{
 "metadata": {
  "name": "",
  "signature": "sha256:509540561ce48e9ff8f4ccbe194cd1a6c9560a35cebe46ba1dd2ef9f17394e6f"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Initialize Earth Engine."
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "# Load code to setup authorization for an IPython Notebook. Note that this is a temporary step\n",
      "# that is required until the Earth Engine Python API is updated to include this logic.\n",
      "%run 'authorize_earth_engine_in_notebook.ipynb'\n",
      "\n",
      "# Initialize Earth Engine\n",
      "# Note that we are calling a function defined in the previously run IPython Notebook, rather than\n",
      "# the typical call to ee.Initialize()\n",
      "ee_initialize()"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "heading",
     "level": 1,
     "metadata": {},
     "source": [
      "Load the Google Maps Interactive Widget"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "%run 'define_google_maps_interactive_widget.ipynb'"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "heading",
     "level": 2,
     "metadata": {},
     "source": [
      "Nile Delta Classification example"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "map = GoogleMapsWidget(lat=31.073, lng=30.732, zoom=8)\n",
      "display(map)\n",
      "\n",
      "composite = ee.Image(\"L7_TOA_1YEAR_2000\")\n",
      "\n",
      "bands = [\"10\", \"20\", \"30\", \"40\", \"50\", \"62\", \"70\"]\n",
      "\n",
      "# Use the MCD12 land-cover as training data.\n",
      "cover = ee.Image(\"MCD12Q1/MCD12Q1_005_2001_01_01\").select(\"Land_Cover_Type_1\")\n",
      "\n",
      "# A region of the image to train with.\n",
      "region = ee.Geometry.Rectangle(30, 30, 32.5, 31.6)\n",
      "\n",
      "PALETTE = ','.join([\n",
      "    'aec3d4', # water\n",
      "    '152106', '225129', '369b47', '30eb5b', '387242', # forest\n",
      "    '6a2325', 'c3aa69', 'b76031', 'd9903d', '91af40', # shrub, grass, savanah\n",
      "    '111149', # wetlands\n",
      "    '8dc33b', # croplands\n",
      "    'cc0013', # urban\n",
      "    '6ca80d', # crop mosaic\n",
      "    'd7cdcc', # snow and ice\n",
      "    'f7e084', # barren\n",
      "    '6f6f6f'  # tundra\n",
      "])\n",
      "\n",
      "map.addLayer(\n",
      "    image=composite, \n",
      "    vis_params={'min':0, 'max':100, 'bands':\"30,20,10\"}, \n",
      "    name='Landsat 7 TOA 2000')\n",
      "map.addLayer(\n",
      "    image=cover,\n",
      "    vis_params={'palette': PALETTE, 'min':0, 'max':17},\n",
      "    name='MODIS land-cover',\n",
      "    visible=False)\n",
      "\n",
      "# Multiple trainings with different numbers of points.\n",
      "nums = [1000, 10000, 100000]\n",
      "for num in nums:\n",
      "    # Extract random points.\n",
      "    points = ee.FeatureCollection.randomPoints(region, num, num, 1)\n",
      "    training = cover.addBands(composite).reduceToVectors(\n",
      "        reducer=\"mean\",\n",
      "        geometry=points,\n",
      "        geometryType=\"centroid\",\n",
      "        scale=30, \n",
      "        crs=\"EPSG:4326\"\n",
      "    )\n",
      "    \n",
      "    classifier = training.trainClassifier(\n",
      "        property_list=bands,\n",
      "        class_property=\"label\",\n",
      "        classifier_name=\"FastNaiveBayes\"\n",
      "    )\n",
      "\n",
      "    # Apply the classifier to the original composite.\n",
      "    out = composite.classify(classifier)\n",
      "    map.addLayer(\n",
      "        image=out,\n",
      "        vis_params={'palette': PALETTE, 'min':0, 'max':17},\n",
      "        name=\"{0} points\".format(num),\n",
      "        visible=False\n",
      "    )\n",
      "\n",
      "# Add the region that constrains the random sampling.\n",
      "map.addLayer(\n",
      "        image=ee.Image().paint(ee.Feature(region), 0, 2),\n",
      "        name=\"Region\"\n",
      ")"
     ],
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}