notebooks/14 document databases/intro-to-accidents.ipynb

   1 {
   2  "metadata": {
   3   "name": "",
   4   "signature": "sha256:3102d98f7297b3c0cd70c28f9fa0390011d2edcac4efb4e92679d52ad77dbeb7"
   5  },
   6  "nbformat": 3,
   7  "nbformat_minor": 0,
   8  "worksheets": [
   9   {
  10    "cells": [
  11     {
  12      "cell_type": "markdown",
  13      "metadata": {},
  14      "source": [
  15       "# Basic CRUD operations in Mongo\n",
  16       "\n",
  17       "This is a notebook to take you through some basic CRUD operations with Mongo."
  18      ]
  19     },
  20     {
  21      "cell_type": "code",
  22      "collapsed": false,
  23      "input": [
  24       "# Import the required libraries and open the connection to Mongo\n",
  25       "\n",
  26       "import collections\n",
  27       "import datetime\n",
  28       "from itertools import chain, repeat\n",
  29       "import matplotlib as mpl\n",
  30       "import matplotlib.pyplot as plt\n",
  31       "from mpl_toolkits.basemap import Basemap\n",
  32       "%matplotlib inline\n",
  33       "\n",
  34       "import numpy as np\n",
  35       "import pandas as pd\n",
  36       "import scipy.stats\n",
  37       "\n",
  38       "import pymongo\n",
  39       "from bson.objectid import ObjectId\n",
  40       "client = pymongo.MongoClient('mongodb://ogedei:27017/')"
  41      ],
  42      "language": "python",
  43      "metadata": {},
  44      "outputs": [],
  45      "prompt_number": 1
  46     },
  47     {
  48      "cell_type": "code",
  49      "collapsed": false,
  50      "input": [
  51       "db = client.small_accidents\n",
  52       "accidents = db.accidents\n",
  53       "labels = db.labels\n",
  54       "roads = db.roads"
  55      ],
  56      "language": "python",
  57      "metadata": {},
  58      "outputs": [],
  59      "prompt_number": 2
  60     },
  61     {
  62      "cell_type": "code",
  63      "collapsed": false,
  64      "input": [
  65       "# Load the expanded names of keys and human-readable codes into memory\n",
  66       "\n",
  67       "expanded_name = collections.defaultdict(str)\n",
  68       "for e in labels.find({'expanded': {\"$exists\": True}}):\n",
  69       "    expanded_name[e['label']] = e['expanded']\n",
  70       "    \n",
  71       "label_of = collections.defaultdict(str)\n",
  72       "for l in labels.find({'codes': {\"$exists\": True}}):\n",
  73       "    for c in l['codes']:\n",
  74       "        try:\n",
  75       "            label_of[(l['label'], int(c))] = l['codes'][c]\n",
  76       "        except ValueError: \n",
  77       "            label_of[(l['label'], c)] = l['codes'][c]"
  78      ],
  79      "language": "python",
  80      "metadata": {},
  81      "outputs": [],
  82      "prompt_number": 3
  83     },
  84     {
  85      "cell_type": "markdown",
  86      "metadata": {},
  87      "source": [
  88       "# Getting a single document\n",
  89       "`find_one()` is the basic method for returning a single document from a collection. With no arguments, it just returns the first document it finds (chosen arbitrarily by Mongo)."
  90      ]
  91     },
  92     {
  93      "cell_type": "code",
  94      "collapsed": false,
  95      "input": [
  96       "accidents.find_one()"
  97      ],
  98      "language": "python",
  99      "metadata": {},
 100      "outputs": [
 101       {
 102        "metadata": {},
 103        "output_type": "pyout",
 104        "prompt_number": 4,
 105        "text": [
 106         "{'Longitude': -2.204997,\n",
 107         " 'Police_Force': 6,\n",
 108         " 'Vehicles': [{'Sex_of_Driver': 2,\n",
 109         "   '1st_Point_of_Impact': 1,\n",
 110         "   'Hit_Object_off_Carriageway': 10,\n",
 111         "   'Vehicle_Reference': 1,\n",
 112         "   'Was_Vehicle_Left_Hand_Drive?': 1,\n",
 113         "   'Vehicle_Leaving_Carriageway': 7,\n",
 114         "   'Vehicle_Location-Restricted_Lane': 0,\n",
 115         "   'Vehicle_Type': 9,\n",
 116         "   'Hit_Object_in_Carriageway': 0,\n",
 117         "   'Junction_Location': 0,\n",
 118         "   'Age_Band_of_Driver': 5,\n",
 119         "   'Driver_Home_Area_Type': 1,\n",
 120         "   'Journey_Purpose_of_Driver': 6,\n",
 121         "   'Propulsion_Code': 1,\n",
 122         "   'Engine_Capacity_(CC)': 1596,\n",
 123         "   'Vehicle_Manoeuvre': 16,\n",
 124         "   'Age_of_Vehicle': 6,\n",
 125         "   'Towing_and_Articulation': 0,\n",
 126         "   'Skidding_and_Overturning': 1,\n",
 127         "   'Driver_IMD_Decile': 3}],\n",
 128         " 'Date': '01/01/2012',\n",
 129         " 'Road_Surface_Conditions': 2,\n",
 130         " 'loc': {'coordinates': [-2.204997, 53.588335], 'type': 'Point'},\n",
 131         " '1st_Road_Class': 5,\n",
 132         " 'Road_Type': 6,\n",
 133         " 'Location_Easting_OSGR': 386528,\n",
 134         " 'Number_of_Casualties': 2,\n",
 135         " 'Datetime': datetime.datetime(2012, 1, 1, 0, 4),\n",
 136         " '1st_Road_Number': 0,\n",
 137         " '2nd_Road_Number': 0,\n",
 138         " 'Weather_Conditions': 2,\n",
 139         " 'Location_Northing_OSGR': 410215,\n",
 140         " 'Pedestrian_Crossing-Human_Control': 0,\n",
 141         " 'Casualties': [{'Casualty_Severity': 3,\n",
 142         "   'Pedestrian_Location': 0,\n",
 143         "   'Vehicle_Reference': 1,\n",
 144         "   'Pedestrian_Road_Maintenance_Worker': 0,\n",
 145         "   'Sex_of_Casualty': 2,\n",
 146         "   'Age_Band_of_Casualty': 5,\n",
 147         "   'Car_Passenger': 0,\n",
 148         "   'Casualty_Type': 9,\n",
 149         "   'Casualty_Home_Area_Type': 1,\n",
 150         "   'Casualty_Reference': 1,\n",
 151         "   'Pedestrian_Movement': 0,\n",
 152         "   'Casualty_Class': 1,\n",
 153         "   'Bus_or_Coach_Passenger': 0},\n",
 154         "  {'Casualty_Severity': 3,\n",
 155         "   'Pedestrian_Location': 0,\n",
 156         "   'Vehicle_Reference': 1,\n",
 157         "   'Pedestrian_Road_Maintenance_Worker': 0,\n",
 158         "   'Sex_of_Casualty': 2,\n",
 159         "   'Age_Band_of_Casualty': 7,\n",
 160         "   'Car_Passenger': 2,\n",
 161         "   'Casualty_Type': 9,\n",
 162         "   'Casualty_Home_Area_Type': -1,\n",
 163         "   'Casualty_Reference': 2,\n",
 164         "   'Pedestrian_Movement': 0,\n",
 165         "   'Casualty_Class': 2,\n",
 166         "   'Bus_or_Coach_Passenger': 0}],\n",
 167         " 'LSOA_of_Accident_Location': 'E01005507',\n",
 168         " 'Accident_Index': '201206P055810',\n",
 169         " 'Local_Authority_(District)': 106,\n",
 170         " 'Junction_Detail': 0,\n",
 171         " 'Accident_Severity': 3,\n",
 172         " 'Speed_limit': 30,\n",
 173         " 'Number_of_Vehicles': 1,\n",
 174         " '_id': ObjectId('52a9c97e92c4e16686d1be6b'),\n",
 175         " 'Local_Authority_(Highway)': 'E08000005',\n",
 176         " 'Latitude': 53.588335,\n",
 177         " 'Carriageway_Hazards': 0,\n",
 178         " 'Time': '00:04',\n",
 179         " 'Day_of_Week': 1,\n",
 180         " 'Did_Police_Officer_Attend_Scene_of_Accident': 1,\n",
 181         " 'Junction_Control': -1,\n",
 182         " 'Pedestrian_Crossing-Physical_Facilities': 0,\n",
 183         " 'Special_Conditions_at_Site': 0,\n",
 184         " 'Urban_or_Rural_Area': 1,\n",
 185         " 'Light_Conditions': 4,\n",
 186         " '2nd_Road_Class': -1}"
 187        ]
 188       }
 189      ],
 190      "prompt_number": 4
 191     },
 192     {
 193      "cell_type": "markdown",
 194      "metadata": {},
 195      "source": [
 196       "That's quite a bit of data, and it's not that useful to select an arbitrary accident report. Let's pull out an accident that happened in a 70 mph zone. We do that by giving a document that must match one in the database before it's returned."
 197      ]
 198     },
 199     {
 200      "cell_type": "code",
 201      "collapsed": false,
 202      "input": [
 203       "accidents.find_one({'Speed_limit': 30})"
 204      ],
 205      "language": "python",
 206      "metadata": {},
 207      "outputs": [
 208       {
 209        "metadata": {},
 210        "output_type": "pyout",
 211        "prompt_number": 5,
 212        "text": [
 213         "{'Longitude': -2.204997,\n",
 214         " 'Police_Force': 6,\n",
 215         " 'Vehicles': [{'Sex_of_Driver': 2,\n",
 216         "   '1st_Point_of_Impact': 1,\n",
 217         "   'Hit_Object_off_Carriageway': 10,\n",
 218         "   'Vehicle_Reference': 1,\n",
 219         "   'Was_Vehicle_Left_Hand_Drive?': 1,\n",
 220         "   'Vehicle_Leaving_Carriageway': 7,\n",
 221         "   'Vehicle_Location-Restricted_Lane': 0,\n",
 222         "   'Vehicle_Type': 9,\n",
 223         "   'Hit_Object_in_Carriageway': 0,\n",
 224         "   'Junction_Location': 0,\n",
 225         "   'Age_Band_of_Driver': 5,\n",
 226         "   'Driver_Home_Area_Type': 1,\n",
 227         "   'Journey_Purpose_of_Driver': 6,\n",
 228         "   'Propulsion_Code': 1,\n",
 229         "   'Engine_Capacity_(CC)': 1596,\n",
 230         "   'Vehicle_Manoeuvre': 16,\n",
 231         "   'Age_of_Vehicle': 6,\n",
 232         "   'Towing_and_Articulation': 0,\n",
 233         "   'Skidding_and_Overturning': 1,\n",
 234         "   'Driver_IMD_Decile': 3}],\n",
 235         " 'Date': '01/01/2012',\n",
 236         " 'Road_Surface_Conditions': 2,\n",
 237         " 'loc': {'coordinates': [-2.204997, 53.588335], 'type': 'Point'},\n",
 238         " '1st_Road_Class': 5,\n",
 239         " 'Road_Type': 6,\n",
 240         " 'Location_Easting_OSGR': 386528,\n",
 241         " 'Number_of_Casualties': 2,\n",
 242         " 'Datetime': datetime.datetime(2012, 1, 1, 0, 4),\n",
 243         " '1st_Road_Number': 0,\n",
 244         " '2nd_Road_Number': 0,\n",
 245         " 'Weather_Conditions': 2,\n",
 246         " 'Location_Northing_OSGR': 410215,\n",
 247         " 'Pedestrian_Crossing-Human_Control': 0,\n",
 248         " 'Casualties': [{'Casualty_Severity': 3,\n",
 249         "   'Pedestrian_Location': 0,\n",
 250         "   'Vehicle_Reference': 1,\n",
 251         "   'Pedestrian_Road_Maintenance_Worker': 0,\n",
 252         "   'Sex_of_Casualty': 2,\n",
 253         "   'Age_Band_of_Casualty': 5,\n",
 254         "   'Car_Passenger': 0,\n",
 255         "   'Casualty_Type': 9,\n",
 256         "   'Casualty_Home_Area_Type': 1,\n",
 257         "   'Casualty_Reference': 1,\n",
 258         "   'Pedestrian_Movement': 0,\n",
 259         "   'Casualty_Class': 1,\n",
 260         "   'Bus_or_Coach_Passenger': 0},\n",
 261         "  {'Casualty_Severity': 3,\n",
 262         "   'Pedestrian_Location': 0,\n",
 263         "   'Vehicle_Reference': 1,\n",
 264         "   'Pedestrian_Road_Maintenance_Worker': 0,\n",
 265         "   'Sex_of_Casualty': 2,\n",
 266         "   'Age_Band_of_Casualty': 7,\n",
 267         "   'Car_Passenger': 2,\n",
 268         "   'Casualty_Type': 9,\n",
 269         "   'Casualty_Home_Area_Type': -1,\n",
 270         "   'Casualty_Reference': 2,\n",
 271         "   'Pedestrian_Movement': 0,\n",
 272         "   'Casualty_Class': 2,\n",
 273         "   'Bus_or_Coach_Passenger': 0}],\n",
 274         " 'LSOA_of_Accident_Location': 'E01005507',\n",
 275         " 'Accident_Index': '201206P055810',\n",
 276         " 'Local_Authority_(District)': 106,\n",
 277         " 'Junction_Detail': 0,\n",
 278         " 'Accident_Severity': 3,\n",
 279         " 'Speed_limit': 30,\n",
 280         " 'Number_of_Vehicles': 1,\n",
 281         " '_id': ObjectId('52a9c97e92c4e16686d1be6b'),\n",
 282         " 'Local_Authority_(Highway)': 'E08000005',\n",
 283         " 'Latitude': 53.588335,\n",
 284         " 'Carriageway_Hazards': 0,\n",
 285         " 'Time': '00:04',\n",
 286         " 'Day_of_Week': 1,\n",
 287         " 'Did_Police_Officer_Attend_Scene_of_Accident': 1,\n",
 288         " 'Junction_Control': -1,\n",
 289         " 'Pedestrian_Crossing-Physical_Facilities': 0,\n",
 290         " 'Special_Conditions_at_Site': 0,\n",
 291         " 'Urban_or_Rural_Area': 1,\n",
 292         " 'Light_Conditions': 4,\n",
 293         " '2nd_Road_Class': -1}"
 294        ]
 295       }
 296      ],
 297      "prompt_number": 5
 298     },
 299     {
 300      "cell_type": "markdown",
 301      "metadata": {},
 302      "source": [
 303       "If we give more than one key-value pair in the query document, the returned document must match all of them (and AND). For instance, to find an accident at 30mph that involved two vehicles and one casuality, we specify that information in the query document:"
 304      ]
 305     },
 306     {
 307      "cell_type": "code",
 308      "collapsed": false,
 309      "input": [
 310       "accidents.find_one({'Speed_limit': 30, 'Number_of_Casualties': 1, 'Number_of_Vehicles': 2})"
 311      ],
 312      "language": "python",
 313      "metadata": {},
 314      "outputs": [
 315       {
 316        "metadata": {},
 317        "output_type": "pyout",
 318        "prompt_number": 6,
 319        "text": [
 320         "{'Longitude': -1.731617,\n",
 321         " 'Police_Force': 13,\n",
 322         " 'Vehicles': [{'Sex_of_Driver': 2,\n",
 323         "   '1st_Point_of_Impact': 1,\n",
 324         "   'Hit_Object_off_Carriageway': 0,\n",
 325         "   'Vehicle_Reference': 1,\n",
 326         "   'Was_Vehicle_Left_Hand_Drive?': 1,\n",
 327         "   'Vehicle_Leaving_Carriageway': 0,\n",
 328         "   'Vehicle_Location-Restricted_Lane': 0,\n",
 329         "   'Vehicle_Type': 9,\n",
 330         "   'Hit_Object_in_Carriageway': 0,\n",
 331         "   'Junction_Location': 1,\n",
 332         "   'Age_Band_of_Driver': 5,\n",
 333         "   'Driver_Home_Area_Type': 1,\n",
 334         "   'Journey_Purpose_of_Driver': 6,\n",
 335         "   'Propulsion_Code': 1,\n",
 336         "   'Engine_Capacity_(CC)': 1587,\n",
 337         "   'Vehicle_Manoeuvre': 5,\n",
 338         "   'Age_of_Vehicle': 10,\n",
 339         "   'Towing_and_Articulation': 0,\n",
 340         "   'Skidding_and_Overturning': 0,\n",
 341         "   'Driver_IMD_Decile': 1},\n",
 342         "  {'Sex_of_Driver': 3,\n",
 343         "   '1st_Point_of_Impact': 2,\n",
 344         "   'Hit_Object_off_Carriageway': 0,\n",
 345         "   'Vehicle_Reference': 2,\n",
 346         "   'Was_Vehicle_Left_Hand_Drive?': 1,\n",
 347         "   'Vehicle_Leaving_Carriageway': 0,\n",
 348         "   'Vehicle_Location-Restricted_Lane': 0,\n",
 349         "   'Vehicle_Type': 19,\n",
 350         "   'Hit_Object_in_Carriageway': 0,\n",
 351         "   'Junction_Location': 1,\n",
 352         "   'Age_Band_of_Driver': -1,\n",
 353         "   'Driver_Home_Area_Type': -1,\n",
 354         "   'Journey_Purpose_of_Driver': 6,\n",
 355         "   'Propulsion_Code': 2,\n",
 356         "   'Engine_Capacity_(CC)': 2402,\n",
 357         "   'Vehicle_Manoeuvre': 2,\n",
 358         "   'Age_of_Vehicle': 7,\n",
 359         "   'Towing_and_Articulation': 0,\n",
 360         "   'Skidding_and_Overturning': 0,\n",
 361         "   'Driver_IMD_Decile': -1}],\n",
 362         " 'Date': '01/01/2012',\n",
 363         " 'Road_Surface_Conditions': 2,\n",
 364         " 'loc': {'coordinates': [-1.731617, 53.810716], 'type': 'Point'},\n",
 365         " '1st_Road_Class': 6,\n",
 366         " 'Road_Type': 6,\n",
 367         " 'Location_Easting_OSGR': 417770,\n",
 368         " 'Number_of_Casualties': 1,\n",
 369         " 'Datetime': datetime.datetime(2012, 1, 1, 0, 15),\n",
 370         " '1st_Road_Number': 0,\n",
 371         " '2nd_Road_Number': -1,\n",
 372         " 'Weather_Conditions': 9,\n",
 373         " 'Location_Northing_OSGR': 434970,\n",
 374         " 'Pedestrian_Crossing-Human_Control': 0,\n",
 375         " 'Casualties': [{'Casualty_Severity': 3,\n",
 376         "   'Pedestrian_Location': 0,\n",
 377         "   'Vehicle_Reference': 1,\n",
 378         "   'Pedestrian_Road_Maintenance_Worker': 0,\n",
 379         "   'Sex_of_Casualty': 2,\n",
 380         "   'Age_Band_of_Casualty': 5,\n",
 381         "   'Car_Passenger': 0,\n",
 382         "   'Casualty_Type': 9,\n",
 383         "   'Casualty_Home_Area_Type': 1,\n",
 384         "   'Casualty_Reference': 1,\n",
 385         "   'Pedestrian_Movement': 0,\n",
 386         "   'Casualty_Class': 1,\n",
 387         "   'Bus_or_Coach_Passenger': 0}],\n",
 388         " 'LSOA_of_Accident_Location': 'E01010828',\n",
 389         " 'Accident_Index': 2012130002806,\n",
 390         " 'Local_Authority_(District)': 200,\n",
 391         " 'Junction_Detail': 8,\n",
 392         " 'Accident_Severity': 3,\n",
 393         " 'Speed_limit': 30,\n",
 394         " 'Number_of_Vehicles': 2,\n",
 395         " '_id': ObjectId('52a9c97f92c4e16686d1e68e'),\n",
 396         " 'Local_Authority_(Highway)': 'E08000032',\n",
 397         " 'Latitude': 53.810716,\n",
 398         " 'Carriageway_Hazards': 0,\n",
 399         " 'Time': '00:15',\n",
 400         " 'Day_of_Week': 1,\n",
 401         " 'Did_Police_Officer_Attend_Scene_of_Accident': 2,\n",
 402         " 'Junction_Control': 4,\n",
 403         " 'Pedestrian_Crossing-Physical_Facilities': 0,\n",
 404         " 'Special_Conditions_at_Site': 0,\n",
 405         " 'Urban_or_Rural_Area': 1,\n",
 406         " 'Light_Conditions': 7,\n",
 407         " '2nd_Road_Class': -1}"
 408        ]
 409       }
 410      ],
 411      "prompt_number": 6
 412     },
 413     {
 414      "cell_type": "markdown",
 415      "metadata": {},
 416      "source": [
 417       "We can also limit the key-value pairs returned by specifying the second argument to `find_one()`. If we give it a list of key names, only those key-value pairs are returned."
 418      ]
 419     },
 420     {
 421      "cell_type": "code",
 422      "collapsed": false,
 423      "input": [
 424       "accidents.find_one({'Speed_limit': 30, 'Number_of_Casualties': 1, 'Number_of_Vehicles': 2}, \n",
 425       "                   ['Accident_Index', 'Speed_limit', 'Number_of_Casualties', 'Number_of_Vehicles'])"
 426      ],
 427      "language": "python",
 428      "metadata": {},
 429      "outputs": [
 430       {
 431        "metadata": {},
 432        "output_type": "pyout",
 433        "prompt_number": 7,
 434        "text": [
 435         "{'Number_of_Vehicles': 2,\n",
 436         " '_id': ObjectId('52a9c97f92c4e16686d1e68e'),\n",
 437         " 'Number_of_Casualties': 1,\n",
 438         " 'Accident_Index': 2012130002806,\n",
 439         " 'Speed_limit': 30}"
 440        ]
 441       }
 442      ],
 443      "prompt_number": 7
 444     },
 445     {
 446      "cell_type": "markdown",
 447      "metadata": {},
 448      "source": [
 449       "You'll notice that Mongo returns the document `_id`. It always does that unless we specifically ask it not to. \n",
 450       "\n",
 451       "The 'list of keys' notation is a shorthand for the full specification for which keys to return. Mongo expects a dict of keys to include or exclude. If the keys in that projection specification have a value `1`, the key is included; if they have a value `0`, they're excludeded. You can only include or exclude keys in one projection, with the exception of the `_id` key. \n",
 452       "\n",
 453       "For instance, the previous query could be specified as:"
 454      ]
 455     },
 456     {
 457      "cell_type": "code",
 458      "collapsed": false,
 459      "input": [
 460       "accidents.find_one({'Speed_limit': 30, 'Number_of_Casualties': 1, 'Number_of_Vehicles': 2}, \n",
 461       "                   {'Accident_Index': 1, 'Speed_limit': 1, \n",
 462       "                    'Number_of_Casualties': 1, 'Number_of_Vehicles': 1})"
 463      ],
 464      "language": "python",
 465      "metadata": {},
 466      "outputs": [
 467       {
 468        "metadata": {},
 469        "output_type": "pyout",
 470        "prompt_number": 8,
 471        "text": [
 472         "{'Number_of_Vehicles': 2,\n",
 473         " '_id': ObjectId('52a9c97f92c4e16686d1e68e'),\n",
 474         " 'Number_of_Casualties': 1,\n",
 475         " 'Accident_Index': 2012130002806,\n",
 476         " 'Speed_limit': 30}"
 477        ]
 478       }
 479      ],
 480      "prompt_number": 8
 481     },
 482     {
 483      "cell_type": "markdown",
 484      "metadata": {},
 485      "source": [
 486       "or, to exclude the `_id`:"
 487      ]
 488     },
 489     {
 490      "cell_type": "code",
 491      "collapsed": false,
 492      "input": [
 493       "accidents.find_one({'Speed_limit': 30, 'Number_of_Casualties': 1, 'Number_of_Vehicles': 2}, \n",
 494       "                   {'_id': 0, \n",
 495       "                    'Accident_Index': 1, 'Speed_limit': 1, \n",
 496       "                    'Number_of_Casualties': 1, 'Number_of_Vehicles': 1})"
 497      ],
 498      "language": "python",
 499      "metadata": {},
 500      "outputs": [
 501       {
 502        "metadata": {},
 503        "output_type": "pyout",
 504        "prompt_number": 9,
 505        "text": [
 506         "{'Number_of_Vehicles': 2,\n",
 507         " 'Number_of_Casualties': 1,\n",
 508         " 'Accident_Index': 2012130002806,\n",
 509         " 'Speed_limit': 30}"
 510        ]
 511       }
 512      ],
 513      "prompt_number": 9
 514     },
 515     {
 516      "cell_type": "markdown",
 517      "metadata": {},
 518      "source": [
 519       "To return all the key-value pairs _except_ for the few we specify:"
 520      ]
 521     },
 522     {
 523      "cell_type": "code",
 524      "collapsed": false,
 525      "input": [
 526       "accidents.find_one({'Speed_limit': 30, 'Number_of_Casualties': 1, 'Number_of_Vehicles': 2}, \n",
 527       "                   {'Casualies': 0, 'Vehicles': 0})"
 528      ],
 529      "language": "python",
 530      "metadata": {},
 531      "outputs": [
 532       {
 533        "metadata": {},
 534        "output_type": "pyout",
 535        "prompt_number": 10,
 536        "text": [
 537         "{'Longitude': -1.731617,\n",
 538         " 'Police_Force': 13,\n",
 539         " 'Day_of_Week': 1,\n",
 540         " 'Date': '01/01/2012',\n",
 541         " 'Road_Surface_Conditions': 2,\n",
 542         " 'Light_Conditions': 7,\n",
 543         " '1st_Road_Class': 6,\n",
 544         " 'Road_Type': 6,\n",
 545         " 'Location_Easting_OSGR': 417770,\n",
 546         " 'Number_of_Casualties': 1,\n",
 547         " '1st_Road_Number': 0,\n",
 548         " '2nd_Road_Number': -1,\n",
 549         " 'Weather_Conditions': 9,\n",
 550         " 'loc': {'coordinates': [-1.731617, 53.810716], 'type': 'Point'},\n",
 551         " 'Location_Northing_OSGR': 434970,\n",
 552         " 'Pedestrian_Crossing-Human_Control': 0,\n",
 553         " 'Casualties': [{'Casualty_Severity': 3,\n",
 554         "   'Pedestrian_Location': 0,\n",
 555         "   'Vehicle_Reference': 1,\n",
 556         "   'Pedestrian_Road_Maintenance_Worker': 0,\n",
 557         "   'Sex_of_Casualty': 2,\n",
 558         "   'Age_Band_of_Casualty': 5,\n",
 559         "   'Car_Passenger': 0,\n",
 560         "   'Casualty_Type': 9,\n",
 561         "   'Casualty_Home_Area_Type': 1,\n",
 562         "   'Casualty_Reference': 1,\n",
 563         "   'Pedestrian_Movement': 0,\n",
 564         "   'Casualty_Class': 1,\n",
 565         "   'Bus_or_Coach_Passenger': 0}],\n",
 566         " 'LSOA_of_Accident_Location': 'E01010828',\n",
 567         " 'Accident_Index': 2012130002806,\n",
 568         " 'Local_Authority_(District)': 200,\n",
 569         " 'Junction_Detail': 8,\n",
 570         " 'Accident_Severity': 3,\n",
 571         " 'Speed_limit': 30,\n",
 572         " 'Number_of_Vehicles': 2,\n",
 573         " '_id': ObjectId('52a9c97f92c4e16686d1e68e'),\n",
 574         " 'Local_Authority_(Highway)': 'E08000032',\n",
 575         " 'Carriageway_Hazards': 0,\n",
 576         " 'Time': '00:15',\n",
 577         " '2nd_Road_Class': -1,\n",
 578         " 'Did_Police_Officer_Attend_Scene_of_Accident': 2,\n",
 579         " 'Junction_Control': 4,\n",
 580         " 'Pedestrian_Crossing-Physical_Facilities': 0,\n",
 581         " 'Datetime': datetime.datetime(2012, 1, 1, 0, 15),\n",
 582         " 'Urban_or_Rural_Area': 1,\n",
 583         " 'Latitude': 53.810716,\n",
 584         " 'Special_Conditions_at_Site': 0}"
 585        ]
 586       }
 587      ],
 588      "prompt_number": 10
 589     },
 590     {
 591      "cell_type": "markdown",
 592      "metadata": {},
 593      "source": [
 594       "## Finding multiple documents\n",
 595       "If we want to find more than one document at a time, we use the imaginatively named `find()` function."
 596      ]
 597     },
 598     {
 599      "cell_type": "code",
 600      "collapsed": false,
 601      "input": [
 602       "accidents.find({'Speed_limit': 30, 'Number_of_Casualties': 1, 'Number_of_Vehicles': 2})"
 603      ],
 604      "language": "python",
 605      "metadata": {},
 606      "outputs": [
 607       {
 608        "metadata": {},
 609        "output_type": "pyout",
 610        "prompt_number": 11,
 611        "text": [
 612         "<pymongo.cursor.Cursor at 0x7f4bac98f4a8>"
 613        ]
 614       }
 615      ],
 616      "prompt_number": 11
 617     },
 618     {
 619      "cell_type": "markdown",
 620      "metadata": {},
 621      "source": [
 622       "This returns a PyMongo `cursor` object, which acts like an iterable in most cases. \n",
 623       "\n",
 624       "One useful thing the cursor can do is tell us how many documents match."
 625      ]
 626     },
 627     {
 628      "cell_type": "code",
 629      "collapsed": false,
 630      "input": [
 631       "accidents.find({'Speed_limit': 30, 'Number_of_Casualties': 1, 'Number_of_Vehicles': 2}).count()"
 632      ],
 633      "language": "python",
 634      "metadata": {},
 635      "outputs": [
 636       {
 637        "metadata": {},
 638        "output_type": "pyout",
 639        "prompt_number": 12,
 640        "text": [
 641         "46715"
 642        ]
 643       }
 644      ],
 645      "prompt_number": 12
 646     },
 647     {
 648      "cell_type": "markdown",
 649      "metadata": {},
 650      "source": [
 651       "While we're exploring, we can limit the number of documents returned with the `limit` keyword parameter. Again, which documents are returned is arbitrary."
 652      ]
 653     },
 654     {
 655      "cell_type": "code",
 656      "collapsed": false,
 657      "input": [
 658       "len(list(accidents.find({'Speed_limit': 30, 'Number_of_Casualties': 1, 'Number_of_Vehicles': 2}, limit=10)))"
 659      ],
 660      "language": "python",
 661      "metadata": {},
 662      "outputs": [
 663       {
 664        "metadata": {},
 665        "output_type": "pyout",
 666        "prompt_number": 13,
 667        "text": [
 668         "10"
 669        ]
 670       }
 671      ],
 672      "prompt_number": 13
 673     },
 674     {
 675      "cell_type": "markdown",
 676      "metadata": {},
 677      "source": [
 678       "We can iterate over the cursor in Python just as we'd iterate over any iterable, and pull values from the document just as we would from any dict:"
 679      ]
 680     },
 681     {
 682      "cell_type": "code",
 683      "collapsed": false,
 684      "input": [
 685       "for a in accidents.find({'Speed_limit': 30, 'Number_of_Casualties': 3, 'Number_of_Vehicles': 2}, limit=10):\n",
 686       "    print(a['Accident_Index'], a['Accident_Severity'])"
 687      ],
 688      "language": "python",
 689      "metadata": {},
 690      "outputs": [
 691       {
 692        "output_type": "stream",
 693        "stream": "stdout",
 694        "text": [
 695         "201201KD50022 3\n",
 696         "2012620213941 3\n",
 697         "201243N035012 2\n",
 698         "201201XD80013 2\n",
 699         "2012130000979 2\n",
 700         "2012540017412 2\n",
 701         "201206N056171 3\n",
 702         "201242I000601 3\n",
 703         "2012160C00741 3\n",
 704         "2012030001017 3\n"
 705        ]
 706       }
 707      ],
 708      "prompt_number": 14
 709     },
 710     {
 711      "cell_type": "markdown",
 712      "metadata": {},
 713      "source": [
 714       "And we can use the `label_of` dict to show what the cryptic codes mean."
 715      ]
 716     },
 717     {
 718      "cell_type": "code",
 719      "collapsed": false,
 720      "input": [
 721       "for a in accidents.find({'Speed_limit': 30, 'Number_of_Casualties': 3, 'Number_of_Vehicles': 2}, limit=10):\n",
 722       "    print(a['Accident_Index'], label_of[('Accident_Severity', a['Accident_Severity'])])"
 723      ],
 724      "language": "python",
 725      "metadata": {},
 726      "outputs": [
 727       {
 728        "output_type": "stream",
 729        "stream": "stdout",
 730        "text": [
 731         "201201KD50022 Slight\n",
 732         "2012620213941 Slight\n",
 733         "201243N035012 Serious\n",
 734         "201201XD80013 Serious\n",
 735         "2012130000979 Serious\n",
 736         "2012540017412 Serious\n",
 737         "201206N056171 Slight\n",
 738         "201242I000601 Slight\n",
 739         "2012160C00741 Slight\n",
 740         "2012030001017 Slight\n"
 741        ]
 742       }
 743      ],
 744      "prompt_number": 15
 745     },
 746     {
 747      "cell_type": "markdown",
 748      "metadata": {},
 749      "source": [
 750       "We can now start to investigate the data. What are the proportions of accident severities in 30mph zones?\n",
 751       "\n",
 752       "First, make sure we have the labels for the severities in the correct order, so we can use that order in the plots."
 753      ]
 754     },
 755     {
 756      "cell_type": "code",
 757      "collapsed": false,
 758      "input": [
 759       "ordered_severity_labels = [label_of[k] for k in \n",
 760       "                  sorted([(l, c) for l, c in label_of.keys() if l == 'Accident_Severity'], \n",
 761       "                         key = lambda ln: ln[1])]"
 762      ],
 763      "language": "python",
 764      "metadata": {},
 765      "outputs": [],
 766      "prompt_number": 16
 767     },
 768     {
 769      "cell_type": "code",
 770      "collapsed": false,
 771      "input": [
 772       "severities_30 = pd.Series(collections.Counter(label_of[('Accident_Severity', a['Accident_Severity'])] \n",
 773       "                                              for a in accidents.find({'Speed_limit': 30}, ['Accident_Severity'])))\n",
 774       "severities_30.reindex(ordered_severity_labels)\n",
 775       "severities_30"
 776      ],
 777      "language": "python",
 778      "metadata": {},
 779      "outputs": [
 780       {
 781        "metadata": {},
 782        "output_type": "pyout",
 783        "prompt_number": 17,
 784        "text": [
 785         "Fatal        582\n",
 786         "Serious    12686\n",
 787         "Slight     81727\n",
 788         "dtype: int64"
 789        ]
 790       }
 791      ],
 792      "prompt_number": 17
 793     },
 794     {
 795      "cell_type": "code",
 796      "collapsed": false,
 797      "input": [
 798       "severities_30.plot(kind='bar')"
 799      ],
 800      "language": "python",
 801      "metadata": {},
 802      "outputs": [
 803       {
 804        "metadata": {},
 805        "output_type": "pyout",
 806        "prompt_number": 18,
 807        "text": [
 808         "<matplotlib.axes.AxesSubplot at 0x7f4bac98f7f0>"
 809        ]
 810       },
 811       {
 812        "metadata": {},
 813        "output_type": "display_data",
 814        "png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAEaCAYAAAD5fVeOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
 815        "text": [
 816         "<matplotlib.figure.Figure at 0x7f4bac993128>"
 817        ]
 818       }
 819      ],
 820      "prompt_number": 18
 821     },
 822     {
 823      "cell_type": "markdown",
 824      "metadata": {},
 825      "source": [
 826       "How about in 70mph zones?"
