Notebooks/ArcGIS/Python walkthrough ArcGIS Data analysis and ML.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## This is an end to end example of using ArcGIS with python \n",
    "this has been mostly followed as a simplefied version from here https://github.com/orhuna/RF-Demo/blob/master/RF_Demo_Fl.ipynb\n",
    "\n",
    "For better visualization, maps have been published into the GIS server with a public read access to display it in the notebook cell\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "import numpy as NUM\n",
    "import arcpy as ARCPY\n",
    "import arcpy.da as DA\n",
    "import pandas as PD\n",
    "import seaborn as SEA\n",
    "import matplotlib.pyplot as PLOT\n",
    "import pandas as PD\n",
    "import arcgisscripting as ARC\n",
    "import arcpy as ARCPY\n",
    "import SSUtilities as UTILS\n",
    "from arcgis.gis import *\n",
    "import os as OS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "mygis = GIS()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from IPython.display import IFrame  \n",
    "def show_web_map_by_url(URL):\n",
    "    url = 'http://www.arcgis.com/home/webmap/viewer.html?url=' + URL\n",
    "    return IFrame(url, width='100%', height=500)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "analysis_map_URL = 'https://services3.arcgis.com/oZfKvdlWHN1MwS48/ArcGIS/rest/services/MachineLearningSeagrass/FeatureServer/1&source=sd'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This is the initial map which has florida data to start the data analysis "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe\n",
       "            width=\"100%\"\n",
       "            height=\"500\"\n",
       "            src=\"http://www.arcgis.com/home/webmap/viewer.html?url=https://services3.arcgis.com/oZfKvdlWHN1MwS48/ArcGIS/rest/services/MachineLearningSeagrass/FeatureServer/1&source=sd\"\n",
       "            frameborder=\"0\"\n",
       "            allowfullscreen\n",
       "        ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.lib.display.IFrame at 0x20bce8590b8>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Where the seagrasses are in florida\n",
    "\n",
    "show_web_map_by_url(analysis_map_URL)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Load the features from the shapefile stored in GIS Desktop\n",
    "inputDir = 'C:/GISDemo/SeaGrass/SeaGrass.gdb/FloridaSeaGrass'\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Names of Prediction Variables\n",
    "predictVars = ['salinity', 'temp', 'phosphate','nitrate', \n",
    "               'silicate', 'dissO2', 'NameEMU']\n",
    "#Name of Classification Variable\n",
    "classVar = ['Present']\n",
    "#List of all Variables\n",
    "allVars = predictVars + classVar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>salinity</th>\n",
       "      <th>temp</th>\n",
       "      <th>phosphate</th>\n",
       "      <th>nitrate</th>\n",
       "      <th>silicate</th>\n",
       "      <th>dissO2</th>\n",
       "      <th>NameEMU</th>\n",
       "      <th>Present</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>35.694950</td>\n",
       "      <td>24.75329</td>\n",
       "      <td>0.070516</td>\n",
       "      <td>0.503541</td>\n",
       "      <td>3.878356</td>\n",
       "      <td>5.038187</td>\n",
       "      <td>Epipelagic, Warm to Very Warm, Euhaline, Oxic,...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>35.985001</td>\n",
       "      <td>24.13740</td>\n",
       "      <td>0.124987</td>\n",
       "      <td>0.836354</td>\n",
       "      <td>1.521608</td>\n",
       "      <td>4.710738</td>\n",
       "      <td>Epipelagic, Warm to Very Warm, Euhaline, Oxic,...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>35.581039</td>\n",
       "      <td>18.00565</td>\n",
       "      <td>1.032444</td>\n",
       "      <td>15.172270</td>\n",
       "      <td>10.742860</td>\n",
       "      <td>3.839985</td>\n",
       "      <td>Epipelagic, Warm to Very Warm, Euhaline, Oxic,...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>35.469360</td>\n",
