImportError Traceback (most recent call last)
in ()
2 import tensorflow as tf
3 import numpy as np
----> 4 from policy_gradient import util
5 from policy_gradient.policy import CategoricalPolicy
6 from policy_gradient.baselines.linear_feature_baseline import LinearFeatureBaseline

/home/haiyang/homework2/policy_gradient/util.py in ()
1 from gym.spaces import Box, Discrete
2 import numpy as np
----> 3 from scipy.signal import lfilter
4
5 def flatten_space(space):

ImportError: No module named 'scipy'

NameError Traceback (most recent call last)
in ()
2
3 # Construct a neural network to represent policy which maps observed state to action.
----> 4 in_dim = util.flatten_space(env.observation_space)
5 out_dim = util.flatten_space(env.action_space)
6 hidden_dim = 8

NameError: name 'util' is not defined

NameError Traceback (most recent call last)
in ()
3 path_length = 200
4 discount_rate = 0.99
----> 5 baseline = LinearFeatureBaseline(env.spec)
6
7 po = PolicyOptimizer(env, policy, baseline, n_iter, n_episode, path_length,

NameError: name 'LinearFeatureBaseline' is not defined

I solve at 109 iters, am I correct? (a little bit more than 80 iters)

Hi all, for people who suffer from using iPython, you can also download and work on this assignment using HW2_Policy_Graident.py provided here.

Basically, this file contains the same code as HW2_Policy_Graident.ipynb.

Assignment Submission

When submitting assignment, just add the newly downloaded HW2_Policy_Graident.py file into your homework2 repo and push it to your forked repository.

Then, open a Pull Request to this repository.

Math Formula

If you are writing assignment in this manner, you can find math formula provided in the description of problem 2 here: