NUTRITION_MODEL_TRAINING.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\mindf\\anaconda3\\lib\\site-packages\\xgboost\\compat.py:36: FutureWarning: pandas.Int64Index is deprecated and will be removed from pandas in a future version. Use pandas.Index with the appropriate dtype instead.\n",
      "  from pandas import MultiIndex, Int64Index\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import xgboost as xgb\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Load csv\n",
    "NUTRITION_MYFITNESS_DATA = pd.read_excel(r'DATASETS\\NUTRITION\\NUTRITION_DATASET.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<bound method NDFrame.head of     CALORIES  CARBOHYDRATES  FAT  PROTEIN  SODIUM  SUGAR  SCORE\n",
       "0       1423            172   61       56    2376     25      0\n",
       "1       2001            343   49       57    2210    145      0\n",
       "2       2001            223   92       67    2112    108      0\n",
       "3       1827            256   62       65    2456    108      0\n",
       "4       2001            250   56      125    2300     81      0\n",
       "5       1355            167   56       44    1266     93      0\n",
       "6       1000            149   27       39    2381     24      0\n",
       "7       1772            288   60       57    1702     79      0\n",
       "8       1975            335   51       60    2050    110      0\n",
       "9       1800            368   60       59    3244    148      0\n",
       "10      1185            132   59       48    1604     56      0\n",
       "11      1879            191   80       62    1920     61      0\n",
       "12       778            105   35       36     795     38      0\n",
       "13      1036            164   34       23    2101     42      0\n",
       "14       998            178   14       54     787     43      0\n",
       "15      1271            179   30       69    2291     25      0\n",
       "16      1575            240   45       49    2230     37      0\n",
       "17       760             98   18       49     735     14      0>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "NUTRITION_MYFITNESS_DATA.head"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " =================================== \n",
      "1423 2250\n",
      "827\n",
      "172 275\n",
      "103\n",
      "61 61\n",
      "0\n",
      "56 112.5\n",
      "56.5\n",
      "2376 2300\n",
      "76\n",
      "25 125\n",
      "100\n",
      "END SCORE:  193.75\n",
      " =================================== \n",
      "2001 2250\n",
      "249\n",
      "343 275\n",
      "68\n",
      "49 61\n",
      "12\n",
      "57 112.5\n",
      "55.5\n",
      "2210 2300\n",
      "90\n",
      "145 125\n",
      "20\n",
      "END SCORE:  82.41666666666667\n",
      " =================================== \n",
      "2001 2250\n",
      "249\n",
      "223 275\n",
      "52\n",
      "92 61\n",
      "31\n",
      "67 112.5\n",
      "45.5\n",
      "2112 2300\n",
      "188\n",
      "108 125\n",
      "17\n",
      "END SCORE:  97.08333333333333\n",
      " =================================== \n",
      "1827 2250\n",
      "423\n",
      "256 275\n",
      "19\n",
      "62 61\n",
      "1\n",
      "65 112.5\n",
      "47.5\n",
      "2456 2300\n",
      "156\n",