       "      <td>14.55408</td>\n",
       "      <td>0.936634</td>\n",
       "      <td>14.109170</td>\n",
       "      <td>9.925514</td>\n",
       "      <td>3.654587</td>\n",
       "      <td>Epipelagic, Warm to Very Warm, Euhaline, Oxic,...</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>35.413799</td>\n",
       "      <td>13.48950</td>\n",
       "      <td>0.980522</td>\n",
       "      <td>14.746720</td>\n",
       "      <td>10.737210</td>\n",
       "      <td>3.721700</td>\n",
       "      <td>Epipelagic, Warm to Very Warm, Euhaline, Oxic,...</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    salinity      temp  phosphate    nitrate   silicate    dissO2  \\\n",
       "0  35.694950  24.75329   0.070516   0.503541   3.878356  5.038187   \n",
       "1  35.985001  24.13740   0.124987   0.836354   1.521608  4.710738   \n",
       "2  35.581039  18.00565   1.032444  15.172270  10.742860  3.839985   \n",
       "3  35.469360  14.55408   0.936634  14.109170   9.925514  3.654587   \n",
       "4  35.413799  13.48950   0.980522  14.746720  10.737210  3.721700   \n",
       "\n",
       "                                             NameEMU  Present  \n",
       "0  Epipelagic, Warm to Very Warm, Euhaline, Oxic,...        0  \n",
       "1  Epipelagic, Warm to Very Warm, Euhaline, Oxic,...        0  \n",
       "2  Epipelagic, Warm to Very Warm, Euhaline, Oxic,...        1  \n",
       "3  Epipelagic, Warm to Very Warm, Euhaline, Oxic,...        1  \n",
       "4  Epipelagic, Warm to Very Warm, Euhaline, Oxic,...        0  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#GIS data to Pandas dataframe\n",
    "\n",
    "#Get Data from ArcGIS Desktop\n",
    "FlLayer = DA.FeatureClassToNumPyArray(inputDir, [\"SHAPE@XY\"] + allVars)\n",
    "#Obtain Spatial Reference\n",
    "spatRef = ARCPY.Describe(inputDir).spatialReference\n",
    "#Define Main Dataframe\n",
    "data = PD.DataFrame(FlLayer, columns = allVars)\n",
    "#Display Portion of the Data Frame\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Process Categorical Data for Analysis flatten the categorical data \n",
    "#Create Numeric Fields for One-Hot Encoding of the Categorical Variable\n",
    "catVars = PD.get_dummies(data[predictVars[-1]])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>salinity</th>\n",
       "      <th>temp</th>\n",
       "      <th>phosphate</th>\n",
       "      <th>nitrate</th>\n",
       "      <th>silicate</th>\n",
       "      <th>dissO2</th>\n",
       "      <th>Present</th>\n",
       "      <th>c1</th>\n",
       "      <th>c2</th>\n",
       "      <th>c3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>35.694950</td>\n",
       "      <td>24.75329</td>\n",
       "      <td>0.070516</td>\n",
       "      <td>0.503541</td>\n",
       "      <td>3.878356</td>\n",
       "      <td>5.038187</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>35.985001</td>\n",
       "      <td>24.13740</td>\n",
       "      <td>0.124987</td>\n",
       "      <td>0.836354</td>\n",
       "      <td>1.521608</td>\n",
       "      <td>4.710738</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>35.581039</td>\n",
       "      <td>18.00565</td>\n",
       "      <td>1.032444</td>\n",
       "      <td>15.172270</td>\n",
       "      <td>10.742860</td>\n",
       "      <td>3.839985</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>35.469360</td>\n",
       "      <td>14.55408</td>\n",
       "      <td>0.936634</td>\n",
       "      <td>14.109170</td>\n",
       "      <td>9.925514</td>\n",
       "      <td>3.654587</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>35.413799</td>\n",
       "      <td>13.48950</td>\n",
       "      <td>0.980522</td>\n",
       "      <td>14.746720</td>\n",
       "      <td>10.737210</td>\n",
       "      <td>3.721700</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    salinity      temp  phosphate    nitrate   silicate    dissO2  Present  \\\n",
       "0  35.694950  24.75329   0.070516   0.503541   3.878356  5.038187        0   \n",
