Case+Study+2+-+Instacart+Market+Basket.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#################### Dataset: Instacart Market Basket ####################"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#### Cargar librerías a usar\n",
    "\n",
    "## Tu código aqui\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import os"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['products.csv', 'orders.csv', 'order_products__train.csv', 'departments.csv', 'aisles.csv', 'order_products__prior.csv', 'sample_submission.csv']\n"
     ]
    }
   ],
   "source": [
    "#### Cargar datos\n",
    "\n",
    "instaFolder = '/Users/Avila/Documents/Kaggle/Instacart Market Basket'\n",
    "\n",
    "os.chdir(instaFolder)\n",
    "print (os.listdir(instaFolder))\n",
    "\n",
    "dfOrderTrain = pd.read_csv('order_products__train.csv')\n",
    "dfOrderPrior = pd.read_csv('order_products__prior.csv')\n",
    "dfOrder = pd.read_csv('orders.csv')\n",
    "dfDepartments = pd.read_csv('departments.csv')\n",
    "dfProducts = pd.read_csv('products.csv')\n",
    "dfAisles = pd.read_csv('aisles.csv')\n",
    "dfSample_sub = pd.read_csv('sample_submission.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#### Limpieza de datos\n",
    "\n",
    "orders_by_dow = dfOrder[['order_id','order_dow']].groupby('order_dow').count()\n",
    "orders_by_hour = dfOrder[['order_id','order_hour_of_day']].groupby('order_hour_of_day').count()\n",
    "orders_since_prior = dfOrder[['order_id','days_since_prior_order']].dropna().groupby('days_since_prior_order').count()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfUAAAFUCAYAAAA9CKVaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHQAAQxB1AAAMQdQBADDE/wfJ3+tXIElmCAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x115166a10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfUAAAFaCAYAAAAHAsQqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHQAAQxDqAAAY4v8D3+8fVu3xB2YAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x115152090>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfUAAAFiCAYAAADvK0HqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11534b890>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#### Gráficas\n",
    "\n",
    "orders_by_dow.plot(kind='bar')\n",
    "orders_by_hour.plot(kind='bar')\n",
    "orders_since_prior.plot(kind='bar')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "      order_count  reorder_proportion  product_id  \\\n",
      "4126           92            0.934783        1729   \n",
      "4152           35            0.914286       30630   \n",
      "4153          368            0.913043       20940   \n",
      "4175           59            0.898305       12193   \n",
      "4190           45            0.888889       31764   \n",
      "4192           81            0.888889       21038   \n",
      "4212           35            0.885714       48927   \n",
      "4215        18726            0.884172       24852   \n",
      "4216          120            0.883333         117   \n",
      "4219          483            0.881988       39180   \n",
      "4220          269            0.881041       12384   \n",
      "4221          461            0.878525       24024   \n",
      "4222           81            0.876543       36717   \n",
      "4223         1206            0.875622       38689   \n",
      "4234           32            0.875000       14446   \n",
      "4270           32            0.875000       15981   \n",
      "4285           79            0.873418        9810   \n",
      "4286          125            0.872000        5514   \n",
      "4287           31            0.870968       27514   \n",
      "4288           31            0.870968       30219   \n",