 827      ]
 828     },
 829     {
 830      "cell_type": "code",
 831      "collapsed": false,
 832      "input": [
 833       "severities_70 = pd.Series(collections.Counter(label_of[('Accident_Severity', a['Accident_Severity'])] \n",
 834       "                                              for a in accidents.find({'Speed_limit': 70}, ['Accident_Severity'])))\n",
 835       "severities_70.reindex(ordered_severity_labels)\n",
 836       "severities_70"
 837      ],
 838      "language": "python",
 839      "metadata": {},
 840      "outputs": [
 841       {
 842        "metadata": {},
 843        "output_type": "pyout",
 844        "prompt_number": 19,
 845        "text": [
 846         "Fatal       198\n",
 847         "Serious    1157\n",
 848         "Slight     8667\n",
 849         "dtype: int64"
 850        ]
 851       }
 852      ],
 853      "prompt_number": 19
 854     },
 855     {
 856      "cell_type": "code",
 857      "collapsed": false,
 858      "input": [
 859       "severities_70.plot(kind='bar')"
 860      ],
 861      "language": "python",
 862      "metadata": {},
 863      "outputs": [
 864       {
 865        "metadata": {},
 866        "output_type": "pyout",
 867        "prompt_number": 20,
 868        "text": [
 869         "<matplotlib.axes.AxesSubplot at 0x7f4bac944940>"
 870        ]
 871       },
 872       {
 873        "metadata": {},
 874        "output_type": "display_data",
 875        "png": "iVBORw0KGgoAAAANSUhEUgAAAXkAAAEaCAYAAAD0YyfJAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
 876        "text": [
 877         "<matplotlib.figure.Figure at 0x7f4bac91f278>"
 878        ]
 879       }
 880      ],
 881      "prompt_number": 20
 882     },
 883     {
 884      "cell_type": "markdown",
 885      "metadata": {},
 886      "source": [
 887       "Let's compare accident severities across speed limits"
 888      ]
 889     },
 890     {
 891      "cell_type": "code",
 892      "collapsed": false,
 893      "input": [
 894       "severity_by_speed_dict = collections.Counter((a['Speed_limit'], a['Accident_Severity']) \n",
 895       "                                               for a in accidents.find({}, ['Speed_limit', 'Accident_Severity']))\n",
 896       "severity_by_speed_dict"
 897      ],
 898      "language": "python",
 899      "metadata": {},
 900      "outputs": [
 901       {
 902        "metadata": {},
 903        "output_type": "pyout",
 904        "prompt_number": 21,
 905        "text": [
 906         "Counter({(30, 3): 81727, (60, 3): 16416, (30, 2): 12686, (40, 3): 9961, (70, 3): 8667, (50, 3): 4356, (60, 2): 4157, (20, 3): 1905, (40, 2): 1798, (70, 2): 1157, (50, 2): 770, (60, 1): 599, (30, 1): 582, (20, 2): 333, (70, 1): 198, (40, 1): 155, (50, 1): 94, (20, 1): 9, (10, 3): 1})"
 907        ]
 908       }
 909      ],
 910      "prompt_number": 21
 911     },
 912     {
 913      "cell_type": "markdown",
 914      "metadata": {},
 915      "source": [
 916       "Capture the severities and speeds and put them in the right order"
 917      ]
 918     },
 919     {
 920      "cell_type": "code",
 921      "collapsed": false,
 922      "input": [
 923       "speeds = sorted(set(k[0] for k in severity_by_speed_dict.keys()))\n",
 924       "severities = sorted(set(k[1] for k in severity_by_speed_dict.keys()))"
 925      ],
 926      "language": "python",
 927      "metadata": {},
 928      "outputs": [],
 929      "prompt_number": 22
 930     },
 931     {
 932      "cell_type": "code",
 933      "collapsed": false,
 934      "input": [
 935       "severity_by_speed = pd.DataFrame(\n",
 936       "        {label_of[('Accident_Severity', sv)]: {sp: severity_by_speed_dict[(sp, sv)]\n",
 937       "              for sp in speeds} \n",
 938       "         for sv in severities})\n",
 939       "severity_by_speed.reindex(columns=ordered_severity_labels)\n",
 940       "severity_by_speed.index.name = 'Speed'\n",
 941       "severity_by_speed.columns.name = 'Severity'\n",
 942       "severity_by_speed"
 943      ],
 944      "language": "python",
 945      "metadata": {},
 946      "outputs": [
 947       {
 948        "html": [
 949         "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
 950         "<table border=\"1\" class=\"dataframe\">\n",
 951         "  <thead>\n",
 952         "    <tr style=\"text-align: right;\">\n",
 953         "      <th>Severity</th>\n",
 954         "      <th>Fatal</th>\n",
 955         "      <th>Serious</th>\n",
 956         "      <th>Slight</th>\n",
 957         "    </tr>\n",
 958         "    <tr>\n",
 959         "      <th>Speed</th>\n",
 960         "      <th></th>\n",
 961         "      <th></th>\n",
 962         "      <th></th>\n",
 963         "    </tr>\n",
 964         "  </thead>\n",
 965         "  <tbody>\n",
 966         "    <tr>\n",
 967         "      <th>10</th>\n",
 968         "      <td>   0</td>\n",
 969         "      <td>     0</td>\n",
 970         "      <td>     1</td>\n",
 971         "    </tr>\n",
 972         "    <tr>\n",
 973         "      <th>20</th>\n",
 974         "      <td>   9</td>\n",
 975         "      <td>   333</td>\n",
 976         "      <td>  1905</td>\n",
 977         "    </tr>\n",
 978         "    <tr>\n",
 979         "      <th>30</th>\n",
 980         "      <td> 582</td>\n",
 981         "      <td> 12686</td>\n",
 982         "      <td> 81727</td>\n",
 983         "    </tr>\n",
 984         "    <tr>\n",
 985         "      <th>40</th>\n",
 986         "      <td> 155</td>\n",
 987         "      <td>  1798</td>\n",
 988         "      <td>  9961</td>\n",
 989         "    </tr>\n",
 990         "    <tr>\n",
 991         "      <th>50</th>\n",
 992         "      <td>  94</td>\n",
 993         "      <td>   770</td>\n",
 994         "      <td>  4356</td>\n",
 995         "    </tr>\n",
 996         "    <tr>\n",
 997         "      <th>60</th>\n",
 998         "      <td> 599</td>\n",
 999         "      <td>  4157</td>\n",
1000         "      <td> 16416</td>\n",
1001         "    </tr>\n",
1002         "    <tr>\n",
1003         "      <th>70</th>\n",
1004         "      <td> 198</td>\n",
1005         "      <td>  1157</td>\n",
1006         "      <td>  8667</td>\n",
1007         "    </tr>\n",
1008         "  </tbody>\n",
1009         "</table>\n",
1010         "<p>7 rows \u00d7 3 columns</p>\n",
1011         "</div>"
1012        ],
1013        "metadata": {},
1014        "output_type": "pyout",
1015        "prompt_number": 23,
1016        "text": [
1017         "Severity  Fatal  Serious  Slight\n",
1018         "Speed                           \n",
1019         "10            0        0       1\n",
1020         "20            9      333    1905\n",
1021         "30          582    12686   81727\n",
1022         "40          155     1798    9961\n",
1023         "50           94      770    4356\n",
1024         "60          599     4157   16416\n",
1025         "70          198     1157    8667\n",
1026         "\n",
1027         "[7 rows x 3 columns]"
1028        ]
1029       }
1030      ],
1031      "prompt_number": 23
1032     },
1033     {
1034      "cell_type": "markdown",
1035      "metadata": {},
1036      "source": [
1037       "Plot this data to see what we've got."
1038      ]
1039     },
1040     {
1041      "cell_type": "code",
1042      "collapsed": false,
1043      "input": [
1044       "severity_by_speed.plot(kind='bar')"
1045      ],
1046      "language": "python",
1047      "metadata": {},
1048      "outputs": [
1049       {
1050        "metadata": {},
1051        "output_type": "pyout",
1052        "prompt_number": 24,
1053        "text": [
1054         "<matplotlib.axes.AxesSubplot at 0x7f4bac8352e8>"
1055        ]
1056       },
1057       {
1058        "metadata": {},
1059        "output_type": "display_data",
1060        "png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAERCAYAAACTuqdNAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
1061        "text": [
1062         "<matplotlib.figure.Figure at 0x7f4bac835c88>"
1063        ]
1064       }
1065      ],
1066      "prompt_number": 24
1067     },
1068     {
1069      "cell_type": "markdown",
1070      "metadata": {},
1071      "source": [
1072       "It's difficult to see the detail because there are so many accidents at 30mph. How many accidents are there at each speed?"
1073      ]
1074     },
1075     {
1076      "cell_type": "code",
1077      "collapsed": false,
1078      "input": [
1079       "# What proportion of the accidents occur at each speed?\n",
1080       "severity_by_speed.sum(axis='columns') / severity_by_speed.sum(axis='columns').sum()"
1081      ],
1082      "language": "python",
1083      "metadata": {},
1084      "outputs": [
1085       {
1086        "metadata": {},
1087        "output_type": "pyout",
1088        "prompt_number": 25,
1089        "text": [
1090         "Speed\n",
1091         "10       0.000007\n",
1092         "20       0.015436\n",
1093         "30       0.652568\n",
1094         "40       0.081843\n",
1095         "50       0.035859\n",
1096         "60       0.145441\n",
1097         "70       0.068846\n",
1098         "dtype: float64"
1099        ]
1100       }
1101      ],
1102      "prompt_number": 25
1103     },
1104     {
1105      "cell_type": "code",
1106      "collapsed": false,
1107      "input": [
1108       "# scale the accident numbers so the largest value is 1.\n",
1109       "severity_by_speed.sum(axis='columns') / severity_by_speed.sum(axis='columns').max()"
1110      ],
1111      "language": "python",
1112      "metadata": {},
1113      "outputs": [
1114       {
1115        "metadata": {},
1116        "output_type": "pyout",
1117        "prompt_number": 26,
1118        "text": [
1119         "Speed\n",
1120         "10       0.000011\n",
1121         "20       0.023654\n",
1122         "30       1.000000\n",
1123         "40       0.125417\n",
1124         "50       0.054950\n",
1125         "60       0.222875\n",
1126         "70       0.105500\n",
1127         "dtype: float64"
1128        ]
1129       }
1130      ],
1131      "prompt_number": 26
1132     },
1133     {
1134      "cell_type": "code",
1135      "collapsed": false,
1136      "input": [
1137       "severity_by_speed.sum(axis=1).plot(kind='bar')"
1138      ],
1139      "language": "python",
1140      "metadata": {},
1141      "outputs": [
1142       {
1143        "metadata": {},
1144        "output_type": "pyout",
1145        "prompt_number": 27,
1146        "text": [
1147         "<matplotlib.axes.AxesSubplot at 0x7f4bab1a0438>"
1148        ]
1149       },
1150       {
1151        "metadata": {},
1152        "output_type": "display_data",
1153        "png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAERCAYAAACU1LsdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
1154        "text": [
1155         "<matplotlib.figure.Figure at 0x7f4bac80cf98>"
1156        ]
1157       }
1158      ],
1159      "prompt_number": 27
1160     },
1161     {
1162      "cell_type": "markdown",
1163      "metadata": {},
1164      "source": [
1165       "There are over four times as many accidents in 30mph zones as at any speed, and those accidents represent about two thirds of all the accidents that occurred. \n",
1166       "\n",
1167       "Let's normalise each row so that we can see the proportions of accidents at each severity at eash speed.\n",
1168       "\n",
1169       "(And push the legend out beyond the right hand side of the plot so that it doesn't obscure the bars.)"