      "108 125\n",
      "17\n",
      "END SCORE:  110.58333333333333\n",
      " =================================== \n",
      "2001 2250\n",
      "249\n",
      "250 275\n",
      "25\n",
      "56 61\n",
      "5\n",
      "125 112.5\n",
      "12.5\n",
      "2300 2300\n",
      "0\n",
      "81 125\n",
      "44\n",
      "END SCORE:  55.916666666666664\n",
      " =================================== \n",
      "1355 2250\n",
      "895\n",
      "167 275\n",
      "108\n",
      "56 61\n",
      "5\n",
      "44 112.5\n",
      "68.5\n",
      "1266 2300\n",
      "1034\n",
      "93 125\n",
      "32\n",
      "END SCORE:  357.0833333333333\n",
      " =================================== \n",
      "1000 2250\n",
      "1250\n",
      "149 275\n",
      "126\n",
      "27 61\n",
      "34\n",
      "39 112.5\n",
      "73.5\n",
      "2381 2300\n",
      "81\n",
      "24 125\n",
      "101\n",
      "END SCORE:  277.5833333333333\n",
      " =================================== \n",
      "1772 2250\n",
      "478\n",
      "288 275\n",
      "13\n",
      "60 61\n",
      "1\n",
      "57 112.5\n",
      "55.5\n",
      "1702 2300\n",
      "598\n",
      "79 125\n",
      "46\n",
      "END SCORE:  198.58333333333334\n",
      " =================================== \n",
      "1975 2250\n",
      "275\n",
      "335 275\n",
      "60\n",
      "51 61\n",
      "10\n",
      "60 112.5\n",
      "52.5\n",
      "2050 2300\n",
      "250\n",
      "110 125\n",
      "15\n",
      "END SCORE:  110.41666666666667\n",
      " =================================== \n",
      "1800 2250\n",
      "450\n",
      "368 275\n",
      "93\n",
      "60 61\n",
      "1\n",
      "59 112.5\n",
      "53.5\n",
      "3244 2300\n",
      "944\n",
      "148 125\n",
      "23\n",
      "END SCORE:  260.75\n",
      " =================================== \n",
      "1185 2250\n",
      "1065\n",
      "132 275\n",
      "143\n",
      "59 61\n",
      "2\n",
      "48 112.5\n",
      "64.5\n",
      "1604 2300\n",
      "696\n",
      "56 125\n",
      "69\n",
      "END SCORE:  339.9166666666667\n",
      " =================================== \n",
      "1879 2250\n",
      "371\n",
      "191 275\n",
      "84\n",
      "80 61\n",
      "19\n",
      "62 112.5\n",
      "50.5\n",
      "1920 2300\n",
      "380\n",
      "61 125\n",
      "64\n",
      "END SCORE:  161.41666666666666\n",
      " =================================== \n",
      "778 2250\n",
      "1472\n",
      "105 275\n",
      "170\n",
      "35 61\n",
      "26\n",
      "36 112.5\n",
      "76.5\n",
      "795 2300\n",
      "1505\n",
      "38 125\n",
      "87\n",
      "END SCORE:  556.0833333333334\n",
      " =================================== \n",
      "1036 2250\n",
      "1214\n",
      "164 275\n",
      "111\n",
      "34 61\n",
      "27\n",
      "23 112.5\n",
      "89.5\n",
      "2101 2300\n",
      "199\n",
      "42 125\n",
      "83\n",
      "END SCORE:  287.25\n",
      " =================================== \n",
      "998 2250\n",
      "1252\n",
      "178 275\n",
      "97\n",
      "14 61\n",
      "47\n",
      "54 112.5\n",
      "58.5\n",
      "787 2300\n",
      "1513\n",
      "43 125\n",
      "82\n",
      "END SCORE:  508.25\n",
