kastnerkyle

kastnerkyle

kastnerkyle

# -*- coding: utf-8 -*-

# Copyright (C) 2017 by Akira TAMAMORI

# Copyright (C) 2016 by hardmaru

# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to deal
# in the Software without restriction, including without limitation the rights
# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell

kastnerkyle

import numpy as np
import random as rand
import datetime

class QLearner(object):
    def __init__(self, \
        num_states=100, \
        num_actions = 4, \
        alpha = 0.2, \
        gamma = 0.9, \

kastnerkyle

# See: Hochreiter, S., & Schmidhuber, J. (1996).
# Bridging long time lags by weight guessing and "Long Short-Term Memory".
# Spatiotemporal models in biological and artificial systems, 37, 65-72.

import numpy as np
import torch
from torch import nn
from torch.nn import functional as F
from torch.autograd import Variable
import sys

kastnerkyle

kastnerkyle

import tensorflow as tf

def gumbel_ST(logits, temp=1.0, hard=False):
    eps = 1e-8
    gumbel_noise = -tf.log(-tf.log(tf.random_uniform(tf.shape(logits)) + eps)
                           + eps)

    y = tf.nn.softmax((logits + gumbel_noise) / temp)

    if hard:

kastnerkyle

# Author: Mathieu Blondel
# License: BSD 3 clause

import numpy as np


def projection_simplex(V, z=1, axis=None):
    """
    Projection of x onto the simplex, scaled by z:
        P(x; z) = argmin_{y >= 0, sum(y) = z} ||y - x||^2

kastnerkyle

kastnerkyle

def policyIterSP(game):
    nnet = initNNet()                                       # initialise random neural network
    examples = []    
    for i in range(numIters):
        for e in range(numEps):
            examples += executeEpisode(game, nnet)          # collect examples from this game
        new_nnet = trainNNet(examples)                  
        frac_win = pit(new_nnet, nnet)                      # compare new net with previous net
        if frac_win > threshold: 
            nnet = new_nnet                                 # replace with new net