      "\n",
      "                               product_name  aisle_id  department_id  \n",
      "4126                   2% Lactose Free Milk        84             16  \n",
      "4152         Yogurt, Sheep Milk, Blackberry       120             16  \n",
      "4153                   Organic Low Fat Milk        84             16  \n",
      "4175              100% Florida Orange Juice        98              7  \n",
      "4190  Original Sparkling Seltzer Water Cans       115              7  \n",
      "4192                Organic Spelt Tortillas       128              3  \n",
      "4212                 Fat Free Plain Yoghurt       120             16  \n",
      "4215                                 Banana        24              4  \n",
      "4216                     Petit Suisse Fruit         2             16  \n",
      "4219                 Organic Lowfat 1% Milk        84             16  \n",
      "4220    Organic Lactose Free 1% Lowfat Milk        91             16  \n",
      "4221                         1% Lowfat Milk        84             16  \n",
      "4222                     Double Fiber Bread       112              3  \n",
      "4223               Organic Reduced Fat Milk        84             16  \n",
      "4234     Pure Premium Low Acid Orange Juice        31              7  \n",
      "4270                    Unsweetened Soymilk        91             16  \n",
      "4285       100% Lactose Free 1% Lowfat Milk        91             16  \n",
      "4286         Organic Homogenized Whole Milk        84             16  \n",
      "4287                 Maple Glazed Honey Ham        96             20  \n",
      "4288          Lactose Free Blueberry Yogurt       120             16  \n"
     ]
    }
   ],
   "source": [
    "#### Cuántos productos compra la gente?\n",
    "## Prior\n",
    "n_products_prior = dfOrderPrior[['order_id','product_id']].groupby('order_id').count()\n",
    "n_products_prior = n_products_prior.sort_index()\n",
    "n_products_prior['order'] = n_products_prior.index\n",
    "n_products_prior = n_products_prior.groupby('product_id').count()\n",
    "\n",
    "n_products_prior.plot(kind='bar')\n",
    "\n",
    "## Train\n",
    "n_products_train = dfOrderTrain[['order_id','product_id']].groupby('order_id').count()\n",
    "n_products_train = n_products_train.sort_index()\n",
    "n_products_train['order'] = n_products_train.index\n",
    "n_products_train = n_products_train.groupby('product_id').count()\n",
    "\n",
    "n_products_train.plot(kind='bar')\n",
    "\n",
    "\n",
    "#### Mejores productos venta (Train)\n",
    "best_products = dfOrderTrain[['order_id','product_id']].groupby('product_id').count()\n",
    "best_products = best_products.sort_values(by=['order_id'],ascending=False)\n",
    "best_products['product'] = best_products.index\n",
    "best_products = best_products.rename(columns={'order_id':'order_count'})\n",
    "best_products = pd.merge(best_products,dfProducts,how='left',left_on='product',right_on='product_id')\n",
    "\n",
    "#### Productos reordenados\n",
    "reordered = dfOrderTrain['reordered'].value_counts()\n",
    "reordered.plot(kind='bar')\n",
    "\n",
    "#### MMás frecuentes reordenados\n",
    "often_reorder = dfOrderTrain[['order_id','product_id','reordered']].groupby('product_id').agg({'order_id':'count','reordered':'mean'}).rename(columns={'order_id':'order_count','reordered':'reorder_proportion'})\n",
    "often_reorder = often_reorder.sort_values(by=['reorder_proportion'],ascending=[False])\n",
    "often_reorder = pd.merge(often_reorder,dfProducts, how='left', left_on=often_reorder.index.values,right_on='product_id')\n",
