demo.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Demo Notebook\n",
    "This notebook is intended to act as a primer and overview of the functionality found in the various libraries written for this repository. Additional comments and information can be found in the inline comments within the library scripts.\n",
    "\n",
    "As a whole, this repository is intended to be a series of functions and libraries to establish a simple, simulated mobile robot in an environment with a smooth function of interest distributed in it. This robot is gathering information about this data, and our code allows the robot to make myopic or nonmyopic planning decisions with respect to the previous observations it has made. We will use this notebook to walk you through setting up an enviroment, creating a robot, and simulating the planning actions it can take."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Primer\n",
    "\n",
    "Several key concepts are critical to the development of this repository. This section intends to walk through some of these in order to build more detailed intuition about what the robot is doing in it's adaptive sampling regime.\n",
    "\n",
    "### Gaussian Processes\n",
    "We assume that the function distributed in the environment is smooth and Gaussian -- think geography (rolling hills), temperature plumes in water, or oil slicks. This allows us to use *Gaussian Processes* (GPs) as a model of our environment, and provides closed-form solutions for incrementally updating the robot's belief model as it explores, allowing for probabilistic predictions of values, and more.\n",
    "\n",
    "### POMDPs\n",
    "A *Partially-Observable Markov Decision Process* (POMDP) is a problem which consists of a value function (a sense of reward), a set of actions (which can be executed to accumulate reward), a set of transition probabilities (the probabiliy to actually executing an action and ending up where we anticipate), observations (which is what we get when we take an action), a set of states (which is the status of the robot between actions or transitions, and where observations can occur), and occasionally other factors like a discount factor (the idea that as we accumulate more actions, their value can decrease). To be *partially observable* means that some of the states, observations, or actions/transitions are not known to us prior to selecting an action, and we end up discovering those things along the way. For an adaptive sampling problem like ours, we assume that we always know where the robot is, but not what it will observe (as in, the environment itself is only partially observable at any action selection).\n",
    "\n",
    "### Monte Carlo Tree Search\n",
    "There are a number of ways to solve POMDP problems, and we've ultimately decided to go with a tree search. A *Monte Carlo Tree Search* (MCTS) walks through potential actions that a vehicle can take at any moment, and predicts the most fruitful chains of actions to explore based on the value function (which we're learning as we go as we take observations). The leafs in the tree search can be of one of two forms: a \"belief\" node -- that is, what the robot believes its state and the environment's state to be, and a belief-action node -- that is, what the robot believes its state and environment's state to be *after* taking an action. MCTS allows us to expand these types of leaves, using accumulated predicted value. The tree is expanded to a certain *horizon* before it determines the accumulated reward of the action sequence it has planned. It will repeat these expansions to a specific computation budget.\n",
    "\n",
    "### Quantifying Valuable Actions\n",
    "In order to select the most valueable type of action at any step, we need to select a reward function which will encourage the type of behavior we desire -- in our case, we want the robot to find, with confidence, the \"best\" or \"most optimal\" or \"largest maxima\" in the environment. You could consider this the place with the largest methane signal in a hydrothermal vent site, or the deepest canyon in a hilly landscape. While it is cananical to use a value function called the *Upper Confidence Bound* (UCB) which weighs the predicted mean and variance of a point; we've opted to use the *Maximum Value Information* (MVI) metric, inspired by work by Wang and Jegelka on [Maximum-Value Entropy Search](https://arxiv.org/pdf/1703.01968.pdf). This function explicitly weighs the entropy of the predicted maximum value to the predicted value of another state in the world. As we observe more, our estimate of the maximum-value improves (since we assume a smooth, Gaussian distribution), and our estimate of the value of a state becomes more accurate. \n",
    "\n",
    "### The Typical Routine\n",
    "The robot initially starts out with no *a priori* information -- that is, everywhere in the world looks just as good as anywhere else in the world. It selects an initially random action. It executes this action, and along the way collects true observations of the environment, and updates its belief model. At the end of the action, the robot selects a new action, now considering the updated belief model. It continues this pattern until some termination is reached (either it has exhausted a time or computation budget, or perhaps it has converged on a desired state). "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports\n",
    "\n",
    "There are several libraries which are external to what is provided by our repository. Please be aware that you will need to be running Python 2.7+, and you will need to have *maptlotlib, sklearn, iPython, numpy, scipy, time* and *itertools* available. Additionally, you will likely need to download *GPy* and *dubins*, non-standard python libraries. You can get these with the following commands (in a terminal):\n",
    "        \n",
    "        ``` pip install GPy ```\n",
    "        ``` pip install dubins ```\n",
    "\n",
    "Or you can visit [this page](https://github.com/SheffieldML/GPy) for GPy, and [this page](https://github.com/AndrewWalker/Dubins-Curves) for dubins download details."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#standard imports\n",
    "import os\n",
    "import time\n",
    "import sys\n",
    "import logging\n",
    "import numpy as np\n",
    "\n",
    "#repository libraries, which contain calls for matplotlib, GPy, dubins, etc.\n",
    "import aq_library as aqlib\n",
    "import mcts_library as mctslib\n",
    "import gpmodel_library as gplib \n",
    "import evaluation_library as evalib \n",
    "import paths_library as pathlib \n",
    "import envmodel_library as envlib \n",
    "import robot_library as roblib\n",
    "import obstacles as obslib"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Default Parameters\n",
    "We use a few \"global\" variables to parameterize our robot and our environment. \n",
    "\n",
    "* SEED - specifies a foundational distribution from which the environment function can be drawn. Takes on any number 0 through 1 million\n",
    "* REWARD_FUNCTION - specifies how a piece of information/an observation by the robot is valued; can take the value *mes* for Maximum Entropy Search value, *mean* for the Upper Confidence Bound value, or *exp_improve* for the Expected Improvement Value\n",
    "* PATHSET - the path primitives available to the vehicle; can take on the value *fully_reachable* for point-to-point or *dubins* for a pre-specified set of dubins curves\n",
    "* USE_COST - a boolean for whether to consider path length when planning\n",
    "* NONMYOPIC - a boolean for whether the planner that the robot uses should be myopic or nonmyopic\n",
    "* GOAL-ONLY - a boolean for whether only the destination point (goal) is used to determine the reward of a trajectory, or whether the entire trajectory is considered\n",
    "* TREE_TYPE - for a nonmyopic planner, the search tree can be selected; can take on the value *belief* for a classical Monte Carlo Search Tree or *dpw* for a tree which utilizes progressive-widening techniques\n",
    "\n",
    "* MIN_COLOR - for plotting, the value of the smallest observation available\n",
    "* MAX_COLOR - for plotting, the value of the largest observation available"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# You can set these depending on the type of robot and environment you want to generate\n",
    "SEED =  0\n",
    "REWARD_FUNCTION = 'mes' #robot will use Maximum Value Information as value function\n",
    "PATHSET = 'dubins' #robot will have a set of pre-defined action sets to use for each step of planning\n",
    "USE_COST = False #since all of the path options will be nominally the same length anyway in the dubins setting, robot doesn't consider path length cost\n",
    "NONMYOPIC = True #robot will consider future outcomes for each action\n",
    "GOAL_ONLY = False #robot will consider the whole trajectory with respect to reward\n",
    "TREE_TYPE = 'dpw' #robot will use progressive widening in the search tree\n",
    "\n",
    "# Parameters for plotting based on the seed world information\n",
    "MIN_COLOR = -25.\n",
    "MAX_COLOR = 25."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Logging and IPython Set-Up\n",
    "The next two code cells set up a logging system which will write files to your local machine; saving the mission run for later review, and allowing the figures and other elements that will generate during a simulated mission to populate within this notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Set up paths for logging the data from the simulation run\n",
    "if not os.path.exists('./figures/' + str(REWARD_FUNCTION)): \n",
    "    os.makedirs('./figures/' + str(REWARD_FUNCTION))\n",
    "logging.basicConfig(filename = './figures/'+ REWARD_FUNCTION + '/robot.log', level = logging.INFO)\n",
    "logger = logging.getLogger('robot')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%aimport ipp_library\n",
    "%autoreload 1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Creating the Environment\n",
    "\n",
    "We will establish a \"world\" in which our robot will explore. We will use the *obstacles* and *Environment Library* in order to create these."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# First, establish the extent of the world; assume rectangular\n",
    "ranges = (0., 10., 0., 10.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Obstacles\n",
