notebooks/LGBM_year_with_train.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### If you haven't installed Hummingbird, do that first, by uncommenting the line below.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#! python -m pip install hummingbird_ml "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Import Hummingbird's convert function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from hummingbird.ml import convert"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Download and unzip the Year data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import zipfile\n",
    "import urllib.request as urllib\n",
    "url = 'https://archive.ics.uci.edu/ml/machine-learning-databases/00203/YearPredictionMSD.txt.zip'\n",
    "\n",
    "filehandle, _ = urllib.urlretrieve(url)\n",
    "zip_file_object = zipfile.ZipFile(filehandle, 'r')\n",
    "filename = zip_file_object.namelist()[0]\n",
    "bytes_data = zip_file_object.open(filename).read()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Convert the bytes to pandas and split it up"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from io import BytesIO\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "year = pd.read_csv(BytesIO(bytes_data), header = None)\n",
    "\n",
    "X = year.iloc[:, 1:]\n",
    "y = year.iloc[:, 0]\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, shuffle=False, train_size=463715, test_size=51630)\n",
    "\n",
    "# Store the test data as numpy by pulling the values out of the pandas dataframe\n",
    "data = X_test.values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>...</th>\n",
       "      <th>81</th>\n",
       "      <th>82</th>\n",
       "      <th>83</th>\n",
       "      <th>84</th>\n",
       "      <th>85</th>\n",
       "      <th>86</th>\n",
       "      <th>87</th>\n",
       "      <th>88</th>\n",
       "      <th>89</th>\n",
       "      <th>90</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>463715</th>\n",
       "      <td>45.44200</td>\n",
       "      <td>-30.74976</td>\n",
       "      <td>31.78587</td>\n",
       "      <td>4.63569</td>\n",
       "      <td>-15.14894</td>\n",
       "      <td>0.23370</td>\n",
       "      <td>-11.97968</td>\n",
       "      <td>-9.59708</td>\n",
       "      <td>6.48111</td>\n",
       "      <td>-8.89073</td>\n",
       "      <td>...</td>\n",
       "      <td>-8.84046</td>\n",
       "      <td>-0.15439</td>\n",
       "      <td>137.44210</td>\n",
       "      <td>77.54739</td>\n",
       "      <td>-4.22875</td>\n",
       "      <td>-61.92657</td>\n",
       "      <td>-33.52722</td>\n",
       "      <td>-3.86253</td>\n",
       "      <td>36.42400</td>\n",
       "      <td>7.17309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463716</th>\n",
       "      <td>52.67814</td>\n",
       "      <td>-2.88914</td>\n",
       "      <td>43.95268</td>\n",
       "      <td>-1.39209</td>\n",
       "      <td>-14.93379</td>\n",
       "      <td>-15.86877</td>\n",
       "      <td>1.19379</td>\n",
       "      <td>0.31401</td>\n",
       "      <td>-4.44235</td>\n",
       "      <td>-5.78934</td>\n",
       "      <td>...</td>\n",
       "      <td>-5.74356</td>\n",
       "      <td>-42.57910</td>\n",
       "      <td>-2.91103</td>\n",
       "      <td>48.72805</td>\n",
       "      <td>-3.08183</td>\n",
       "      <td>-9.38888</td>\n",
       "      <td>-7.27179</td>\n",
       "      <td>-4.00966</td>\n",
       "      <td>-68.96211</td>\n",
       "      <td>-5.21525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463717</th>\n",
       "      <td>45.74235</td>\n",
       "      <td>12.02291</td>\n",
       "      <td>11.03009</td>\n",
       "      <td>-11.60763</td>\n",
       "      <td>11.80054</td>\n",
       "      <td>-11.12389</td>\n",
       "      <td>-5.39058</td>\n",
       "      <td>-1.11981</td>\n",
       "      <td>-7.74086</td>\n",
       "      <td>-3.33421</td>\n",
       "      <td>...</td>\n",
       "      <td>-4.70606</td>\n",
       "      <td>-24.22599</td>\n",
       "      <td>-35.22686</td>\n",
       "      <td>27.77729</td>\n",
       "      <td>15.38934</td>\n",
       "      <td>58.20036</td>\n",
       "      <td>-61.12698</td>\n",
       "      <td>-10.92522</td>\n",
       "      <td>26.75348</td>\n",
       "      <td>-5.78743</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463718</th>\n",
       "      <td>52.55883</td>\n",
       "      <td>2.87222</td>\n",
       "      <td>27.38848</td>\n",
       "      <td>-5.76235</td>\n",
       "      <td>-15.35766</td>\n",
       "      <td>-15.01592</td>\n",
       "      <td>-5.86893</td>\n",
       "      <td>-0.31447</td>\n",
       "      <td>-5.06922</td>\n",
       "      <td>-4.62734</td>\n",
       "      <td>...</td>\n",
       "      <td>-8.35215</td>\n",
       "      <td>-16.86791</td>\n",
       "      <td>-10.58277</td>\n",
       "      <td>40.10173</td>\n",
       "      <td>-0.54005</td>\n",
       "      <td>-11.54746</td>\n",
       "      <td>-45.35860</td>\n",
       "      <td>-4.55694</td>\n",
       "      <td>-43.17368</td>\n",
       "      <td>-3.33725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463719</th>\n",
       "      <td>51.34809</td>\n",
       "      <td>9.02702</td>\n",
       "      <td>25.33757</td>\n",
       "      <td>-6.62537</td>\n",
       "      <td>0.03367</td>\n",
       "      <td>-12.69565</td>\n",
       "      <td>-3.13400</td>\n",
       "      <td>2.98649</td>\n",
       "      <td>-6.71750</td>\n",
       "      <td>-1.85804</td>\n",
       "      <td>...</td>\n",
       "      <td>-6.87366</td>\n",
       "      <td>-20.03371</td>\n",
       "      <td>-66.38940</td>\n",
       "      <td>50.56569</td>\n",
       "      <td>0.27747</td>\n",
       "      <td>67.05657</td>\n",
       "      <td>-55.58846</td>\n",
       "      <td>-7.50859</td>\n",
       "      <td>28.23511</td>\n",
       "      <td>-0.72045</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 90 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "              1         2         3         4         5         6         7   \\\n",
       "463715  45.44200 -30.74976  31.78587   4.63569 -15.14894   0.23370 -11.97968   \n",
       "463716  52.67814  -2.88914  43.95268  -1.39209 -14.93379 -15.86877   1.19379   \n",
       "463717  45.74235  12.02291  11.03009 -11.60763  11.80054 -11.12389  -5.39058   \n",
       "463718  52.55883   2.87222  27.38848  -5.76235 -15.35766 -15.01592  -5.86893   \n",
       "463719  51.34809   9.02702  25.33757  -6.62537   0.03367 -12.69565  -3.13400   \n",
       "\n",
       "             8        9        10  ...       81        82         83  \\\n",
       "463715 -9.59708  6.48111 -8.89073  ... -8.84046  -0.15439  137.44210   \n",
       "463716  0.31401 -4.44235 -5.78934  ... -5.74356 -42.57910   -2.91103   \n",
       "463717 -1.11981 -7.74086 -3.33421  ... -4.70606 -24.22599  -35.22686   \n",
       "463718 -0.31447 -5.06922 -4.62734  ... -8.35215 -16.86791  -10.58277   \n",
       "463719  2.98649 -6.71750 -1.85804  ... -6.87366 -20.03371  -66.38940   \n",
       "\n",
       "              84        85        86        87        88        89       90  \n",
       "463715  77.54739  -4.22875 -61.92657 -33.52722  -3.86253  36.42400  7.17309  \n",
       "463716  48.72805  -3.08183  -9.38888  -7.27179  -4.00966 -68.96211 -5.21525  \n",
       "463717  27.77729  15.38934  58.20036 -61.12698 -10.92522  26.75348 -5.78743  \n",
       "463718  40.10173  -0.54005 -11.54746 -45.35860  -4.55694 -43.17368 -3.33725  \n",
       "463719  50.56569   0.27747  67.05657 -55.58846  -7.50859  28.23511 -0.72045  \n",
       "\n",
       "[5 rows x 90 columns]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# peak at the data if desired\n",
    "X_test.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Train the model  (Note, this may take a bit of time. )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LGBMRegressor(boosting_type='gbdt', class_weight=None, colsample_bytree=1.0,\n",
       "              importance_type='split', learning_rate=0.1, max_depth=8,\n",
       "              min_child_samples=20, min_child_weight=0.001, min_split_gain=0.0,\n",
       "              n_estimators=500, n_jobs=6, num_leaves=256,\n",
       "              objective='regression', random_state=None, reg_alpha=0.0,\n",
       "              reg_lambda=1, silent=True, subsample=1.0,\n",