      " =================================== \n",
      "1271 2250\n",
      "979\n",
      "179 275\n",
      "96\n",
      "30 61\n",
      "31\n",
      "69 112.5\n",
      "43.5\n",
      "2291 2300\n",
      "9\n",
      "25 125\n",
      "100\n",
      "END SCORE:  209.75\n",
      " =================================== \n",
      "1575 2250\n",
      "675\n",
      "240 275\n",
      "35\n",
      "45 61\n",
      "16\n",
      "49 112.5\n",
      "63.5\n",
      "2230 2300\n",
      "70\n",
      "37 125\n",
      "88\n",
      "END SCORE:  157.91666666666666\n",
      " =================================== \n",
      "760 2250\n",
      "1490\n",
      "98 275\n",
      "177\n",
      "18 61\n",
      "43\n",
      "49 112.5\n",
      "63.5\n",
      "735 2300\n",
      "1565\n",
      "14 125\n",
      "111\n",
      "END SCORE:  574.9166666666666\n"
     ]
    }
   ],
   "source": [
    "#Create dictionary of ideal values for each score; calculate L2 norm to generate score)\n",
    "IDEAL_VALS = {'CALORIES': 2250, 'CARBOHYDRATES': 275, 'FAT': 61, 'PROTEIN': 112.5,\n",
    "              'SODIUM': 2300, 'SUGAR': 125}\n",
    "columns = NUTRITION_MYFITNESS_DATA.columns.values.tolist()\n",
    "columns = columns[:6]\n",
    "#Generate scores for each lever - compare iteratively and generate average score on its basis\n",
    "def generate_score(row):\n",
    "    #Iterate over each column except for gender\n",
    "    print(\" =================================== \")\n",
    "    total_score = 0\n",
    "    for column in columns:\n",
    "        if column == \"SCORE\": continue\n",
    "        print(row[column], IDEAL_VALS[column])\n",
    "        column_score = abs(row[column] - IDEAL_VALS[column])\n",
    "        print(column_score)\n",
    "        total_score += column_score\n",
    "    #Average out the score\n",
    "    row[\"SCORE\"] = total_score / len(columns)\n",
    "    print('END SCORE: ', row[\"SCORE\"])\n",
    "    return row\n",
    "\n",
    "NUTRITION_MYFITNESS_DATA = NUTRITION_MYFITNESS_DATA.apply(generate_score, axis = \"columns\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    CALORIES  CARBOHYDRATES   FAT  PROTEIN  SODIUM  SUGAR       SCORE\n",
      "0     1423.0          172.0  61.0     56.0  2376.0   25.0  193.750000\n",
      "1     2001.0          343.0  49.0     57.0  2210.0  145.0   82.416667\n",
      "2     2001.0          223.0  92.0     67.0  2112.0  108.0   97.083333\n",
      "3     1827.0          256.0  62.0     65.0  2456.0  108.0  110.583333\n",
      "4     2001.0          250.0  56.0    125.0  2300.0   81.0   55.916667\n",
      "5     1355.0          167.0  56.0     44.0  1266.0   93.0  357.083333\n",
      "6     1000.0          149.0  27.0     39.0  2381.0   24.0  277.583333\n",
      "7     1772.0          288.0  60.0     57.0  1702.0   79.0  198.583333\n",
      "8     1975.0          335.0  51.0     60.0  2050.0  110.0  110.416667\n",
      "9     1800.0          368.0  60.0     59.0  3244.0  148.0  260.750000\n",
      "10    1185.0          132.0  59.0     48.0  1604.0   56.0  339.916667\n",
      "11    1879.0          191.0  80.0     62.0  1920.0   61.0  161.416667\n",
      "12     778.0          105.0  35.0     36.0   795.0   38.0  556.083333\n",
      "13    1036.0          164.0  34.0     23.0  2101.0   42.0  287.250000\n",