    "We can create a number of different types of obstacle worlds. Examples below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "ow = obslib.FreeWorld() #a world without obstacles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f84282f8d90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ow_channel = obslib.ChannelWorld(extent=ranges, \n",
    "                                 opening_location=(3.5, 7.), \n",
    "                                 opening_size= 3., \n",
    "                                 wall_thickness=0.3)\n",
    "ow_channel.draw_obstacles()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f83f6be6890>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ow_bug = obslib.BugTrap(extent=ranges, \n",
    "                        opening_location=(2.2, 3.0), \n",
    "                        opening_size=4.6, \n",
    "                        orientation = 'left', \n",
    "                        width = 5.0)\n",
    "ow_bug.draw_obstacles()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f83f60d1510>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ow_blocks = obslib.BlockWorld(extent=ranges,\n",
    "                              num_blocks=12, \n",
    "                              dim_blocks=(1., 1.), \n",
    "                              centers=[(2.5, 2.5), (7.,4.), (5., 8.), (8.75, 6.), (3.5,6.), (6.,1.5), (1.75,5.), (6.2,6.), (8.,8.5), (4.2, 3.8), (8.75,2.5), (2.2,8.2)])\n",
    "ow_blocks.draw_obstacles()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Function\n",
    "Now, we need to create the underlying function that we want the robot to discover."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Current environment in violation of boundary constraint. Regenerating!\n",
      "Environment initialized with bounds X1: ( 0.0 , 10.0 )  X2:( 0.0 , 10.0 )\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f83f5e63350>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "world = envlib.Environment(ranges = ranges,\n",
    "                           NUM_PTS = 20, \n",
    "                           variance = 100.0, \n",
    "                           lengthscale = 1.0, \n",
    "                           visualize = True,\n",
    "                           seed = SEED,\n",
    "                           MIN_COLOR=MIN_COLOR, \n",
    "                           MAX_COLOR=MAX_COLOR, \n",
    "                           obstacle_world = ow,\n",
    "                           noise=10.0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluation\n",
    "\n",
    "We use the *Evaluation Library* in order to create an \"evaluator\" object which can be used by the robot to determine how well it is doing as it explores."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "World max value 27.744079714634324 at location [2.10526316 1.05263158]\n"
     ]
    }
   ],
   "source": [
    "# Create the evaluation class used to quantify the simulation metrics\n",
    "evaluation = evalib.Evaluation(world = world, reward_function = REWARD_FUNCTION)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The Robot\n",
    "Finally, we establish our point robot using a number of different parameters, which are fully detailed in the *Robot Library* in this repository."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "robot = roblib.Robot(sample_world = world.sample_value, #function handle for collecting observations\n",
    "                     start_loc = (5.0, 5.0, 0.0), #where robot is instantiated\n",
    "                     extent = ranges, #extent of the explorable environment\n",
    "                     kernel_file = None,\n",
    "                     kernel_dataset = None,\n",
    "                     #prior_dataset =  (data, observations), \n",
    "                     prior_dataset = None,\n",
    "                     init_lengthscale = 1.0, \n",
    "                     init_variance = 100.0, \n",
    "                     noise = 10.0001,\n",
    "                     #noise = 0.5000,\n",
    "                     path_generator = PATHSET, #options: default, dubins, equal_dubins, fully_reachable_goal, fully_reachable_step\n",
    "                     goal_only = GOAL_ONLY, #select only if using fully reachable step and you want the reward of the step to only be the goal\n",
    "                     frontier_size = 15,\n",
    "                     horizon_length = 1.5, \n",
    "                     turning_radius = 0.05,\n",
    "                     sample_step = 0.5,\n",
    "                     evaluation = evaluation, \n",
    "                     f_rew = REWARD_FUNCTION, \n",
    "                     create_animation = True, #logs images to the file folder\n",
    "                     learn_params = False, #if kernel params should be trained online\n",
    "                     nonmyopic = NONMYOPIC,\n",
    "                     discretization = (20, 20), #parameterizes the fully reachable sets\n",
    "                     use_cost = USE_COST, #select if you want to use a cost heuristic\n",
    "                     MIN_COLOR = MIN_COLOR,\n",
    "                     MAX_COLOR = MAX_COLOR,\n",
    "                     computation_budget = 250.0,\n",
    "                     rollout_length = 5,\n",
    "                     obstacle_world = ow, \n",