    "print (often_reorder[often_reorder['order_count'] > 30].head(20))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "    product_id  add_to_cart_order  product_count       pct  \\\n",
      "0        45004                  1             39  0.661017   \n",
      "1        11885                  1             41  0.594203   \n",
      "2        13128                  1             12  0.571429   \n",
      "3         4100                  1             14  0.560000   \n",
      "4         1729                  1             48  0.521739   \n",
      "5         6729                  1             35  0.486111   \n",
      "6         9285                  1             13  0.481481   \n",
      "7         6848                  1             12  0.461538   \n",
      "8        12640                  1             12  0.461538   \n",
      "9        26405                  1             47  0.447619   \n",
      "10       13948                  1             27  0.442623   \n",
      "11       21883                  1             11  0.440000   \n",
      "12        4731                  1             22  0.431373   \n",
      "13       26900                  1             24  0.428571   \n",
      "14       40939                  1            145  0.423977   \n",
      "15       35470                  1             16  0.421053   \n",
      "16        5351                  1             12  0.413793   \n",
      "17       17924                  1             38  0.408602   \n",
      "18       14408                  1             12  0.400000   \n",
      "19       43739                  1             31  0.387500   \n",
      "\n",
      "                                       product_name  aisle_id  department_id  \n",
      "0                            White Multifold Towels        54             17  \n",
      "1                          Sparkling Water, Bottles       115              7  \n",
      "2       Purified Alkalkine Water with Minerals pH10       115              7  \n",
      "3                                Organic Dark Roast        26              7  \n",
      "4                              2% Lactose Free Milk        84             16  \n",
      "5                                       Cookie Tray         8              3  \n",
      "6                       Boneless Pork Shoulder Butt         7             12  \n",
      "7                                    Party Tumblers       111             17  \n",
      "8   Sport Bottle with Flip Cap Natural Spring Water       115              7  \n",
      "9                  XL Pick-A-Size Paper Towel Rolls        54             17  \n",
      "10                               Wastebasket Liners        60             17  \n",
      "11                                    Irish Whiskey       124              5  \n",
      "12                                    Classic Roast        26              7  \n",
      "13                         USDA AA Extra Large Eggs        86             16  \n",
      "14                                   Drinking Water       115              7  \n",
      "15                                    Water Mineral       115              7  \n",
      "16                           Smoked Whitefish Salad         1             20  \n",
      "17                                    Seltzer Water       115              7  \n",
      "18             Unflavored Oral Electrolyte Solution        92             18  \n",