       "1  35.985001  24.13740   0.124987   0.836354   1.521608  4.710738        0   \n",
       "2  35.581039  18.00565   1.032444  15.172270  10.742860  3.839985        1   \n",
       "3  35.469360  14.55408   0.936634  14.109170   9.925514  3.654587        1   \n",
       "4  35.413799  13.48950   0.980522  14.746720  10.737210  3.721700        0   \n",
       "\n",
       "   c1  c2  c3  \n",
       "0   0   1   0  \n",
       "1   0   1   0  \n",
       "2   0   1   0  \n",
       "3   0   1   0  \n",
       "4   0   1   0  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Remove raw Categories from Dataset\n",
    "data = data.drop(predictVars[-1], axis = 1)\n",
    "#Add Processed Categories Back into the Data Frame\n",
    "data = data.join(catVars)\n",
    "#Abbreviate Long Categorical Variable Names\n",
    "newNames = ['c1','c2','c3']\n",
    "for ind, name in enumerate(newNames):\n",
    "    data.rename(columns={data.columns[len(predictVars)+ind]:name}, inplace=True)\n",
    "#Update Predict Variable Names\n",
    "predictVarsNew = predictVars[:-1] + newNames\n",
    "#Display Portion of the Data Frame\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAV4AAAEmCAYAAAAqWvi2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ncBhjeA0Gg8Fh/h/TH9jeL6zZyQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x20bcc3c24e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "##EVALUATE CORRELATION BETWEEN PREDICTORS- EXCLUDE CATEGORICAL\n",
    "#Calculate Correlation Coefficient between Prediction Variables\n",
    "corr = data.drop(data.columns[-3:], axis = 1).astype('float64').corr()\n",
    "\n",
    "#Plot Correlation Matrix Between Prediction Variables\n",
    "ax = SEA.heatmap(corr, cmap=SEA.diverging_palette(220, 10, as_cmap=True),\n",
    "            square=True, annot = True, linecolor = 'k', linewidths = 1)\n",
    "\n",
    "PLOT.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Data Size = 300000\n",
      "Test Data Size = 700000\n"
     ]
    }
   ],
   "source": [
    "##PERFORM RANDOM FOREST CLASSIFICATION\n",
    "#Fraction of data to be used in Training\n",
    "fracNum = 0.3\n",
    "#Seperate the Data into Training and Test Datasets\n",
    "train_c1 = data[data[classVar[0]] == 1].sample(frac = fracNum)\n",
    "train_c0 = data[data[classVar[0]] == 0].sample(frac = fracNum)\n",
    "#Create Training Dataset\n",
    "train_set = PD.concat([train_c0, train_c1])\n",
    "#Create Testing Dataset\n",
    "test_set = data.drop(train_set.index)\n",
    "#Encode Seagrass Presence as Classes\n",
    "indicator, _ = PD.factorize(train_set[classVar[0]])\n",
    "#Print Test and Train Data Set Sizes\n",
    "print('Training Data Size = ' + str(train_set.size))\n",
    "print('Test Data Size = ' + str(test_set.size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=500, n_jobs=1,\n",
       "            oob_score=True, random_state=None, verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "##Create Random Forest Classification Object\n",
    "rfco = RandomForestClassifier(n_estimators = 500, oob_score = True)\n",
    "#Perform Classification Using Training Set\n",
    "rfco.fit(train_set[predictVars[:-1]], indicator)\n",
    "#Plot Variable Importance\n",
    "#Dummy Variables for Y-Ticks\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Predict Seagrass Occurance for the Test Dataset\n",
    "seagrassPred = rfco.predict(test_set[predictVars[:-1]])\n",
    "#Calculate Prediction Accuracy\n",
    "test_seagrass = test_set['Present'].values.flatten()\n",
    "#Calculate Estimation Error\n",
    "error = NUM.sum(test_seagrass - seagrassPred)/len(seagrassPred) * 100\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy = 99.66285714285715 % \n",
      "Locations with Seagrass = 21612\n",
      "Predicted Locations with Seagrass = 21376\n"
     ]
    }
   ],
   "source": [
    "#Print Accuracy Metrics\n",
    "print('Accuracy = ' + str(100 - NUM.abs(error)) + ' % ')\n",
    "print('Locations with Seagrass = ' + str(len(NUM.where(test_seagrass==1)[0])) )\n",