1170      ]
1171     },
1172     {
1173      "cell_type": "code",
1174      "collapsed": false,
1175      "input": [
1176       "severity_by_speed_norm_row = severity_by_speed.div(severity_by_speed.sum(axis='columns'), axis='index')\n",
1177       "ax = severity_by_speed_norm_row.plot(kind='bar', legend=False)\n",
1178       "ax.legend(loc='center left', bbox_to_anchor=(1, 0.5))"
1179      ],
1180      "language": "python",
1181      "metadata": {},
1182      "outputs": [
1183       {
1184        "metadata": {},
1185        "output_type": "pyout",
1186        "prompt_number": 28,
1187        "text": [
1188         "<matplotlib.legend.Legend at 0x7f4bac7d8cf8>"
1189        ]
1190       },
1191       {
1192        "metadata": {},
1193        "output_type": "display_data",
1194        "png": "iVBORw0KGgoAAAANSUhEUgAAAcwAAAERCAYAAAD2e5d3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niAzE/wfBD5sVYPb3OAAAAABJRU5ErkJggg==\n",
1195        "text": [
1196         "<matplotlib.figure.Figure at 0x7f4bac7d82b0>"
1197        ]
1198       }
1199      ],
1200      "prompt_number": 28
1201     },
1202     {
1203      "cell_type": "markdown",
1204      "metadata": {},
1205      "source": [
1206       "Is there a difference of accident severity with speed? We'll normalise the data down each column, so we can see the proportion of fatal accidents that occur at each speed (and the same for all the severities).\n",
1207       "\n",
1208       "We'll plot the data grouped by severity."
1209      ]
1210     },
1211     {
1212      "cell_type": "code",
1213      "collapsed": false,
1214      "input": [
1215       "severity_by_speed_norm_col = severity_by_speed / severity_by_speed.sum()\n",
1216       "ax = severity_by_speed_norm_col.T.plot(kind='bar')\n",
1217       "ax.legend(loc='center left', bbox_to_anchor=(1, 0.5))"
1218      ],
1219      "language": "python",
1220      "metadata": {},
1221      "outputs": [
1222       {
1223        "metadata": {},
1224        "output_type": "pyout",
1225        "prompt_number": 29,
1226        "text": [
1227         "<matplotlib.legend.Legend at 0x7f4bac83fba8>"
1228        ]
1229       },
1230       {
1231        "metadata": {},
1232        "output_type": "display_data",
1233        "png": "iVBORw0KGgoAAAANSUhEUgAAAa4AAAEpCAYAAADLQMb6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
1234        "text": [
1235         "<matplotlib.figure.Figure at 0x7f4bac93be48>"
1236        ]
1237       }
1238      ],
1239      "prompt_number": 29
1240     },
1241     {
1242      "cell_type": "markdown",
1243      "metadata": {},
1244      "source": [
1245       "Which shows that serious and fatal accidents tend to happen at higher speeds. \n",
1246       "\n",
1247       "Let's now look at how accident the number of casualites and vehicles correlate.\n",
1248       "\n",
1249       "First of all, what's the range of number of casualities and vehicles involved in accidents? (The `distinct` function gives the distinct values for a particular key in a cursor.)"
1250      ]
1251     },
1252     {
1253      "cell_type": "code",
1254      "collapsed": false,
1255      "input": [
1256       "sorted(accidents.find().distinct('Number_of_Casualties'))"
1257      ],
1258      "language": "python",
1259      "metadata": {},
1260      "outputs": [
1261       {
1262        "metadata": {},
1263        "output_type": "pyout",
1264        "prompt_number": 30,
1265        "text": [
1266         "[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 19, 25, 33, 38, 42]"
1267        ]
1268       }
1269      ],
1270      "prompt_number": 30
1271     },
1272     {
1273      "cell_type": "code",
1274      "collapsed": false,
1275      "input": [
1276       "sorted(accidents.find().distinct('Number_of_Vehicles'))"
1277      ],
1278      "language": "python",
1279      "metadata": {},
1280      "outputs": [
1281       {
1282        "metadata": {},
1283        "output_type": "pyout",
1284        "prompt_number": 31,
1285        "text": [
1286         "[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 16, 18]"
1287        ]
1288       }
1289      ],
1290      "prompt_number": 31
1291     },
1292     {
1293      "cell_type": "markdown",
1294      "metadata": {},
1295      "source": [
1296       "Now let's look at the distribution of the numbers."
1297      ]
1298     },
1299     {
1300      "cell_type": "code",
1301      "collapsed": false,
1302      "input": [
1303       "cas_veh_dict = collections.Counter((a['Number_of_Casualties'], a['Number_of_Vehicles']) for a in accidents.find())\n",
1304       "cas_veh_dict"
1305      ],
1306      "language": "python",
1307      "metadata": {},
1308      "outputs": [
1309       {
1310        "metadata": {},
1311        "output_type": "pyout",
1312        "prompt_number": 32,
1313        "text": [
1314         "Counter({(1, 2): 65680, (1, 1): 39217, (2, 2): 15035, (1, 3): 6168, (3, 2): 4186, (2, 1): 3781, (2, 3): 3002, (4, 2): 1440, (3, 3): 1203, (1, 4): 1056, (2, 4): 716, (3, 1): 682, (5, 2): 521, (4, 3): 481, (3, 4): 319, (4, 1): 287, (1, 5): 197, (6, 2): 183, (5, 3): 169, (4, 4): 166, (2, 5): 128, (5, 1): 79, (1, 6): 76, (3, 5): 73, (6, 3): 72, (5, 4): 63, (7, 2): 52, (4, 5): 48, (6, 4): 46, (2, 6): 41, (5, 5): 29, (7, 3): 28, (1, 7): 27, (3, 6): 25, (8, 2): 24, (6, 1): 22, (2, 7): 14, (6, 5): 13, (3, 7): 12, (8, 3): 12, (7, 4): 12, (4, 6): 11, (9, 2): 11, (5, 6): 10, (8, 4): 9, (6, 7): 7, (7, 5): 7, (7, 1): 7, (10, 2): 7, (3, 8): 6, (2, 9): 6, (5, 7): 6, (9, 3): 5, (8, 1): 5, (2, 8): 4, (6, 6): 4, (11, 4): 4, (1, 8): 4, (4, 8): 3, (8, 5): 3, (10, 5): 3, (11, 2): 3, (10, 3): 3, (2, 10): 3, (7, 6): 2, (13, 2): 2, (25, 2): 2, (4, 7): 2, (9, 1): 2, (4, 9): 2, (8, 6): 2, (11, 9): 2, (7, 8): 2, (3, 9): 2, (9, 5): 2, (12, 1): 1, (13, 4): 1, (1, 11): 1, (5, 8): 1, (42, 2): 1, (4, 13): 1, (8, 10): 1, (17, 2): 1, (2, 16): 1, (12, 3): 1, (13, 1): 1, (15, 3): 1, (7, 7): 1, (1, 9): 1, (14, 1): 1, (6, 8): 1, (11, 8): 1, (5, 9): 1, (4, 18): 1, (3, 11): 1, (2, 12): 1, (9, 4): 1, (16, 1): 1, (12, 2): 1, (1, 10): 1, (33, 2): 1, (12, 7): 1, (13, 6): 1, (19, 2): 1, (38, 4): 1, (15, 2): 1, (11, 3): 1, (16, 3): 1})"
1315        ]
1316       }
1317      ],
1318      "prompt_number": 32
1319     },
1320     {
1321      "cell_type": "code",
1322      "collapsed": false,
1323      "input": [
1324       "cas_veh = pd.DataFrame(\n",
1325       "    {'Casualties': [v for c, v in sorted(cas_veh_dict.keys())],\n",
1326       "     'Vehicles': [c for c, v in sorted(cas_veh_dict.keys())],\n",
1327       "     'Count': [cas_veh_dict[(c, v)] for c, v in sorted(cas_veh_dict.keys())]}\n",
1328       "    )\n",
1329       "cas_veh"
1330      ],
1331      "language": "python",
1332      "metadata": {},
1333      "outputs": [
1334       {
1335        "html": [
1336         "<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
1337         "<table border=\"1\" class=\"dataframe\">\n",
1338         "  <thead>\n",
1339         "    <tr style=\"text-align: right;\">\n",
1340         "      <th></th>\n",
1341         "      <th>Casualties</th>\n",
1342         "      <th>Count</th>\n",
1343         "      <th>Vehicles</th>\n",
1344         "    </tr>\n",
1345         "  </thead>\n",
1346         "  <tbody>\n",
1347         "    <tr>\n",
1348         "      <th>0 </th>\n",
1349         "      <td>  1</td>\n",
1350         "      <td> 39217</td>\n",
1351         "      <td> 1</td>\n",
1352         "    </tr>\n",
1353         "    <tr>\n",
1354         "      <th>1 </th>\n",
1355         "      <td>  2</td>\n",
1356         "      <td> 65680</td>\n",
1357         "      <td> 1</td>\n",
1358         "    </tr>\n",
1359         "    <tr>\n",
1360         "      <th>2 </th>\n",
1361         "      <td>  3</td>\n",
1362         "      <td>  6168</td>\n",
1363         "      <td> 1</td>\n",
1364         "    </tr>\n",
1365         "    <tr>\n",
1366         "      <th>3 </th>\n",
1367         "      <td>  4</td>\n",
1368         "      <td>  1056</td>\n",
1369         "      <td> 1</td>\n",
1370         "    </tr>\n",
1371         "    <tr>\n",
1372         "      <th>4 </th>\n",
1373         "      <td>  5</td>\n",
1374         "      <td>   197</td>\n",
1375         "      <td> 1</td>\n",
1376         "    </tr>\n",
1377         "    <tr>\n",
1378         "      <th>5 </th>\n",
1379         "      <td>  6</td>\n",
1380         "      <td>    76</td>\n",
1381         "      <td> 1</td>\n",
1382         "    </tr>\n",
1383         "    <tr>\n",
1384         "      <th>6 </th>\n",
1385         "      <td>  7</td>\n",
1386         "      <td>    27</td>\n",
1387         "      <td> 1</td>\n",
1388         "    </tr>\n",
1389         "    <tr>\n",
1390         "      <th>7 </th>\n",
1391         "      <td>  8</td>\n",
1392         "      <td>     4</td>\n",
1393         "      <td> 1</td>\n",
1394         "    </tr>\n",
1395         "    <tr>\n",
1396         "      <th>8 </th>\n",
1397         "      <td>  9</td>\n",
1398         "      <td>     1</td>\n",
1399         "      <td> 1</td>\n",
1400         "    </tr>\n",
1401         "    <tr>\n",
1402         "      <th>9 </th>\n",
1403         "      <td> 10</td>\n",
1404         "      <td>     1</td>\n",
1405         "      <td> 1</td>\n",
1406         "    </tr>\n",
1407         "    <tr>\n",
1408         "      <th>10</th>\n",
1409         "      <td> 11</td>\n",
1410         "      <td>     1</td>\n",
1411         "      <td> 1</td>\n",
1412         "    </tr>\n",
1413         "    <tr>\n",
1414         "      <th>11</th>\n",
1415         "      <td>  1</td>\n",
1416         "      <td>  3781</td>\n",
1417         "      <td> 2</td>\n",
1418         "    </tr>\n",
1419         "    <tr>\n",
1420         "      <th>12</th>\n",
1421         "      <td>  2</td>\n",
1422         "      <td> 15035</td>\n",
1423         "      <td> 2</td>\n",
1424         "    </tr>\n",
1425         "    <tr>\n",
1426         "      <th>13</th>\n",
1427         "      <td>  3</td>\n",
1428         "      <td>  3002</td>\n",
1429         "      <td> 2</td>\n",
1430         "    </tr>\n",
1431         "    <tr>\n",
1432         "      <th>14</th>\n",
1433         "      <td>  4</td>\n",
1434         "      <td>   716</td>\n",
1435         "      <td> 2</td>\n",
1436         "    </tr>\n",
1437         "    <tr>\n",
1438         "      <th>15</th>\n",
1439         "      <td>  5</td>\n",
1440         "      <td>   128</td>\n",
1441         "      <td> 2</td>\n",
1442         "    </tr>\n",
1443         "    <tr>\n",
1444         "      <th>16</th>\n",
1445         "      <td>  6</td>\n",
1446         "      <td>    41</td>\n",
1447         "      <td> 2</td>\n",
1448         "    </tr>\n",
1449         "    <tr>\n",
1450         "      <th>17</th>\n",
1451         "      <td>  7</td>\n",
1452         "      <td>    14</td>\n",
1453         "      <td> 2</td>\n",
1454         "    </tr>\n",
1455         "    <tr>\n",
1456         "      <th>18</th>\n",
1457         "      <td>  8</td>\n",
1458         "      <td>     4</td>\n",
1459         "      <td> 2</td>\n",
1460         "    </tr>\n",
1461         "    <tr>\n",
1462         "      <th>19</th>\n",
1463         "      <td>  9</td>\n",
1464         "      <td>     6</td>\n",
1465         "      <td> 2</td>\n",
1466         "    </tr>\n",
1467         "    <tr>\n",
1468         "      <th>20</th>\n",
1469         "      <td> 10</td>\n",
1470         "      <td>     3</td>\n",
1471         "      <td> 2</td>\n",
1472         "    </tr>\n",
1473         "    <tr>\n",
1474         "      <th>21</th>\n",
1475         "      <td> 12</td>\n",
1476         "      <td>     1</td>\n",
1477         "      <td> 2</td>\n",
