H2O and XGBoost on GPU.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# GPU Performance Test: H2O - XGBoost on GPU"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![GPU vs CPU](https://s18.postimg.org/ataodekvt/GPU_vs_CPU.jpg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Import the H2O Library and its XGBoost Estimator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import h2o\n",
    "from h2o.estimators.xgboost import H2OXGBoostEstimator"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Connect to a Running H2O instance or start a new one"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture\n",
    "h2o.connect(ip=\"35.227.47.29\")\n",
    "h2o.no_progress()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Load Some Data From an Amazon S3 Bucket"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import some data from Amazon S3\n",
    "h2oDF = h2o.import_file(path=\"https://s3-us-west-1.amazonaws.com/dsclouddata/LendingClubData/LoansGoodBad.csv\")\n",
    "\n",
    "\n",
    "# Stratified Split into Train/Test\n",
    "stratsplit = h2oDF[\"Bad_Loan\"].stratified_split(test_frac=0.3, seed=12349453)\n",
    "train = h2oDF[stratsplit==\"train\"]\n",
    "test = h2oDF[stratsplit==\"test\"]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Specify the Response Column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead>\n",
       "<tr><th style=\"text-align: right;\">  RowID</th><th style=\"text-align: right;\">  Loan_Amount</th><th>Term     </th><th style=\"text-align: right;\">  Interest_Rate</th><th style=\"text-align: right;\">  Employment_Years</th><th>Home_Ownership  </th><th style=\"text-align: right;\">  Annual_Income</th><th>Verification_Status     </th><th>Loan_Purpose  </th><th>State  </th><th style=\"text-align: right;\">  Debt_to_Income</th><th style=\"text-align: right;\">  Delinquent_2yr</th><th style=\"text-align: right;\">  Revolving_Cr_Util</th><th style=\"text-align: right;\">  Total_Accounts</th><th>Bad_Loan  </th><th style=\"text-align: right;\">  Longest_Credit_Length</th></tr>\n",
       "</thead>\n",
       "<tbody>\n",
       "<tr><td style=\"text-align: right;\">      2</td><td style=\"text-align: right;\">         2500</td><td>60 months</td><td style=\"text-align: right;\">          15.27</td><td style=\"text-align: right;\">               0.5</td><td>RENT            </td><td style=\"text-align: right;\">          30000</td><td>VERIFIED - income source</td><td>car           </td><td>GA     </td><td style=\"text-align: right;\">            1   </td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">                9.4</td><td style=\"text-align: right;\">               4</td><td>BAD       </td><td style=\"text-align: right;\">                     12</td></tr>\n",
       "<tr><td style=\"text-align: right;\">      3</td><td style=\"text-align: right;\">         2400</td><td>36 months</td><td style=\"text-align: right;\">          15.96</td><td style=\"text-align: right;\">              10  </td><td>RENT            </td><td style=\"text-align: right;\">          12252</td><td>not verified            </td><td>small_business</td><td>IL     </td><td style=\"text-align: right;\">            8.72</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">               98.5</td><td style=\"text-align: right;\">              10</td><td>GOOD      </td><td style=\"text-align: right;\">                     10</td></tr>\n",
       "<tr><td style=\"text-align: right;\">      4</td><td style=\"text-align: right;\">        10000</td><td>36 months</td><td style=\"text-align: right;\">          13.49</td><td style=\"text-align: right;\">              10  </td><td>RENT            </td><td style=\"text-align: right;\">          49200</td><td>VERIFIED - income source</td><td>other         </td><td>CA     </td><td style=\"text-align: right;\">           20   </td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">               21  </td><td style=\"text-align: right;\">              37</td><td>GOOD      </td><td style=\"text-align: right;\">                     15</td></tr>\n",
       "<tr><td style=\"text-align: right;\">      5</td><td style=\"text-align: right;\">         5000</td><td>36 months</td><td style=\"text-align: right;\">           7.9 </td><td style=\"text-align: right;\">               3  </td><td>RENT            </td><td style=\"text-align: right;\">          36000</td><td>VERIFIED - income source</td><td>wedding       </td><td>AZ     </td><td style=\"text-align: right;\">           11.2 </td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">               28.3</td><td style=\"text-align: right;\">              12</td><td>GOOD      </td><td style=\"text-align: right;\">                      7</td></tr>\n",
       "<tr><td style=\"text-align: right;\">      6</td><td style=\"text-align: right;\">         3000</td><td>36 months</td><td style=\"text-align: right;\">          18.64</td><td style=\"text-align: right;\">               9  </td><td>RENT            </td><td style=\"text-align: right;\">          48000</td><td>VERIFIED - income source</td><td>car           </td><td>CA     </td><td style=\"text-align: right;\">            5.35</td><td style=\"text-align: right;\">               0</td><td style=\"text-align: right;\">               87.5</td><td style=\"text-align: right;\">               4</td><td>GOOD      </td><td style=\"text-align: right;\">                      4</td></tr>\n",
       "</tbody>\n",
       "</table>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": []
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Identify predictors and response\n",
    "x = train.columns\n",
    "y = \"Bad_Loan\"\n",
    "x.remove(y)\n",
    "\n",
    "# For binary classification, response should be a factor\n",
    "train[y] = train[y].asfactor()\n",
    "test[y] = test[y].asfactor()\n",
    "train.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Train an XGBoost Model using GPU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy AUC: 0.874694485388\n",
      "CPU times: user 67.8 ms, sys: 8.89 ms, total: 76.7 ms\n",
      "Wall time: 11.5 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "XGB_GPU = H2OXGBoostEstimator(model_id=\"XGB_on_GPU\", ntrees=200, max_depth=9, learn_rate=0.05, backend=\"gpu\", gpu_id=0)\n",
    "XGB_GPU.train(x=x, y=y, training_frame=train, validation_frame=test)\n",
    "print \"Accuracy AUC: \" + str(XGB_GPU.auc())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Train an XGBoost Model using CPU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy AUC: 0.859213997196\n",
      "CPU times: user 226 ms, sys: 21.3 ms, total: 247 ms\n",
      "Wall time: 1min 24s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "XGB_CPU = H2OXGBoostEstimator(model_id=\"XGB_on_CPU\", ntrees=200, max_depth=9, learn_rate=0.05, backend=\"cpu\")\n",
    "XGB_CPU.train(x=x, y=y, training_frame=train, validation_frame=test)\n",
    "print \"Accuracy AUC: \" + str(XGB_CPU.auc())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7. Which Option is Faster? Which Option Costs Less?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fddac35f2d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt; plt.rcdefaults()\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    " \n",
    "objects = ('GPU', 'CPU')\n",
    "y_pos = np.arange(len(objects))\n",
    "seconds = [11,84]\n",
    " \n",
    "plt.barh(y_pos, seconds, align='center', alpha=0.5)\n",
    "plt.yticks(y_pos, objects)\n",
    "plt.xlabel('Seconds')\n",
    "plt.title('XGBoost Training Time in Seconds')\n",
    " \n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fdda78d0c10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "GPU_Cost = (0.9 / 3600) * 11\n",
    "CPU_Cost = (0.7 / 3600) * 84\n",
    "objects = ('GPU Cost', 'CPU Cost')\n",
    "y_pos = np.arange(len(objects))\n",
    "seconds = [GPU_Cost,CPU_Cost]\n",
    " \n",
    "plt.barh(y_pos, seconds, align='center', alpha=0.5)\n",
    "plt.yticks(y_pos, objects)\n",
    "plt.xlabel('Cents')\n",
    "plt.title('Cost in Cents')\n",
    " \n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}