    "print('Predicted Locations with Seagrass = ' + str(len(NUM.where(seagrassPred==1)[0])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Result 'C:/GISDemo/SeaGrass/SeaGrass.gdb\\\\Florida_Seagrass_Prediction'>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    " #Delete the feature  as ARCPY.da.NumPyArrayToFeatureClass does not overwrite \n",
    "#ARCPY.DeleteFeatures_management(\"C:/GISDemo/SeaGrass/SeaGrass.gdb\\Florida_Seagrass_Prediction\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "##BRING OUTPUT BACK INTO ArcGIS Desktop\n",
    "#Allow overwriting Feature Classes\n",
    "ARCPY.env.overwriteOutput = True\n",
    "#Get Indexes for the Test Dataset\n",
    "outputDir = r'C:/GISDemo/SeaGrass/SeaGrass.gdb'\n",
    "nameFC = 'Florida_Seagrass_Prediction_Py'\n",
    "#Locations with Seagrass\n",
    "grassExists = FlLayer[[\"SHAPE@XY\"]][test_set.index[NUM.where(seagrassPred==1)]]\n",
    "# Write Locations with Seagrass to Feature Class\n",
    "ARCPY.da.NumPyArrayToFeatureClass(grassExists, OS.path.join(outputDir, nameFC), ['SHAPE@XY'], spatRef)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "predicted_fl_map_url ='https://www.arcgis.com/home/webmap/viewer.html?webmap=2e65c37b14764e7a875efb5b5bad4633'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe\n",
       "            width=\"100%\"\n",
       "            height=\"500\"\n",
       "            src=\"https://www.arcgis.com/home/webmap/viewer.html?webmap=2e65c37b14764e7a875efb5b5bad4633\"\n",
       "            frameborder=\"0\"\n",
       "            allowfullscreen\n",
       "        ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.lib.display.IFrame at 0x23e810bdba8>"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "IFrame(predicted_fl_map_url, width='100%', height=500)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "Now create a model from the USA data set and we will use this to predict the sea grass locations worldwide"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>salinity</th>\n",
       "      <th>temp</th>\n",
       "      <th>phosphate</th>\n",
       "      <th>nitrate</th>\n",
       "      <th>silicate</th>\n",
       "      <th>dissO2</th>\n",
       "      <th>Present</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>33.757988</td>\n",
       "      <td>8.378799</td>\n",
       "      <td>2.04564</td>\n",
       "      <td>25.473619</td>\n",
       "      <td>35.493446</td>\n",
       "      <td>3.365844</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>33.757988</td>\n",
       "      <td>8.378799</td>\n",
       "      <td>2.04564</td>\n",
       "      <td>25.473619</td>\n",
       "      <td>35.493446</td>\n",
       "      <td>3.365844</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>33.757988</td>\n",
       "      <td>8.378799</td>\n",
       "      <td>2.04564</td>\n",
       "      <td>25.473619</td>\n",
       "      <td>35.493446</td>\n",
       "      <td>3.365844</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>33.757988</td>\n",
       "      <td>8.378799</td>\n",
       "      <td>2.04564</td>\n",
       "      <td>25.473619</td>\n",
       "      <td>35.493446</td>\n",
       "      <td>3.365844</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>33.757988</td>\n",
       "      <td>8.378799</td>\n",
       "      <td>2.04564</td>\n",
       "      <td>25.473619</td>\n",
       "      <td>35.493446</td>\n",
       "      <td>3.365844</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    salinity      temp  phosphate    nitrate   silicate    dissO2  Present\n",
       "0  33.757988  8.378799    2.04564  25.473619  35.493446  3.365844        0\n",
       "1  33.757988  8.378799    2.04564  25.473619  35.493446  3.365844        0\n",
       "2  33.757988  8.378799    2.04564  25.473619  35.493446  3.365844        0\n",
       "3  33.757988  8.378799    2.04564  25.473619  35.493446  3.365844        0\n",
       "4  33.757988  8.378799    2.04564  25.473619  35.493446  3.365844        0"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Names of Prediction Variables#Import USA Seagrass Data \n",
    "inputDir = 'C:/GISDemo/SeaGrass/SeaGrass.gdb/USASeaGrass_training'\n",