      "14     998.0          178.0  14.0     54.0   787.0   43.0  508.250000\n",
      "15    1271.0          179.0  30.0     69.0  2291.0   25.0  209.750000\n",
      "16    1575.0          240.0  45.0     49.0  2230.0   37.0  157.916667\n",
      "17     760.0           98.0  18.0     49.0   735.0   14.0  574.916667\n"
     ]
    }
   ],
   "source": [
    "print(NUTRITION_MYFITNESS_DATA)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<bound method NDFrame.head of     CALORIES  CARBOHYDRATES   FAT  PROTEIN  SODIUM  SUGAR       SCORE\n",
      "0     1423.0          172.0  61.0     56.0  2376.0   25.0  193.750000\n",
      "1     2001.0          343.0  49.0     57.0  2210.0  145.0   82.416667\n",
      "2     2001.0          223.0  92.0     67.0  2112.0  108.0   97.083333\n",
      "3     1827.0          256.0  62.0     65.0  2456.0  108.0  110.583333\n",
      "4     2001.0          250.0  56.0    125.0  2300.0   81.0   55.916667\n",
      "5     1355.0          167.0  56.0     44.0  1266.0   93.0  357.083333\n",
      "6     1000.0          149.0  27.0     39.0  2381.0   24.0  277.583333\n",
      "7     1772.0          288.0  60.0     57.0  1702.0   79.0  198.583333\n",
      "8     1975.0          335.0  51.0     60.0  2050.0  110.0  110.416667\n",
      "9     1800.0          368.0  60.0     59.0  3244.0  148.0  260.750000\n",
      "10    1185.0          132.0  59.0     48.0  1604.0   56.0  339.916667\n",
      "11    1879.0          191.0  80.0     62.0  1920.0   61.0  161.416667\n",
      "12     778.0          105.0  35.0     36.0   795.0   38.0  556.083333\n",
      "13    1036.0          164.0  34.0     23.0  2101.0   42.0  287.250000\n",
      "14     998.0          178.0  14.0     54.0   787.0   43.0  508.250000\n",
      "15    1271.0          179.0  30.0     69.0  2291.0   25.0  209.750000\n",
      "16    1575.0          240.0  45.0     49.0  2230.0   37.0  157.916667\n",
      "17     760.0           98.0  18.0     49.0   735.0   14.0  574.916667>\n"
     ]
    }
   ],
   "source": [
    "print(NUTRITION_MYFITNESS_DATA.head)\n",
    "NUTRITION_MYFITNESS_DATA.to_pickle('PREPROCESSED_NUTRITION_MYFITNESS_DATA.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    CALORIES  CARBOHYDRATES   FAT  PROTEIN  SODIUM  SUGAR\n",
      "0     1423.0          172.0  61.0     56.0  2376.0   25.0\n",
      "1     2001.0          343.0  49.0     57.0  2210.0  145.0\n",
      "2     2001.0          223.0  92.0     67.0  2112.0  108.0\n",
      "3     1827.0          256.0  62.0     65.0  2456.0  108.0\n",
      "4     2001.0          250.0  56.0    125.0  2300.0   81.0\n",
      "5     1355.0          167.0  56.0     44.0  1266.0   93.0\n",
      "6     1000.0          149.0  27.0     39.0  2381.0   24.0\n",
      "7     1772.0          288.0  60.0     57.0  1702.0   79.0\n",
      "8     1975.0          335.0  51.0     60.0  2050.0  110.0\n",
      "9     1800.0          368.0  60.0     59.0  3244.0  148.0\n",
      "10    1185.0          132.0  59.0     48.0  1604.0   56.0\n",
      "11    1879.0          191.0  80.0     62.0  1920.0   61.0\n",
      "12     778.0          105.0  35.0     36.0   795.0   38.0\n",
      "13    1036.0          164.0  34.0     23.0  2101.0   42.0\n",
