Skip to content

Instantly share code, notes, and snippets.

@yashpatel5400

yashpatel5400/cartpole.py

Created July 25, 2017 08:15
Show Gist options
  • Save yashpatel5400/69f0d8b10560343d3c6903f1cf77c648 to your computer and use it in GitHub Desktop.
Save yashpatel5400/69f0d8b10560343d3c6903f1cf77c648 to your computer and use it in GitHub Desktop.
Download ZIP
OpenAI CartPole w/ Keras
Raw
cartpole.py
"""
__name__ = predict.py
__author__ = Yash Patel
__description__ = Full prediction code of OpenAI Cartpole environment using Keras
"""
import gym
import numpy as np
from keras.models import Sequential
from keras.layers import Dense, Dropout
def gather_data(env):
num_trials = 10000
min_score = 50
sim_steps = 500
trainingX, trainingY = [], []
scores = []
for _ in range(num_trials):
observation = env.reset()
score = 0
training_sampleX, training_sampleY = [], []
for step in range(sim_steps):
# action corresponds to the previous observation so record before step
action = np.random.randint(0, 2)
one_hot_action = np.zeros(2)
one_hot_action[action] = 1
training_sampleX.append(observation)
training_sampleY.append(one_hot_action)
observation, reward, done, _ = env.step(action)
score += reward
if done:
break
if score > min_score:
scores.append(score)
trainingX += training_sampleX
trainingY += training_sampleY
trainingX, trainingY = np.array(trainingX), np.array(trainingY)
print("Average: {}".format(np.mean(scores)))
print("Median: {}".format(np.median(scores)))
return trainingX, trainingY
def create_model():
model = Sequential()
model.add(Dense(128, input_shape=(4,), activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(256, activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(512, activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(256, activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(128, activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(2, activation="softmax"))
model.compile(
loss="categorical_crossentropy",
optimizer="adam",
metrics=["accuracy"])
return model
def predict():
env = gym.make("CartPole-v0")
trainingX, trainingY = gather_data(env)
model = create_model()
model.fit(trainingX, trainingY, epochs=5)
scores = []
num_trials = 50
sim_steps = 500
for _ in range(num_trials):
observation = env.reset()
score = 0
for step in range(sim_steps):
action = np.argmax(model.predict(observation.reshape(1,4)))
observation, reward, done, _ = env.step(action)
score += reward
if done:
break
scores.append(score)
print(np.mean(scores))
if __name__ == "__main__":
predict()
Raw
data.py
"""
__name__ = data.py
__author__ = Yash Patel
__description__ = Gathers the data for the Cartpole environment into the
X and Y numpy arrays for training
"""
import gym
import numpy as np
def gather_data(env):
min_score = 50
sim_steps = 500
trainingX, trainingY = [], []
scores = []
for _ in range(10000):
observation = env.reset()
score = 0
training_sampleX, training_sampleY = [], []
for step in range(sim_steps):
# action corresponds to the previous observation so record before step
action = np.random.randint(0, 2)
one_hot_action = np.zeros(2)
one_hot_action[action] = 1
training_sampleX.append(observation)
training_sampleY.append(one_hot_action)
observation, reward, done, _ = env.step(action)
score += reward
if done:
break
if score > min_score:
scores.append(score)
trainingX += training_sampleX
trainingY += training_sampleY
trainingX, trainingY = np.array(trainingX), np.array(trainingY)
print("Average: {}".format(np.mean(scores)))
print("Median: {}".format(np.median(scores)))
return trainingX, trainingY
if __name__ == "__main__":
env = gym.make("CartPole-v0")
trainingX, trainingY = gather_data(env)
Raw
model.py
"""
__name__ = model.py
__author__ = Yash Patel
__description__ = Defines model to be trained on the Cartpole data,
predicting the directioal action to take given 4D observation state
"""
from keras.models import Sequential
from keras.layers import Dense, Dropout
def create_model():
model = Sequential()
model.add(Dense(128, input_shape=(4,), activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(256, activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(512, activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(256, activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(128, activation="relu"))
model.add(Dropout(0.6))
model.add(Dense(2, activation="softmax"))
model.compile(
loss="categorical_crossentropy",
optimizer="adam",
metrics=["accuracy"])
return model
if __name__ == "__main__":
model = create_model()
Raw
predict.py
"""
__name__ = predict.py
__author__ = Yash Patel
__description__ = Does the prediction using the defined model and data
"""
import gym
import numpy as np
from data import gather_data
from model import create_model
def predict():
env = gym.make("CartPole-v0")
trainingX, trainingY = gather_data(env)
model = create_model()
model.fit(trainingX, trainingY, epochs=5)
scores = []
num_trials = 50
sim_steps = 500
for trial in range(num_trials):
observation = env.reset()
score = 0
for step in range(sim_steps):
action = np.argmax(model.predict(observation.reshape(1,4)))
observation, reward, done, _ = env.step(action)
score += reward
if done:
break
scores.append(score)
print(np.mean(scores))
if __name__ == "__main__":
predict()
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment