tsu-nera

import numpy as np
import ev3dev.ev3 as ev3

import random, os, time
import pickle


# qlearningAgents.py
# ------------------
## based on http://inst.eecs.berkeley.edu/~cs188/sp09/pacman.html

tsu-nera

import gym
import numpy as np
from qlearning_answer import QLearningAgent

env = gym.make("CartPole-v0")
n_actions = env.action_space.n

def build_state(features):
    """get our features and put all together converting into an integer"""
    return int("".join(map(lambda feature: str(int(feature)), features)))

tsu-nera

import gym
import numpy as np
import gym_tic_tac_toe
from math import floor

env = gym.make('tic_tac_toe-v0')
n_states = 3 ** 9 # 状態数
n_actions = 9 # 行動数
eM = 1000 # 評価を行うエピソード数

tsu-nera

import numpy as np
import gym
from gym.spaces import Discrete, Box

# ================================================================
# Policies
# ================================================================

class DeterministicDiscreteActionLinearPolicy(object):

tsu-nera

tsu-nera

import gym
import numpy as np
import gym_tic_tac_toe
import random
from math import floor
import matplotlib.pyplot as plt

def random_plus_middle_move(moves, p):
    if ([p, 4] in moves):
        m = [p, 4]

tsu-nera

import gym
import numpy as np
from gym import wrappers

env = gym.make("FrozenLake8x8-v0")
env = wrappers.Monitor(env, '/tmp/frozenlake-experiment-2')
env.reset()
n_states = env.observation_space.n
n_actions = env.action_space.n

tsu-nera

import gym
import numpy as np
from gym import wrappers

env = gym.make("FrozenLake-v0")
env = wrappers.Monitor(env, '/tmp/cartpole-experiment-5')

env.reset();
n_states = env.observation_space.n
n_actions = env.action_space.n

tsu-nera

import gym
import numpy as np
from gym import wrappers

env = gym.make("FrozenLake-v0")
env = wrappers.Monitor(env, '/tmp/cartpole-experiment-3')

env.reset();
n_states = env.observation_space.n
n_actions = env.action_space.n

tsu-nera

import gym
import numpy as np
from gym import wrappers

env = gym.make("FrozenLake-v0")
env = wrappers.Monitor(env, '/tmp/cartpole-experiment-2')

env.reset();
n_states = env.observation_space.n
n_actions = env.action_space.n