      "14     998.0          178.0  14.0     54.0   787.0   43.0\n",
      "15    1271.0          179.0  30.0     69.0  2291.0   25.0\n",
      "16    1575.0          240.0  45.0     49.0  2230.0   37.0\n",
      "17     760.0           98.0  18.0     49.0   735.0   14.0\n"
     ]
    }
   ],
   "source": [
    "#Reshape tensors for XGBOOST\n",
    "import tensorflow as tf\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "x_features = NUTRITION_MYFITNESS_DATA.iloc[:, :6]\n",
    "print(x_features)\n",
    "y_labels = NUTRITION_MYFITNESS_DATA[\"SCORE\"]\n",
    "train_x, valid_x, train_y, valid_y = train_test_split(x_features, y_labels, random_state = 2, shuffle = True)\n",
    "\n",
    "train_x = tf.convert_to_tensor(train_x)\n",
    "train_x = tf.reshape(train_x, [len(train_x), 6])\n",
    "\n",
    "valid_x = tf.convert_to_tensor(valid_x)\n",
    "valid_x = tf.reshape(valid_x, [len(valid_x), 6])\n",
    "\n",
    "train_y = tf.convert_to_tensor(train_y)\n",
    "train_y = tf.reshape(train_y, [len(train_y), 1])\n",
    "\n",
    "valid_y = tf.convert_to_tensor(valid_y)\n",
    "valid_y = tf.reshape(valid_y, [len(valid_y), 1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[17:52:04] WARNING: C:\\Windows\\Temp\\abs_557yfx631l\\croots\\recipe\\xgboost-split_1659548953302\\work\\src\\learner.cc:576: \n",
      "Parameters: { \"n_estimators\" } might not be used.\n",
      "\n",
      "  This could be a false alarm, with some parameters getting used by language bindings but\n",
      "  then being mistakenly passed down to XGBoost core, or some parameter actually being used\n",
      "  but getting flagged wrongly here. Please open an issue if you find any such cases.\n",
      "\n",
      "\n",
      "[0]\tTRAIN_DATA-rmse:297.20090\tVALID_DATA-rmse:202.51248\n",
      "[1]\tTRAIN_DATA-rmse:283.40820\tVALID_DATA-rmse:187.21147\n",
      "[2]\tTRAIN_DATA-rmse:265.71304\tVALID_DATA-rmse:172.67262\n",
      "[3]\tTRAIN_DATA-rmse:246.44974\tVALID_DATA-rmse:160.54864\n",
      "[4]\tTRAIN_DATA-rmse:227.82654\tVALID_DATA-rmse:152.33933\n",
      "[5]\tTRAIN_DATA-rmse:210.52936\tVALID_DATA-rmse:145.16296\n",
      "[6]\tTRAIN_DATA-rmse:196.65736\tVALID_DATA-rmse:134.63860\n",
      "[7]\tTRAIN_DATA-rmse:184.08818\tVALID_DATA-rmse:125.31136\n",
      "[8]\tTRAIN_DATA-rmse:179.93411\tVALID_DATA-rmse:121.27068\n",
      "[9]\tTRAIN_DATA-rmse:174.71794\tVALID_DATA-rmse:113.86499\n",
      "[10]\tTRAIN_DATA-rmse:164.03902\tVALID_DATA-rmse:106.86322\n",
      "[11]\tTRAIN_DATA-rmse:155.55806\tVALID_DATA-rmse:103.62668\n",
      "[12]\tTRAIN_DATA-rmse:147.32350\tVALID_DATA-rmse:100.57404\n",
      "[13]\tTRAIN_DATA-rmse:140.81805\tVALID_DATA-rmse:94.27182\n",
      "[14]\tTRAIN_DATA-rmse:131.77872\tVALID_DATA-rmse:93.62853\n",
      "[15]\tTRAIN_DATA-rmse:125.56552\tVALID_DATA-rmse:91.48016\n",
      "[16]\tTRAIN_DATA-rmse:120.17564\tVALID_DATA-rmse:89.23794\n",
      "[17]\tTRAIN_DATA-rmse:117.73700\tVALID_DATA-rmse:87.22914\n",
      "[18]\tTRAIN_DATA-rmse:110.41700\tVALID_DATA-rmse:86.85922\n",
      "[19]\tTRAIN_DATA-rmse:105.61785\tVALID_DATA-rmse:84.86755\n",
      "[20]\tTRAIN_DATA-rmse:99.10721\tVALID_DATA-rmse:84.80103\n",
      "[21]\tTRAIN_DATA-rmse:94.44423\tVALID_DATA-rmse:84.05016\n",
      "[22]\tTRAIN_DATA-rmse:88.79846\tVALID_DATA-rmse:83.74805\n",
      "[23]\tTRAIN_DATA-rmse:85.54735\tVALID_DATA-rmse:82.90606\n",
      "[24]\tTRAIN_DATA-rmse:80.61731\tVALID_DATA-rmse:81.43794\n",
      "[25]\tTRAIN_DATA-rmse:77.99147\tVALID_DATA-rmse:80.83730\n",
      "[26]\tTRAIN_DATA-rmse:75.09465\tVALID_DATA-rmse:79.90430\n",
      "[27]\tTRAIN_DATA-rmse:71.00880\tVALID_DATA-rmse:79.68434\n",
      "[28]\tTRAIN_DATA-rmse:70.88951\tVALID_DATA-rmse:79.08846\n",
      "[29]\tTRAIN_DATA-rmse:66.95009\tVALID_DATA-rmse:78.70700\n",
      "[30]\tTRAIN_DATA-rmse:62.36389\tVALID_DATA-rmse:79.33321\n",
      "[31]\tTRAIN_DATA-rmse:59.71778\tVALID_DATA-rmse:79.27235\n",
      "[32]\tTRAIN_DATA-rmse:57.26330\tVALID_DATA-rmse:78.46295\n",
      "[33]\tTRAIN_DATA-rmse:54.26033\tVALID_DATA-rmse:78.56889\n",
      "[34]\tTRAIN_DATA-rmse:51.45464\tVALID_DATA-rmse:78.44868\n",
      "[35]\tTRAIN_DATA-rmse:49.44253\tVALID_DATA-rmse:78.94871\n",
      "[36]\tTRAIN_DATA-rmse:47.35149\tVALID_DATA-rmse:78.76566\n",
      "[37]\tTRAIN_DATA-rmse:47.17563\tVALID_DATA-rmse:78.76566\n",
      "[38]\tTRAIN_DATA-rmse:45.87542\tVALID_DATA-rmse:78.89365\n",
      "[39]\tTRAIN_DATA-rmse:42.95526\tVALID_DATA-rmse:78.90099\n",
      "[40]\tTRAIN_DATA-rmse:43.09212\tVALID_DATA-rmse:78.76427\n",
      "[41]\tTRAIN_DATA-rmse:41.54624\tVALID_DATA-rmse:78.16225\n",
      "[42]\tTRAIN_DATA-rmse:39.49706\tVALID_DATA-rmse:78.13058\n",
      "[43]\tTRAIN_DATA-rmse:37.51750\tVALID_DATA-rmse:78.11154\n",
      "[44]\tTRAIN_DATA-rmse:36.23056\tVALID_DATA-rmse:77.43577\n",
      "[45]\tTRAIN_DATA-rmse:34.31623\tVALID_DATA-rmse:77.27669\n",
      "[46]\tTRAIN_DATA-rmse:33.41976\tVALID_DATA-rmse:77.27669\n",
      "[47]\tTRAIN_DATA-rmse:32.09063\tVALID_DATA-rmse:77.27669\n",
      "[48]\tTRAIN_DATA-rmse:30.47965\tVALID_DATA-rmse:76.95940\n",
      "[49]\tTRAIN_DATA-rmse:29.36261\tVALID_DATA-rmse:76.68198\n",
      "[50]\tTRAIN_DATA-rmse:28.55513\tVALID_DATA-rmse:76.92211\n",
      "[51]\tTRAIN_DATA-rmse:28.55513\tVALID_DATA-rmse:76.92211\n",
      "[52]\tTRAIN_DATA-rmse:27.37699\tVALID_DATA-rmse:76.62343\n",
      "[53]\tTRAIN_DATA-rmse:26.43834\tVALID_DATA-rmse:76.38386\n",
      "[54]\tTRAIN_DATA-rmse:25.10372\tVALID_DATA-rmse:76.36006\n",
      "[55]\tTRAIN_DATA-rmse:23.89573\tVALID_DATA-rmse:76.36755\n",
      "[56]\tTRAIN_DATA-rmse:23.76930\tVALID_DATA-rmse:76.55351\n",
      "[57]\tTRAIN_DATA-rmse:22.63180\tVALID_DATA-rmse:76.56319\n",
      "[58]\tTRAIN_DATA-rmse:21.44319\tVALID_DATA-rmse:76.61402\n",
      "[59]\tTRAIN_DATA-rmse:20.78621\tVALID_DATA-rmse:76.72518\n",
      "[60]\tTRAIN_DATA-rmse:19.74490\tVALID_DATA-rmse:76.74592\n",
      "[61]\tTRAIN_DATA-rmse:19.04459\tVALID_DATA-rmse:76.56432\n",
      "[62]\tTRAIN_DATA-rmse:18.52350\tVALID_DATA-rmse:76.56432\n",
      "[63]\tTRAIN_DATA-rmse:18.04791\tVALID_DATA-rmse:76.36581\n",
      "[64]\tTRAIN_DATA-rmse:17.73651\tVALID_DATA-rmse:76.38222\n",
      "[65]\tTRAIN_DATA-rmse:17.41191\tVALID_DATA-rmse:76.35093\n",
      "[66]\tTRAIN_DATA-rmse:16.87620\tVALID_DATA-rmse:76.22350\n",
      "[67]\tTRAIN_DATA-rmse:17.00949\tVALID_DATA-rmse:76.23705\n",
      "[68]\tTRAIN_DATA-rmse:16.85001\tVALID_DATA-rmse:76.44459\n",
      "[69]\tTRAIN_DATA-rmse:16.27712\tVALID_DATA-rmse:76.44459\n",
      "[70]\tTRAIN_DATA-rmse:16.02098\tVALID_DATA-rmse:76.45827\n",
      "[71]\tTRAIN_DATA-rmse:15.30458\tVALID_DATA-rmse:76.51901\n",