1478         "    </tr>\n",
1479         "    <tr>\n",
1480         "      <th>22</th>\n",
1481         "      <td> 16</td>\n",
1482         "      <td>     1</td>\n",
1483         "      <td> 2</td>\n",
1484         "    </tr>\n",
1485         "    <tr>\n",
1486         "      <th>23</th>\n",
1487         "      <td>  1</td>\n",
1488         "      <td>   682</td>\n",
1489         "      <td> 3</td>\n",
1490         "    </tr>\n",
1491         "    <tr>\n",
1492         "      <th>24</th>\n",
1493         "      <td>  2</td>\n",
1494         "      <td>  4186</td>\n",
1495         "      <td> 3</td>\n",
1496         "    </tr>\n",
1497         "    <tr>\n",
1498         "      <th>25</th>\n",
1499         "      <td>  3</td>\n",
1500         "      <td>  1203</td>\n",
1501         "      <td> 3</td>\n",
1502         "    </tr>\n",
1503         "    <tr>\n",
1504         "      <th>26</th>\n",
1505         "      <td>  4</td>\n",
1506         "      <td>   319</td>\n",
1507         "      <td> 3</td>\n",
1508         "    </tr>\n",
1509         "    <tr>\n",
1510         "      <th>27</th>\n",
1511         "      <td>  5</td>\n",
1512         "      <td>    73</td>\n",
1513         "      <td> 3</td>\n",
1514         "    </tr>\n",
1515         "    <tr>\n",
1516         "      <th>28</th>\n",
1517         "      <td>  6</td>\n",
1518         "      <td>    25</td>\n",
1519         "      <td> 3</td>\n",
1520         "    </tr>\n",
1521         "    <tr>\n",
1522         "      <th>29</th>\n",
1523         "      <td>  7</td>\n",
1524         "      <td>    12</td>\n",
1525         "      <td> 3</td>\n",
1526         "    </tr>\n",
1527         "    <tr>\n",
1528         "      <th>30</th>\n",
1529         "      <td>  8</td>\n",
1530         "      <td>     6</td>\n",
1531         "      <td> 3</td>\n",
1532         "    </tr>\n",
1533         "    <tr>\n",
1534         "      <th>31</th>\n",
1535         "      <td>  9</td>\n",
1536         "      <td>     2</td>\n",
1537         "      <td> 3</td>\n",
1538         "    </tr>\n",
1539         "    <tr>\n",
1540         "      <th>32</th>\n",
1541         "      <td> 11</td>\n",
1542         "      <td>     1</td>\n",
1543         "      <td> 3</td>\n",
1544         "    </tr>\n",
1545         "    <tr>\n",
1546         "      <th>33</th>\n",
1547         "      <td>  1</td>\n",
1548         "      <td>   287</td>\n",
1549         "      <td> 4</td>\n",
1550         "    </tr>\n",
1551         "    <tr>\n",
1552         "      <th>34</th>\n",
1553         "      <td>  2</td>\n",
1554         "      <td>  1440</td>\n",
1555         "      <td> 4</td>\n",
1556         "    </tr>\n",
1557         "    <tr>\n",
1558         "      <th>35</th>\n",
1559         "      <td>  3</td>\n",
1560         "      <td>   481</td>\n",
1561         "      <td> 4</td>\n",
1562         "    </tr>\n",
1563         "    <tr>\n",
1564         "      <th>36</th>\n",
1565         "      <td>  4</td>\n",
1566         "      <td>   166</td>\n",
1567         "      <td> 4</td>\n",
1568         "    </tr>\n",
1569         "    <tr>\n",
1570         "      <th>37</th>\n",
1571         "      <td>  5</td>\n",
1572         "      <td>    48</td>\n",
1573         "      <td> 4</td>\n",
1574         "    </tr>\n",
1575         "    <tr>\n",
1576         "      <th>38</th>\n",
1577         "      <td>  6</td>\n",
1578         "      <td>    11</td>\n",
1579         "      <td> 4</td>\n",
1580         "    </tr>\n",
1581         "    <tr>\n",
1582         "      <th>39</th>\n",
1583         "      <td>  7</td>\n",
1584         "      <td>     2</td>\n",
1585         "      <td> 4</td>\n",
1586         "    </tr>\n",
1587         "    <tr>\n",
1588         "      <th>40</th>\n",
1589         "      <td>  8</td>\n",
1590         "      <td>     3</td>\n",
1591         "      <td> 4</td>\n",
1592         "    </tr>\n",
1593         "    <tr>\n",
1594         "      <th>41</th>\n",
1595         "      <td>  9</td>\n",
1596         "      <td>     2</td>\n",
1597         "      <td> 4</td>\n",
1598         "    </tr>\n",
1599         "    <tr>\n",
1600         "      <th>42</th>\n",
1601         "      <td> 13</td>\n",
1602         "      <td>     1</td>\n",
1603         "      <td> 4</td>\n",
1604         "    </tr>\n",
1605         "    <tr>\n",
1606         "      <th>43</th>\n",
1607         "      <td> 18</td>\n",
1608         "      <td>     1</td>\n",
1609         "      <td> 4</td>\n",
1610         "    </tr>\n",
1611         "    <tr>\n",
1612         "      <th>44</th>\n",
1613         "      <td>  1</td>\n",
1614         "      <td>    79</td>\n",
1615         "      <td> 5</td>\n",
1616         "    </tr>\n",
1617         "    <tr>\n",
1618         "      <th>45</th>\n",
1619         "      <td>  2</td>\n",
1620         "      <td>   521</td>\n",
1621         "      <td> 5</td>\n",
1622         "    </tr>\n",
1623         "    <tr>\n",
1624         "      <th>46</th>\n",
1625         "      <td>  3</td>\n",
1626         "      <td>   169</td>\n",
1627         "      <td> 5</td>\n",
1628         "    </tr>\n",
1629         "    <tr>\n",
1630         "      <th>47</th>\n",
1631         "      <td>  4</td>\n",
1632         "      <td>    63</td>\n",
1633         "      <td> 5</td>\n",
1634         "    </tr>\n",
1635         "    <tr>\n",
1636         "      <th>48</th>\n",
1637         "      <td>  5</td>\n",
1638         "      <td>    29</td>\n",
1639         "      <td> 5</td>\n",
1640         "    </tr>\n",
1641         "    <tr>\n",
1642         "      <th>49</th>\n",
1643         "      <td>  6</td>\n",
1644         "      <td>    10</td>\n",
1645         "      <td> 5</td>\n",
1646         "    </tr>\n",
1647         "    <tr>\n",
1648         "      <th>50</th>\n",
1649         "      <td>  7</td>\n",
1650         "      <td>     6</td>\n",
1651         "      <td> 5</td>\n",
1652         "    </tr>\n",
1653         "    <tr>\n",
1654         "      <th>51</th>\n",
1655         "      <td>  8</td>\n",
1656         "      <td>     1</td>\n",
1657         "      <td> 5</td>\n",
1658         "    </tr>\n",
1659         "    <tr>\n",
1660         "      <th>52</th>\n",
1661         "      <td>  9</td>\n",
1662         "      <td>     1</td>\n",
1663         "      <td> 5</td>\n",
1664         "    </tr>\n",
1665         "    <tr>\n",
1666         "      <th>53</th>\n",
1667         "      <td>  1</td>\n",
1668         "      <td>    22</td>\n",
1669         "      <td> 6</td>\n",
1670         "    </tr>\n",
1671         "    <tr>\n",
1672         "      <th>54</th>\n",
1673         "      <td>  2</td>\n",
1674         "      <td>   183</td>\n",
1675         "      <td> 6</td>\n",
1676         "    </tr>\n",
1677         "    <tr>\n",
1678         "      <th>55</th>\n",
1679         "      <td>  3</td>\n",
1680         "      <td>    72</td>\n",
1681         "      <td> 6</td>\n",
1682         "    </tr>\n",
1683         "    <tr>\n",
1684         "      <th>56</th>\n",
1685         "      <td>  4</td>\n",
1686         "      <td>    46</td>\n",
1687         "      <td> 6</td>\n",
1688         "    </tr>\n",
1689         "    <tr>\n",
1690         "      <th>57</th>\n",
1691         "      <td>  5</td>\n",
1692         "      <td>    13</td>\n",
1693         "      <td> 6</td>\n",
1694         "    </tr>\n",
1695         "    <tr>\n",
1696         "      <th>58</th>\n",
1697         "      <td>  6</td>\n",
1698         "      <td>     4</td>\n",
1699         "      <td> 6</td>\n",
1700         "    </tr>\n",
1701         "    <tr>\n",
1702         "      <th>59</th>\n",
1703         "      <td>  7</td>\n",
1704         "      <td>     7</td>\n",
1705         "      <td> 6</td>\n",
1706         "    </tr>\n",
1707         "    <tr>\n",
1708         "      <th></th>\n",
1709         "      <td>...</td>\n",
1710         "      <td>...</td>\n",
1711         "      <td>...</td>\n",
1712         "    </tr>\n",
1713         "  </tbody>\n",
1714         "</table>\n",
1715         "<p>108 rows \u00d7 3 columns</p>\n",
1716         "</div>"
1717        ],
1718        "metadata": {},
1719        "output_type": "pyout",
1720        "prompt_number": 33,
1721        "text": [
1722         "    Casualties  Count  Vehicles\n",
1723         "0            1  39217         1\n",
1724         "1            2  65680         1\n",
1725         "2            3   6168         1\n",
1726         "3            4   1056         1\n",
1727         "4            5    197         1\n",
1728         "5            6     76         1\n",
1729         "6            7     27         1\n",
1730         "7            8      4         1\n",
1731         "8            9      1         1\n",
1732         "9           10      1         1\n",
1733         "10          11      1         1\n",
1734         "11           1   3781         2\n",
1735         "12           2  15035         2\n",
1736         "13           3   3002         2\n",
1737         "14           4    716         2\n",
1738         "15           5    128         2\n",
1739         "16           6     41         2\n",
1740         "17           7     14         2\n",
1741         "18           8      4         2\n",
1742         "19           9      6         2\n",
1743         "20          10      3         2\n",
1744         "21          12      1         2\n",
1745         "22          16      1         2\n",
1746         "23           1    682         3\n",
1747         "24           2   4186         3\n",
1748         "25           3   1203         3\n",
1749         "26           4    319         3\n",
1750         "27           5     73         3\n",
1751         "28           6     25         3\n",
1752         "29           7     12         3\n",
1753         "30           8      6         3\n",
1754         "31           9      2         3\n",
1755         "32          11      1         3\n",
1756         "33           1    287         4\n",
1757         "34           2   1440         4\n",
1758         "35           3    481         4\n",
1759         "36           4    166         4\n",
1760         "37           5     48         4\n",
1761         "38           6     11         4\n",
1762         "39           7      2         4\n",
1763         "40           8      3         4\n",
1764         "41           9      2         4\n",
1765         "42          13      1         4\n",
1766         "43          18      1         4\n",
1767         "44           1     79         5\n",
1768         "45           2    521         5\n",
1769         "46           3    169         5\n",
1770         "47           4     63         5\n",
1771         "48           5     29         5\n",
1772         "49           6     10         5\n",
1773         "50           7      6         5\n",
1774         "51           8      1         5\n",
1775         "52           9      1         5\n",
1776         "53           1     22         6\n",
1777         "54           2    183         6\n",
1778         "55           3     72         6\n",
1779         "56           4     46         6\n",
1780         "57           5     13         6\n",
1781         "58           6      4         6\n",
1782         "59           7      7         6\n",
1783         "           ...    ...       ...\n",
1784         "\n",
1785         "[108 rows x 3 columns]"
1786        ]
1787       }
1788      ],
1789      "prompt_number": 33
1790     },
1791     {
1792      "cell_type": "markdown",
1793      "metadata": {},
1794      "source": [
1795       "Plot the data on a bubble chart: the area of the bubble is the number of accidents with that number of casualties and vehicles."