    "                     tree_type = TREE_TYPE) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Executing a Mission\n",
    "We have created simple method calls within the robot class in order to execute a mission."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0 ] Current Location:   (5.0, 5.0, 0.0)\n",
      "Current predicted max and value: \t[0. 0.] \t0.0\n",
      "Setting c to : 1.0\n",
      "Rollouts completed in 1.68485593796s\n",
      "Number of rollouts: 250\n",
      "# nodes in tree: 10950\n",
      "[(17, 5.0), (17, 5.0), (17, 5.0), (17, 5.0), (16, 5.0), (17, 5.0), (16, 5.0), (16, 5.0), (17, 5.0), (16, 5.0), (16, 5.0), (17, 5.0), (17, 5.0), (17, 5.0), (17, 5.0)]\n",
      "[ 1 ] Current Location:   (5.335149778313376, 6.44263505832286, 1.3695422350712745)\n",
      "Current predicted max and value: \t[5.51724138 6.89655172] \t4.9449031356635995\n",
      "Setting c to : 1.0\n",
      "Starting global optimization 0 of 3\n",
      "Max Value in Optimization \t \t[23.94244467]\n",
      "Starting global optimization 1 of 3\n",
      "Max Value in Optimization \t \t[31.98301938]\n",
      "Starting global optimization 2 of 3\n",
      "Max Value in Optimization \t \t[36.57238735]\n",
      "Rollouts completed in 3.03742408752s\n",
      "Number of rollouts: 250\n",
      "# nodes in tree: 8230\n",
      "[(2, 3.1077275376689384), (2, 3.58387482401934), (2, 3.7229715318278624), (2, 3.7558707470018127), (3, 4.270147255360079), (4, 4.405765696072762), (3, 4.198927261527207), (3, 3.9255601560293827), (5, 4.517367684610121), (4, 4.454761894099985), (208, 6.358245430556713), (4, 4.331220328506457), (3, 4.156109293669881), (2, 3.5408666670084914), (3, 4.136676628543246)]\n",
      "[ 2 ] Current Location:   (4.259038641062613, 7.475699947544114, 2.3926666608418206)\n",
      "Current predicted max and value: \t[5.51724138 6.89655172] \t5.930126448930239\n",
      "Setting c to : 1.0\n",
      "Starting global optimization 0 of 3\n",
      "Max Value in Optimization \t \t[24.67027073]\n",
      "Starting global optimization 1 of 3\n",
      "Max Value in Optimization \t \t[22.84316292]\n",
      "Starting global optimization 2 of 3\n",
      "Max Value in Optimization \t \t[19.35874949]\n",
      "Rollouts completed in 2.94990706444s\n",
      "Number of rollouts: 250\n",
      "# nodes in tree: 5850\n",
      "[(1, 4.944418907477537), (2, 5.28882452254316), (1, 5.020513905729065), (4, 6.313293979466501), (6, 6.903245003847236), (2, 5.967241192850201), (6, 6.869617795598372), (3, 6.3562579863428), (208, 8.895293274480546), (6, 6.9334297579946), (2, 5.897951301342231), (3, 6.410295962462153), (2, 5.786715101608825), (2, 5.041674254702767), (2, 5.978685199082117)]\n",
      "[ 3 ] Current Location:   (2.885574574067233, 8.077902643472463, 2.730262550672748)\n",
      "Current predicted max and value: \t[2.4137931  8.27586207] \t7.116810257554767\n",
      "Setting c to : 1.0\n",
      "Starting global optimization 0 of 3\n",
      "Max Value in Optimization \t \t[21.22854989]\n",
      "Starting global optimization 1 of 3\n",
      "Max Value in Optimization \t \t[31.36711827]\n",
      "Starting global optimization 2 of 3\n",
      "Max Value in Optimization \t \t[17.73270349]\n",
      "Rollouts completed in 2.95842885971s\n",
      "Number of rollouts: 250\n",
      "# nodes in tree: 5975\n",
      "[(1, 4.519686072096609), (2, 5.5794392771617725), (1, 4.790894682509349), (3, 5.917023886333531), (5, 6.510178778232175), (3, 6.139472687407177), (3, 5.986859436162365), (2, 5.3439589643808185), (3, 5.876927665210651), (6, 6.643801209201784), (2, 5.463834714992711), (209, 8.853979382659256), (2, 5.394838306459221), (3, 6.130068223224929), (5, 6.547247473041102)]\n",
      "[ 4 ] Current Location:   (2.0015725969398335, 6.889600581772427, 4.0998047857440225)\n",
      "Current predicted max and value: \t[2.06896552 7.5862069 ] \t12.793833672574113\n",
      "Setting c to : 1.0\n",
      "Starting global optimization 0 of 3\n",
      "Max Value in Optimization \t \t[26.58019643]\n",
      "Starting global optimization 1 of 3\n",
      "Max Value in Optimization \t \t[28.06659533]\n",
      "Starting global optimization 2 of 3\n",
      "Max Value in Optimization \t \t[25.34701714]\n",
      "Rollouts completed in 2.97776794434s\n",
      "Number of rollouts: 250\n",
      "# nodes in tree: 5009\n",
      "[(2, 4.729534603215753), (2, 4.890939507840167), (2, 5.013884923123339), (2, 5.098192585378339), (2, 5.269801482555728), (2, 5.404990813932779), (226, 8.464661349802459), (2, 5.225112247908052), (2, 5.039005998843944), (2, 5.584139701327141), (2, 5.3472871581150105), (1, 4.824550450559592), (1, 4.774009123012092), (1, 4.79687071267464), (1, 4.5990917238666)]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f83e730d590>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Establisht the number of \"steps\" or \"actions\" the robot should take in the total mission using T\n",
    "robot.planner(T = 5) #robot will plan 5 actions\n",
    "robot.visualize_trajectory(screen = True) #creates a summary trajectory image\n",
    "# robot.plot_information() #plots all of the metrics of interest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.14"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}