    "predictVars = ['salinity', 'temp', 'phosphate','nitrate', \n",
    "               'silicate', 'dissO2']\n",
    "#Name of Classification Variable\n",
    "classVar = ['Present']\n",
    "#List of all Variables\n",
    "allVars = predictVars + classVar\n",
    "#Create a Data Object\n",
    "USAlayer = DA.FeatureClassToNumPyArray(inputDir, [\"SHAPE@XY\"] + allVars)\n",
    "#Obtain Spatial Reference\n",
    "spatRefGlobal = ARCPY.Describe(inputDir).spatialReference\n",
    "#Define Main Dataframe\n",
    "USA_Train = PD.DataFrame(USAlayer, columns = allVars)\n",
    "#Display Portion of the Data Frame\n",
    "USA_Train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
       "            max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
       "            min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "            min_samples_leaf=1, min_samples_split=2,\n",
       "            min_weight_fraction_leaf=0.0, n_estimators=500, n_jobs=1,\n",
       "            oob_score=False, random_state=None, verbose=0,\n",
       "            warm_start=False)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "##Train Random Forest Using USA Data\n",
    "#Encode Seagrass Presence as Classes\n",
    "indicatorUSA, _ = PD.factorize(USA_Train['Present'])\n",
    "#Create Random Forest Classification Object\n",
    "rfco = RandomForestClassifier(n_estimators = 500)\n",
    "#Perform Classification Using Training Set\n",
    "rfco.fit(USA_Train[predictVars[:-1]], indicatorUSA)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['salinity', 'temp', 'phosphate', 'nitrate', 'silicate']"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictVars[:-1]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "## Import Global Data for Prediction ##\n",
    "##globalUrl = r'https://services3.arcgis.com/oZfKvdlWHN1MwS48/arcgis/rest/services/MachineLearningSeagrass/FeatureServer/2?token=W4CSlYAQxBZKNONRrS_MVm0qNa7j21xD5BNMByQOuXTQ6xhXxFydToukAXYrndK0RRrMNjvbFHcbCOioIVNAgIe6f0iV7rclZ4kNGqoI3i1Ldjhv_UgwzeBGc7KJacuAshcwYXkb6PIgrGipbwD3YOV3AGhhRmh9e5RzMSWB7HJS9fSYzU2l8ZL9vkLmWccKvo_DGJaqm9GIkz9HoPM2hv2yGDdf9SEtcoQvBsNkCWk.'\n",
    "globalUrl = 'C:/GISDemo/SeaGrass/SeaGrass.gdb/Global_Seagrass_Prediction'\n",
    "#Create a SS Data Object\n",
    "globalData = DA.FeatureClassToNumPyArray(globalUrl, [\"SHAPE@XY\"] + predictVars[:-1])\n",
    "#Obtain Spatial Reference\n",
    "spatRefGlobal = ARCPY.Describe(globalUrl).spatialReference\n",
    "#Define Dataframe\n",
    "globalTrain = PD.DataFrame(globalData, columns = predictVars[:-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Predict Global Seagrass Occurance\n",
    "seagrassPredGlobal = rfco.predict(globalTrain)\n",
    "#Bring Output Back to ArcGIS\n",
    "nameFC = 'GlobSeagrass'\n",
    "outputDir = r'C:/GISDemo/SeaGrass/SeaGrass.gdb'\n",
    "nameFC = 'Global_Seagrass_Prediction_Done'\n",
    "grassExists = globalData[[\"SHAPE@XY\"]][globalTrain.index[NUM.where(seagrassPredGlobal==1)]]\n",
    "# Write Locations with Seagrass to Feature Class\n",
    "ARCPY.da.NumPyArrayToFeatureClass(grassExists, OS.path.join(outputDir, nameFC), ['SHAPE@XY'], spatRefGlobal)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "predicted_gl_map_url='https://www.arcgis.com/home/webmap/viewer.html?webmap=503efd8acb7b48e3b92b4f4ac4d6e82e'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Global prediction data as shown on the map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe\n",
       "            width=\"100%\"\n",
       "            height=\"500\"\n",
       "            src=\"https://www.arcgis.com/home/webmap/viewer.html?webmap=503efd8acb7b48e3b92b4f4ac4d6e82e\"\n",
       "            frameborder=\"0\"\n",
       "            allowfullscreen\n",
       "        ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.lib.display.IFrame at 0x23e890a2710>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "IFrame(predicted_gl_map_url, width='100%', height=500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [default]",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}