1796      ]
1797     },
1798     {
1799      "cell_type": "code",
1800      "collapsed": false,
1801      "input": [
1802       "plt.scatter(cas_veh['Vehicles'], cas_veh['Casualties'], s=np.sqrt(cas_veh['Count']))\n",
1803       "plt.xlabel('Number of casualties')\n",
1804       "plt.ylabel('Number of vehicles')\n",
1805       "plt.show()"
1806      ],
1807      "language": "python",
1808      "metadata": {},
1809      "outputs": [
1810       {
1811        "metadata": {},
1812        "output_type": "display_data",
1813        "png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAEPCAYAAACk43iMAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
1814        "text": [
1815         "<matplotlib.figure.Figure at 0x7f4bac6a78d0>"
1816        ]
1817       }
1818      ],
1819      "prompt_number": 34
1820     },
1821     {
1822      "cell_type": "markdown",
1823      "metadata": {},
1824      "source": [
1825       "That shows a few outliers of some very large accidents. How large?"
1826      ]
1827     },
1828     {
1829      "cell_type": "code",
1830      "collapsed": false,
1831      "input": [
1832       "print('Casualties:', cas_veh['Casualties'].max(), 'Vehicles:', cas_veh['Vehicles'].max())"
1833      ],
1834      "language": "python",
1835      "metadata": {},
1836      "outputs": [
1837       {
1838        "output_type": "stream",
1839        "stream": "stdout",
1840        "text": [
1841         "Casualties: 18 Vehicles: 42\n"
1842        ]
1843       }
1844      ],
1845      "prompt_number": 35
1846     },
1847     {
1848      "cell_type": "markdown",
1849      "metadata": {},
1850      "source": [
1851       "Let's limit the plot to smallish values. It looks like 10 vehicles and 20 casualties will capture most of the data."
1852      ]
1853     },
1854     {
1855      "cell_type": "code",
1856      "collapsed": false,
1857      "input": [
1858       "small_cas_veh = cas_veh[cas_veh['Casualties'] <= 20][cas_veh['Vehicles'] <= 10]\n",
1859       "plt.scatter(small_cas_veh['Vehicles'], small_cas_veh['Casualties'], s=np.sqrt(small_cas_veh['Count']))\n",
1860       "plt.xlabel('Number of casualties')\n",
1861       "plt.ylabel('Number of vehicles')\n",
1862       "plt.show()"
1863      ],
1864      "language": "python",
1865      "metadata": {},
1866      "outputs": [
1867       {
1868        "metadata": {},
1869        "output_type": "display_data",
1870        "png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEPCAYAAABLIROyAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
1871        "text": [
1872         "<matplotlib.figure.Figure at 0x7f4bac63ac50>"
1873        ]
1874       }
1875      ],
1876      "prompt_number": 36
1877     },
1878     {
1879      "cell_type": "markdown",
1880      "metadata": {},
1881      "source": [
1882       "It seems that there isn't much of a correlation. We can easily find the Pearson's R correlation coefficient, which confirms what the graph shows us.\n",
1883       "\n",
1884       "The calculation assumes there's one row in the data for each data point. That means we need to expand out the data into one long sequence of (casualties, vehicles) pairs. We'll use some functions from the `itertools` built-in library to do that."
1885      ]
1886     },
1887     {
1888      "cell_type": "code",
1889      "collapsed": false,
1890      "input": [
1891       "# Create an iterator of length 'count' for each (casualties, vehicles) pair\n",
1892       "iters = [repeat(cv, n) for cv, n in cas_veh_dict.items()]\n",
1893       "# Chain all the iterators together into one long one, run them, and store the results in a list.\n",
1894       "all_cv = list(chain(*iters))\n",
1895       "# Finally, find the correlation coefficient\n",
1896       "scipy.stats.pearsonr([cv[0] for cv in all_cv], [cv[1] for cv in all_cv])"
1897      ],
1898      "language": "python",
1899      "metadata": {},
1900      "outputs": [
1901       {
1902        "metadata": {},
1903        "output_type": "pyout",
1904        "prompt_number": 37,
1905        "text": [
1906         "(0.24158490939307017, 0.0)"
1907        ]
1908       }
1909      ],
1910      "prompt_number": 37
1911     },
1912     {
1913      "cell_type": "markdown",
1914      "metadata": {},
1915      "source": [
1916       "There were accidents with 42 and 38 casualties. Let's look at those in a bit more detail. First, find the accident documents:"
1917      ]
1918     },
1919     {
1920      "cell_type": "code",
1921      "collapsed": false,
1922      "input": [
1923       "[a for a in accidents.find({'Number_of_Casualties': {'$gt': 35}}, \n",
1924       "                           ['Accident_Index', 'Number_of_Casualties', 'Number_of_Vehicles',\n",
1925       "                            'Datetime', 'Accident_Severity'])]"
1926      ],
1927      "language": "python",
1928      "metadata": {},
1929      "outputs": [
1930       {
1931        "metadata": {},
1932        "output_type": "pyout",
1933        "prompt_number": 38,
1934        "text": [
1935         "[{'Accident_Index': '201220Z000792',\n",
1936         "  'Datetime': datetime.datetime(2012, 3, 24, 6, 20),\n",
1937         "  'Number_of_Casualties': 42,\n",
1938         "  'Number_of_Vehicles': 2,\n",
1939         "  '_id': ObjectId('52a9c98092c4e16686d2287a'),\n",
1940         "  'Accident_Severity': 1},\n",
1941         " {'Accident_Index': '201291NM05840',\n",
1942         "  'Datetime': datetime.datetime(2012, 6, 8, 15, 40),\n",
1943         "  'Number_of_Casualties': 38,\n",
1944         "  'Number_of_Vehicles': 4,\n",
1945         "  '_id': ObjectId('52a9c98492c4e16686d34644'),\n",
1946         "  'Accident_Severity': 1}]"
1947        ]
1948       }
1949      ],
1950      "prompt_number": 38
1951     },
1952     {
1953      "cell_type": "markdown",
1954      "metadata": {},
1955      "source": [
1956       "Let's look into these in a bit more detail, without being overwhelmed by the data."
1957      ]
1958     },
1959     {
1960      "cell_type": "code",
1961      "collapsed": false,
1962      "input": [
1963       "big_accidents = [a for a in accidents.find({'Accident_Index': {'$in': ['201220Z000792', '201291NM05840']}})]"
1964      ],
1965      "language": "python",
1966      "metadata": {},
1967      "outputs": [],
1968      "prompt_number": 39
1969     },
1970     {
1971      "cell_type": "code",
1972      "collapsed": false,
1973      "input": [
1974       "for a in big_accidents:\n",
1975       "    print(a['Accident_Index'], a['Datetime'], \n",
1976       "      label_of[('1st_Road_Class', a['1st_Road_Class'])][0] + str(a['1st_Road_Number']),\n",
1977       "      label_of[('Accident_Severity', a['Accident_Severity'])],\n",
1978       "      a['Number_of_Casualties'], 'casualties,',\n",
1979       "      a['Number_of_Vehicles'], 'vehicles')\n",
1980       "    for v in a['Vehicles']:\n",
1981       "        print(' ', label_of[('Vehicle_Type', v['Vehicle_Type'])], ':',\n",
1982       "              len([c for c in a['Casualties'] if c['Vehicle_Reference'] == v['Vehicle_Reference']]),\n",
1983       "              'casualties')"
1984      ],
1985      "language": "python",
1986      "metadata": {},
1987      "outputs": [
1988       {
1989        "output_type": "stream",
1990        "stream": "stdout",
1991        "text": [
1992         "201220Z000792 2012-03-24 06:20:00 M5 Fatal 42 casualties, 2 vehicles\n",
1993         "  Bus or coach (17 or more pass seats) : 41 casualties\n",
1994         "  Goods 7.5 tonnes mgw and over : 1 casualties\n",
1995         "201291NM05840 2012-06-08 15:40:00 A9 Fatal 38 casualties, 4 vehicles\n",
1996         "  Van / Goods 3.5 tonnes mgw or under : 2 casualties\n",
1997         "  Bus or coach (17 or more pass seats) : 36 casualties\n",
1998         "  Car : 0 casualties\n",
1999         "  Car : 0 casualties\n"
2000        ]
2001       }
2002      ],
2003      "prompt_number": 40
2004     },
2005     {
2006      "cell_type": "markdown",
2007      "metadata": {},
2008      "source": [
2009       "# Exercises"
2010      ]
2011     },
2012     {
2013      "cell_type": "markdown",
2014      "metadata": {},
2015      "source": [
2016       "How many accidents occurred in the Milton Keynes district authority (`'Local_Authority_(District)'` = 479)"
2017      ]
2018     },
2019     {
2020      "cell_type": "code",
2021      "collapsed": false,
2022      "input": [
2023       "accidents.find({'Local_Authority_(District)': 479}).count()"
2024      ],
2025      "language": "python",
2026      "metadata": {},
2027      "outputs": [
2028       {
2029        "metadata": {},
2030        "output_type": "pyout",
2031        "prompt_number": 41,
2032        "text": [
2033         "674"
2034        ]
2035       }
2036      ],
2037      "prompt_number": 41
2038     },
2039     {
2040      "cell_type": "markdown",
2041      "metadata": {},
2042      "source": [
2043       "Which highway authorities are responsible for the roads in the Milton Keynes district authority?"