      "[72]\tTRAIN_DATA-rmse:15.12049\tVALID_DATA-rmse:76.66824\n",
      "[73]\tTRAIN_DATA-rmse:14.51624\tVALID_DATA-rmse:76.56400\n",
      "[74]\tTRAIN_DATA-rmse:14.59896\tVALID_DATA-rmse:76.61585\n",
      "[75]\tTRAIN_DATA-rmse:14.24354\tVALID_DATA-rmse:76.61585\n",
      "[76]\tTRAIN_DATA-rmse:13.84568\tVALID_DATA-rmse:76.38177\n",
      "[77]\tTRAIN_DATA-rmse:13.34095\tVALID_DATA-rmse:76.38833\n",
      "[78]\tTRAIN_DATA-rmse:12.88326\tVALID_DATA-rmse:76.17014\n",
      "[79]\tTRAIN_DATA-rmse:12.75559\tVALID_DATA-rmse:76.12925\n",
      "[80]\tTRAIN_DATA-rmse:12.39263\tVALID_DATA-rmse:76.15875\n",
      "[81]\tTRAIN_DATA-rmse:12.39721\tVALID_DATA-rmse:76.18195\n",
      "[82]\tTRAIN_DATA-rmse:12.00366\tVALID_DATA-rmse:76.18195\n",
      "[83]\tTRAIN_DATA-rmse:11.99417\tVALID_DATA-rmse:76.19449\n",
      "[84]\tTRAIN_DATA-rmse:11.67113\tVALID_DATA-rmse:76.19449\n",
      "[85]\tTRAIN_DATA-rmse:11.62514\tVALID_DATA-rmse:76.19449\n",
      "[86]\tTRAIN_DATA-rmse:11.26487\tVALID_DATA-rmse:76.23738\n",
      "[87]\tTRAIN_DATA-rmse:11.26487\tVALID_DATA-rmse:76.23738\n",
      "[88]\tTRAIN_DATA-rmse:11.01068\tVALID_DATA-rmse:76.23738\n",
      "[89]\tTRAIN_DATA-rmse:10.98334\tVALID_DATA-rmse:76.24720\n",
      "[90]\tTRAIN_DATA-rmse:10.79855\tVALID_DATA-rmse:76.19545\n",
      "[91]\tTRAIN_DATA-rmse:10.55653\tVALID_DATA-rmse:76.08918\n",
      "[92]\tTRAIN_DATA-rmse:10.42114\tVALID_DATA-rmse:76.12162\n",
      "[93]\tTRAIN_DATA-rmse:10.42114\tVALID_DATA-rmse:76.12162\n",
      "[94]\tTRAIN_DATA-rmse:10.32603\tVALID_DATA-rmse:76.12162\n",
      "[95]\tTRAIN_DATA-rmse:10.11163\tVALID_DATA-rmse:76.12162\n",
      "[96]\tTRAIN_DATA-rmse:10.11163\tVALID_DATA-rmse:76.12162\n",
      "[97]\tTRAIN_DATA-rmse:9.86243\tVALID_DATA-rmse:76.12162\n",
      "[98]\tTRAIN_DATA-rmse:9.79346\tVALID_DATA-rmse:76.02471\n",
      "[99]\tTRAIN_DATA-rmse:9.72979\tVALID_DATA-rmse:76.02471\n",
      "[100]\tTRAIN_DATA-rmse:9.71849\tVALID_DATA-rmse:76.00565\n",
      "[101]\tTRAIN_DATA-rmse:9.71849\tVALID_DATA-rmse:76.00565\n",
      "[102]\tTRAIN_DATA-rmse:9.56965\tVALID_DATA-rmse:75.95418\n",
      "[103]\tTRAIN_DATA-rmse:9.40105\tVALID_DATA-rmse:76.02977\n",
      "[104]\tTRAIN_DATA-rmse:9.23133\tVALID_DATA-rmse:75.99819\n",
      "[105]\tTRAIN_DATA-rmse:9.06515\tVALID_DATA-rmse:75.97559\n",
      "[106]\tTRAIN_DATA-rmse:9.04609\tVALID_DATA-rmse:75.96224\n",
      "[107]\tTRAIN_DATA-rmse:8.89903\tVALID_DATA-rmse:75.98132\n",
      "[108]\tTRAIN_DATA-rmse:8.85479\tVALID_DATA-rmse:75.98946\n",
      "[109]\tTRAIN_DATA-rmse:8.71013\tVALID_DATA-rmse:76.01144\n",
      "[110]\tTRAIN_DATA-rmse:8.68484\tVALID_DATA-rmse:76.02424\n",
      "[111]\tTRAIN_DATA-rmse:8.65511\tVALID_DATA-rmse:76.01569\n",
      "[112]\tTRAIN_DATA-rmse:8.64768\tVALID_DATA-rmse:76.00925\n",
      "[113]\tTRAIN_DATA-rmse:8.64768\tVALID_DATA-rmse:76.00925\n",
      "[114]\tTRAIN_DATA-rmse:8.61737\tVALID_DATA-rmse:76.03007\n",
      "[115]\tTRAIN_DATA-rmse:8.50211\tVALID_DATA-rmse:76.05343\n",
      "[116]\tTRAIN_DATA-rmse:8.40404\tVALID_DATA-rmse:76.03306\n",
      "[117]\tTRAIN_DATA-rmse:8.33763\tVALID_DATA-rmse:76.00754\n",
      "[118]\tTRAIN_DATA-rmse:8.28857\tVALID_DATA-rmse:76.03715\n",
      "[119]\tTRAIN_DATA-rmse:8.27598\tVALID_DATA-rmse:76.00061\n",
      "[120]\tTRAIN_DATA-rmse:8.27598\tVALID_DATA-rmse:76.00061\n",