       "              subsample_for_bin=200000, subsample_freq=0)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import lightgbm as lgb\n",
    "\n",
    "lgbm_model = lgb.LGBMRegressor(max_depth=8,n_estimators=500, num_leaves=256, objective=\"regression\", reg_lambda=1, n_jobs=6)\n",
    "\n",
    "lgbm_model.fit(X_train, y_train)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### LGBM (CPU only)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "844 ms ± 6.59 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
     ]
    }
   ],
   "source": [
    "lgbm_time = %timeit -o lgbm_model.predict(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Convert LGBM model to PyTorch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Finished loading model, total used 500 iterations\n"
     ]
    }
   ],
   "source": [
    "model = convert(lgbm_model, 'torch')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Time PyTorch - CPU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1.78 s ± 17 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
     ]
    }
   ],
   "source": [
    "pred_cpu_hb = %timeit -o model.predict(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(51630, 90)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.values.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Switch PyTorch from CPU to GPU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture \n",
    "model.to('cuda')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Time PyTorch - GPU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "50 ms ± 243 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n"
     ]
    }
   ],
   "source": [
    "pred_gpu_hb = %timeit -o model.predict(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Note: Continuing in this notebook requires TVM built with LLVM support. Install instructions [here](https://tvm.apache.org/docs/install/index.html)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Convert LGBM model to TVM (CPU)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Finished loading model, total used 500 iterations\n"
     ]
    }
   ],
   "source": [
    "model_tvm = convert(lgbm_model, 'tvm', data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Time TVM - CPU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "803 ms ± 8.4 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)\n"
     ]
    }
   ],
   "source": [
    "pred_cpu_tvm = %timeit -o model_tvm.predict(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Convert LBGM model to TVM (GPU)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Finished loading model, total used 500 iterations\n"
     ]
    }
   ],
   "source": [
    "model_tvm = convert(lgbm_model, 'tvm', data, 'cuda')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Time TVM - GPU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "31.1 ms ± 402 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)\n"
     ]
    }
   ],
   "source": [
    "pred_gpu_tvm = %timeit -o model_tvm.predict(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Plot the results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot(title, lgbm_time, pred_cpu_hb, pred_gpu_hb, pred_cpu_tvm, pred_gpu_tvm):\n",
    "    import matplotlib.pyplot as plt\n",
    "    import numpy as np\n",
    "    from matplotlib.pyplot import cm\n",
    "\n",
    "    fig = plt.figure()\n",
    "\n",
    "    x = ['lgbm','pyt-cpu','pyt-gpu','tvm-cpu','tvm-gpu']\n",
    "    height = [lgbm_time.average,pred_cpu_hb.average,pred_gpu_hb.average,pred_cpu_tvm.average,pred_gpu_tvm.average]\n",
    "    width = 1.0\n",
    "    plt.ylabel('time in seconds')\n",
    "    plt.xlabel(title)\n",
    "\n",
    "    rects = plt.bar(x, height, width, color=cm.rainbow(np.linspace(0,1,5)))\n",
    "    def autolabel(rects):\n",
    "\n",
    "        for rect in rects:\n",
    "            height = rect.get_height()\n",
    "            plt.text(rect.get_x() + rect.get_width()/2., 1.05*height,\n",
    "                    '%.4f' % (height),\n",
    "                    ha='center', va='bottom')\n",
    "\n",
    "    autolabel(rects)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot(\"LGBM (500) - Year Dataset\", lgbm_time, pred_cpu_hb, pred_gpu_hb, pred_cpu_tvm, pred_gpu_tvm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}