2044      ]
2045     },
2046     {
2047      "cell_type": "code",
2048      "collapsed": false,
2049      "input": [
2050       "lahs = accidents.find({'Local_Authority_(District)': 479}, ['Local_Authority_(Highway)']).distinct('Local_Authority_(Highway)')\n",
2051       "[(lah, label_of[('Local_Authority_(Highway)', lah)]) for lah in lahs]"
2052      ],
2053      "language": "python",
2054      "metadata": {},
2055      "outputs": [
2056       {
2057        "metadata": {},
2058        "output_type": "pyout",
2059        "prompt_number": 42,
2060        "text": [
2061         "[('E06000042', 'Milton Keynes')]"
2062        ]
2063       }
2064      ],
2065      "prompt_number": 42
2066     },
2067     {
2068      "cell_type": "markdown",
2069      "metadata": {},
2070      "source": [
2071       "Which district authorities are served by Milton Keynes highways?"
2072      ]
2073     },
2074     {
2075      "cell_type": "code",
2076      "collapsed": false,
2077      "input": [
2078       "lads = accidents.find({'Local_Authority_(Highway)': 'E06000042'}, ['Local_Authority_(District)']).distinct('Local_Authority_(District)')\n",
2079       "[(lad, label_of[('Local_Authority_(District)', lad)]) for lad in lads]"
2080      ],
2081      "language": "python",
2082      "metadata": {},
2083      "outputs": [
2084       {
2085        "metadata": {},
2086        "output_type": "pyout",
2087        "prompt_number": 43,
2088        "text": [
2089         "[(479, 'Milton Keynes')]"
2090        ]
2091       }
2092      ],
2093      "prompt_number": 43
2094     },
2095     {
2096      "cell_type": "markdown",
2097      "metadata": {},
2098      "source": [
2099       "Which police forces are responsible for the roads in the Milton Keynes district authority?"
2100      ]
2101     },
2102     {
2103      "cell_type": "code",
2104      "collapsed": false,
2105      "input": [
2106       "pfs = accidents.find({'Local_Authority_(District)': 479}, ['Police_Force']).distinct('Police_Force')\n",
2107       "[(pf, label_of[('Police_Force', pf)]) for pf in pfs]"
2108      ],
2109      "language": "python",
2110      "metadata": {},
2111      "outputs": [
2112       {
2113        "metadata": {},
2114        "output_type": "pyout",
2115        "prompt_number": 44,
2116        "text": [
2117         "[(43, 'Thames Valley')]"
2118        ]
2119       }
2120      ],
2121      "prompt_number": 44
2122     },
2123     {
2124      "cell_type": "markdown",
2125      "metadata": {},
2126      "source": [
2127       "Which district authorities are served by Thames Valley police?"
2128      ]
2129     },
2130     {
2131      "cell_type": "code",
2132      "collapsed": false,
2133      "input": [
2134       "lads = accidents.find({'Police_Force': 43}, ['Local_Authority_(District)']).distinct('Local_Authority_(District)')\n",
2135       "[(lad, label_of[('Local_Authority_(District)', lad)]) for lad in lads]"
2136      ],
2137      "language": "python",
2138      "metadata": {},
2139      "outputs": [
2140       {
2141        "metadata": {},
2142        "output_type": "pyout",
2143        "prompt_number": 45,
2144        "text": [
2145         "[(479, 'Milton Keynes'),\n",
2146         " (477, 'South Bucks'),\n",
2147         " (475, 'Wokingham'),\n",
2148         " (470, 'Bracknell Forest'),\n",
2149         " (484, 'South Oxfordshire'),\n",
2150         " (471, 'West Berkshire'),\n",
2151         " (472, 'Reading'),\n",
2152         " (485, 'West Oxfordshire'),\n",
2153         " (481, 'Cherwell'),\n",
2154         " (480, 'Wycombe'),\n",
2155         " (476, 'Aylesbury Vale'),\n",
2156         " (474, 'Windsor and Maidenhead'),\n",
2157         " (483, 'Vale of White Horse'),\n",
2158         " (473, 'Slough'),\n",
2159         " (482, 'Oxford'),\n",
2160         " (478, 'Chiltern')]"
2161        ]
2162       }
2163      ],
2164      "prompt_number": 45
2165     },
2166     {
2167      "cell_type": "markdown",
2168      "metadata": {},
2169      "source": [
2170       "How many casualties were involved in accident index '201243S022012'"
2171      ]
2172     },
2173     {
2174      "cell_type": "code",
2175      "collapsed": false,
2176      "input": [
2177       "accidents.find_one({'Accident_Index': '201243S022012'}, ['Number_of_Casualties'])"
2178      ],
2179      "language": "python",
2180      "metadata": {},
2181      "outputs": [
2182       {
2183        "metadata": {},
2184        "output_type": "pyout",
2185        "prompt_number": 46,
2186        "text": [
2187         "{'_id': ObjectId('52a9c98292c4e16686d2b607'), 'Number_of_Casualties': 4}"
2188        ]
2189       }
2190      ],
2191      "prompt_number": 46
2192     },
2193     {
2194      "cell_type": "code",
2195      "collapsed": false,
2196      "input": [
2197       "[a for a in accidents.find({'Local_Authority_(District)': 479}, \n",
2198       "                           ['Accident_Index', 'Number_of_Casualties', 'Number_of_Vehicles'],\n",
2199       "                           limit=10)]"
2200      ],
2201      "language": "python",
2202      "metadata": {},
2203      "outputs": [
2204       {
2205        "metadata": {},
2206        "output_type": "pyout",
2207        "prompt_number": 47,
2208        "text": [
2209         "[{'Number_of_Vehicles': 1,\n",
2210         "  '_id': ObjectId('52a9c98292c4e16686d2b5f6'),\n",
2211         "  'Accident_Index': '201243S018012',\n",
2212         "  'Number_of_Casualties': 1},\n",
2213         " {'Number_of_Vehicles': 2,\n",
2214         "  '_id': ObjectId('52a9c98292c4e16686d2b6a5'),\n",
2215         "  'Accident_Index': '201243S056012',\n",
2216         "  'Number_of_Casualties': 1},\n",
2217         " {'Number_of_Vehicles': 2,\n",
2218         "  '_id': ObjectId('52a9c98292c4e16686d2b5e6'),\n",
2219         "  'Accident_Index': '201243S014012',\n",
2220         "  'Number_of_Casualties': 1},\n",
2221         " {'Number_of_Vehicles': 2,\n",
2222         "  '_id': ObjectId('52a9c98292c4e16686d2b60c'),\n",
2223         "  'Accident_Index': '201243S023012',\n",
2224         "  'Number_of_Casualties': 1},\n",
2225         " {'Number_of_Vehicles': 2,\n",
2226         "  '_id': ObjectId('52a9c98292c4e16686d2b5fe'),\n",
2227         "  'Accident_Index': '201243S020012',\n",
2228         "  'Number_of_Casualties': 1},\n",
2229         " {'Number_of_Vehicles': 2,\n",
2230         "  '_id': ObjectId('52a9c98292c4e16686d2b5f9'),\n",
2231         "  'Accident_Index': '201243S019012',\n",
2232         "  'Number_of_Casualties': 1},\n",
2233         " {'Number_of_Vehicles': 2,\n",
2234         "  '_id': ObjectId('52a9c98292c4e16686d2b607'),\n",
2235         "  'Accident_Index': '201243S022012',\n",
2236         "  'Number_of_Casualties': 4},\n",
2237         " {'Number_of_Vehicles': 2,\n",
2238         "  '_id': ObjectId('52a9c98292c4e16686d2b649'),\n",
2239         "  'Accident_Index': '201243S036012',\n",
2240         "  'Number_of_Casualties': 1},\n",
2241         " {'Number_of_Vehicles': 2,\n",
2242         "  '_id': ObjectId('52a9c98292c4e16686d2b683'),\n",
2243         "  'Accident_Index': '201243S048012',\n",
2244         "  'Number_of_Casualties': 1},\n",
2245         " {'Number_of_Vehicles': 2,\n",
2246         "  '_id': ObjectId('52a9c98292c4e16686d2b621'),\n",
2247         "  'Accident_Index': '201243S028012',\n",
2248         "  'Number_of_Casualties': 1}]"
2249        ]
2250       }
2251      ],
2252      "prompt_number": 47
2253     },
2254     {
2255      "cell_type": "code",
2256      "collapsed": false,
2257      "input": [
2258       "[(v, label_of[('Police_Force', v)]) for k, v in label_of.keys() if k == 'Police_Force']"
2259      ],
2260      "language": "python",
2261      "metadata": {},
2262      "outputs": [
2263       {
2264        "metadata": {},
2265        "output_type": "pyout",
2266        "prompt_number": 48,
2267        "text": [
2268         "[(14, 'South Yorkshire'),\n",
2269         " (32, 'Lincolnshire'),\n",
2270         " (92, 'Grampian'),\n",
2271         " (1, 'Metropolitan Police'),\n",
2272         " (61, 'Gwent'),\n",
2273         " (96, 'Central'),\n",
2274         " (47, 'Sussex'),\n",
2275         " (91, 'Northern'),\n",
2276         " (12, 'North Yorkshire'),\n",
2277         " (7, 'Cheshire'),\n",
2278         " (34, 'Northamptonshire'),\n",
2279         " (16, 'Humberside'),\n",
2280         " (42, 'Essex'),\n",
2281         " (11, 'Durham'),\n",
2282         " (45, 'Surrey'),\n",
2283         " (44, 'Hampshire'),\n",
2284         " (36, 'Norfolk'),\n",
2285         " (62, 'South Wales'),\n",
2286         " (5, 'Merseyside'),\n",
2287         " (31, 'Nottinghamshire'),\n",
2288         " (46, 'Kent'),\n",
2289         " (22, 'West Mercia'),\n",
2290         " (40, 'Bedfordshire'),\n",
2291         " (95, 'Lothian and Borders'),\n",
2292         " (60, 'North Wales'),\n",
2293         " (55, 'Dorset'),\n",
2294         " (20, 'West Midlands'),\n",
2295         " (33, 'Leicestershire'),\n",
2296         " (93, 'Tayside'),\n",
2297         " (52, 'Avon and Somerset'),\n",
2298         " (6, 'Greater Manchester'),\n",
2299         " (50, 'Devon and Cornwall'),\n",
2300         " (35, 'Cambridgeshire'),\n",
2301         " (53, 'Gloucestershire'),\n",
2302         " (4, 'Lancashire'),\n",
2303         " (97, 'Strathclyde'),\n",
2304         " (10, 'Northumbria'),\n",
2305         " (13, 'West Yorkshire'),\n",
2306         " (30, 'Derbyshire'),\n",
2307         " (21, 'Staffordshire'),\n",
2308         " (48, 'City of London'),\n",
2309         " (17, 'Cleveland'),\n",
2310         " (43, 'Thames Valley'),\n",
2311         " (94, 'Fife'),\n",
2312         " (3, 'Cumbria'),\n",
2313         " (37, 'Suffolk'),\n",
2314         " (63, 'Dyfed-Powys'),\n",
2315         " (54, 'Wiltshire'),\n",
2316         " (98, 'Dumfries and Galloway'),\n",
2317         " (23, 'Warwickshire'),\n",
2318         " (41, 'Hertfordshire')]"
2319        ]
2320       }
2321      ],
2322      "prompt_number": 48
2323     },
2324     {
2325      "cell_type": "code",
2326      "collapsed": false,
2327      "input": [],
2328      "language": "python",
2329      "metadata": {},
2330      "outputs": [],
2331      "prompt_number": 48
2332     }
2333    ],
2334    "metadata": {}
2335   }
2336  ]
2337 }