      "[121]\tTRAIN_DATA-rmse:8.27528\tVALID_DATA-rmse:76.00302\n",
      "[122]\tTRAIN_DATA-rmse:8.21760\tVALID_DATA-rmse:75.98119\n",
      "[123]\tTRAIN_DATA-rmse:8.21760\tVALID_DATA-rmse:75.98119\n",
      "[124]\tTRAIN_DATA-rmse:8.17341\tVALID_DATA-rmse:75.98119\n",
      "[125]\tTRAIN_DATA-rmse:8.08632\tVALID_DATA-rmse:75.96678\n",
      "[126]\tTRAIN_DATA-rmse:8.01043\tVALID_DATA-rmse:75.97849\n",
      "[127]\tTRAIN_DATA-rmse:7.97512\tVALID_DATA-rmse:75.98229\n",
      "[128]\tTRAIN_DATA-rmse:7.97512\tVALID_DATA-rmse:75.98229\n",
      "[129]\tTRAIN_DATA-rmse:7.89729\tVALID_DATA-rmse:75.99699\n",
      "[130]\tTRAIN_DATA-rmse:7.88426\tVALID_DATA-rmse:75.99940\n",
      "[131]\tTRAIN_DATA-rmse:7.83107\tVALID_DATA-rmse:76.03590\n",
      "[132]\tTRAIN_DATA-rmse:7.82891\tVALID_DATA-rmse:76.04577\n",
      "[133]\tTRAIN_DATA-rmse:7.79050\tVALID_DATA-rmse:76.03029\n",
      "[134]\tTRAIN_DATA-rmse:7.79050\tVALID_DATA-rmse:76.03029\n",
      "[135]\tTRAIN_DATA-rmse:7.78105\tVALID_DATA-rmse:76.01951\n",
      "[136]\tTRAIN_DATA-rmse:7.74653\tVALID_DATA-rmse:76.00607\n",
      "[137]\tTRAIN_DATA-rmse:7.74769\tVALID_DATA-rmse:75.98290\n",
      "[138]\tTRAIN_DATA-rmse:7.69374\tVALID_DATA-rmse:76.01176\n",
      "[139]\tTRAIN_DATA-rmse:7.64641\tVALID_DATA-rmse:75.98786\n",
      "[140]\tTRAIN_DATA-rmse:7.64641\tVALID_DATA-rmse:75.98786\n",
      "[141]\tTRAIN_DATA-rmse:7.64641\tVALID_DATA-rmse:75.98786\n",
      "ERROR:  63.23971811930339\n"
     ]
    }
   ],
   "source": [
    "#TRAIN GRADIENT BOOSTED TREE\n",
    "from sklearn.metrics import mean_squared_error, mean_absolute_error\n",
    "\n",
    "#Convert to Dmatrix\n",
    "TRAIN_DATA = xgb.DMatrix(train_x, train_y, feature_names = columns[:6])\n",
    "VALID_DATA = xgb.DMatrix(valid_x, valid_y, feature_names = columns[:6])\n",
    "#Parameters for boosted tree\n",
    "XGBOOST_PARAMS = {\"objective\": \"reg:squarederror\", \"subsample\": 0.6,\n",
    "                  \"colsample_bytree\" : 0.6, \"learning_rate\" : 0.1, \"max_depth\" : 100,\n",
    "                  \"alpha\": 20, \"n_estimators\": 12}\n",
    "#Train the xgboost model\n",
    "XGB_NUTRITION_MODEL = xgb.train(XGBOOST_PARAMS, TRAIN_DATA, evals = [(TRAIN_DATA, \"TRAIN_DATA\"), (VALID_DATA, \"VALID_DATA\")],\n",
    "                            num_boost_round = 200, early_stopping_rounds = 40)\n",
    "predictions = XGB_NUTRITION_MODEL.predict(VALID_DATA)\n",
    "error = mean_absolute_error(valid_y, predictions)\n",
    "print('ERROR: ', error)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Save Model\n",
    "XGB_NUTRITION_MODEL.save_model('METRICS_NUTRITION_GRADBOOSTED_MODELS[63.24].model')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2160x2160 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Print desicion tree -> transparency into exactly what the model is doing\n",
    "ig, ax = plt.subplots(figsize=(30, 30))\n",
    "xgb.plot_tree(XGB_NUTRITION_MODEL, num_trees=10, ax=ax)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "bb33c15c13a14f140d690812bf0a39a2e5e1e289641f45cea7a1bbeb79eff1a7"
  },
  "kernelspec": {
   "display_name": "Python 3.8.13 ('base')",
   "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": "3.8.13"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}