      "19                                 Variety Diet Tea        94              7  \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeEAAAFiCAYAAADBbmB4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3XtclHX+//+nMOJp0LDIytSSxM92JGzZXD94xj4eK03B\nA+p21NbdNqnUNFNUpLWTq+aun892crdC0w6UtYViFrcyJVHJdF0PpLtZeBZQCOb9/cOf84sUrmEE\n3g487rebtxsz1/Wa14vDXE+vmeu6poExxggAANS6INsDAABQXxHCAABYQggDAGAJIQwAgCWEMAAA\nlhDCAABY4qrthvn5JypcFhbWVEeOFPn1uP7WBlLPQJrVRs9AmtVGz0Ca1UbPQJrVRs9AmtVGT6e6\n8PDQc95/Qe0Ju1zBtV4bSD0DaVYbPQNpVhs9A2lWGz0DaVYbPQNpVhs9/a27oEIYAID6hBAGAMAS\nQhgAAEsIYQAALCGEAQCwhBAGAMASQhgAAEtq/WIdAIDAcnfqmmp9vBcn96zWx/NFXt5ezZuXooUL\nl1xQj8eeMAAAlrAnDAC44Kxala6PPnpfxcU/6q674rVs2esKCgrSjTdGafz43+nEiROaNesJFRYW\nqqysTPfdN16dOv1SiYnD1KZNO7ndTXT//b9XcvI0GWPUsuXF3sfetClbS5a8oODgYF1xRWs99thU\nffTRB3r//XcVHNxAo0ffq+PHjyst7e/leh48ePCcj3c+CGEAwAWpefPmeuSRqXrwwXv1f/+3VI0b\nN9asWU9ow4YvtH79F7rlll9p2LDhys//QQ8+eK+WLXtHJ0+e1Nix96hLl19qypRp6t37Ng0adKdW\nr/5Ib731powxeuqpOVq8+P8UFtZS//u/i7VqVbpcLpdCQ0P117/+r3bt2n/Onp9++slZj3e+CGEA\nwAXp6quv1v79+3T06BE98sjvJUlFRUX697/3Ky9vj/r0+R9JUnj4pWratJmOHDksSWrb9ipJ0r59\n32rgwDslSTfccJPeeutNHT16RIcOHdQTT0yWJBUXF+uXv/yVrryyjdq2bSdJFfY81+OdL+shXNkb\n/jbevAcAXBiCgoJ0+eWtdemlrfT88y/I5XJp1ap0degQqf3792vz5hxFRv6X8vN/0IkTx9W8eQtJ\nUoMGDSRJV13VXl9/vUUdOkTqm2+2SZJatLhIl156qVJTn5Xb7dZnn32iJk2a6vvvD6hBg9OHSVXU\nMy8v76zHO1/WQxgAgIqEhYUpPn6kJky4X2VlZbr88ivUs2ecRo/+jebOTdbatatVXFysxx6bKper\nfKSNGXOPkpOnKSPjI11xRWtJp4P9oYce0aOPPiRjjJo2baYnnpip778/4NjzXI93vghhAEClXpzc\nU+HhoZV+Hnxl/Knt12+gt+622/rpttv6lVveuHFjzZ37zFl1b76Z7v36oosu0rPPLjxrnZiYWxUT\nc+tZ/X6qop7nerzzwSlKAABYQggDAGCJYwh7PB5Nnz5d8fHxSkxMVF5eXrnl7777ru68804NGTJE\nr732Wo0NCgBAXeP4nnBGRoZKSkqUlpamnJwcpaamavHixd7lf/zjH/Xee++padOm6t+/v/r3768W\nLVrU6NAAANQFjiGcnZ2t2NhYSVJUVJRyc3PLLe/YsaNOnDghl8slY4z30HAAAFA5xxAuKCiQ2+32\n3g4ODlZpaan3UPAOHTpoyJAhatKkieLi4tS8efOamxYAgDrEMYTdbrcKCwu9tz0ejzeAt2/frrVr\n12r16tVq2rSpHn30UX3wwQfq27dvhY8XFtZULlewT8OFh4f6tJ6/659vnY2egTSrjZ6BNKuNnoE0\nq42egTSrjZ6BNKuNnv7UOYZwdHS0MjMz1a9fP+Xk5CgyMtK7LDQ0VI0bN1ajRo0UHBysli1b6vjx\n45U+3pEjRT4PV5Xzyvw9h622z32zUVdfegbSrDZ6BtKsNnoG0qw2egbSrDZ6OtVVFNCOIRwXF6es\nrCwlJCTIGKOUlBSlp6erqKhI8fHxio+P14gRI9SwYUO1bdtWd955Z5WHBwCgPnIM4aCgICUnJ5e7\nLyIiwvv18OHDNXz48OqfDACAOo6LdQAAYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACW\nEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACA\nJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWuJxW8Hg8mjFj\nhnbs2KGQkBDNnj1b7dq1kyTl5+dr4sSJ3nW/+eYbJSUlafjw4TU3MQAAdYRjCGdkZKikpERpaWnK\nyclRamqqFi9eLEkKDw/X0qVLJUmbNm3Sc889p2HDhtXsxAAA1BGOIZydna3Y2FhJUlRUlHJzc89a\nxxijWbNm6emnn1ZwcHD1TwkAQB3kGMIFBQVyu93e28HBwSotLZXL9f+XrlmzRh06dFD79u0dG4aF\nNZXL5VtQh4eH+rSev+ufb52NnrUx68Ckdypclv7M7TXS03ZdfekZSLPa6BlIs9roGUiz2ujpT51j\nCLvdbhUWFnpvezyecgEsSe+++65Gjx7tU8MjR4p8Hi4//4TP64aHh1Zp/fOts9HTxqw/dyH/TgLp\nd2mjZyDNaqNnIM1qo2cgzWqjp1NdRQHteHR0dHS01q1bJ0nKyclRZGTkWevk5uYqOjra11kBAIB8\n2BOOi4tTVlaWEhISZIxRSkqK0tPTVVRUpPj4eB0+fFhut1sNGjSojXkBAKgzHEM4KChIycnJ5e6L\niIjwft2yZUu9807F7yECAIBz42IdAABYQggDAGAJIQwAgCWO7wkjMN2duqbCZS9O7llnegJAICOE\na0Fl4SQRUABQXxHCFzj2LgGg7uI9YQAALCGEAQCwhBAGAMASQhgAAEs4MAvWcfQ4gPqKPWEAACxh\nT7gKOF0IEn8HAKoPIQxc4Hi5Hqi7CGEENPZKAQQy3hMGAMASQhgAAEsIYQAALOE9YdRL9eWjHnnP\nHLiwsScMAIAlhDAAAJYQwgAAWEIIAwBgSb08MIuDVYCaw/ML8F29DGEAlasvQVpfvk9cuAI2hHny\nALDlfLY/bLvwU44h7PF4NGPGDO3YsUMhISGaPXu22rVr512+ZcsWpaamyhij8PBwzZs3T40aNarR\noQEAqAscD8zKyMhQSUmJ0tLSlJSUpNTUVO8yY4yeeOIJzZ07V6+//rpiY2P173//u0YHBgCgrnDc\nE87OzlZsbKwkKSoqSrm5ud5le/bs0UUXXaSXX35ZO3fuVLdu3dS+ffuamxZAncXLtKiPHEO4oKBA\nbrfbezs4OFilpaVyuVw6cuSINm3apOnTp6tt27YaN26crr/+enXu3LlGhwaAM/i8ZQQyxxB2u90q\nLCz03vZ4PHK5TpdddNFFateunSIiIiRJsbGxys3NrTSEw8KayuUK9mm48PBQn9arrrr60jOQZrXR\nM5BmtdEzkGa10bM2Z63L31sg9vSnzjGEo6OjlZmZqX79+iknJ0eRkZHeZW3atFFhYaHy8vLUrl07\nbdy4UXfddVelj3fkSJHPw+Xnn/B53eqoqy89A2lWGz0DaVYbPQNpVhs9a2rW6nq5Pjw81K8Z/a2r\nLz2d6ioKaMcQjouLU1ZWlhISEmSMUUpKitLT01VUVKT4+HjNmTNHSUlJMsbo5ptvVvfu3as8PAAA\n9ZFjCAcFBSk5ObncfWdefpakzp07680336z+yQAAqOMC9mIdAACcL9sH9hHCAIALhr/vfQfqKW58\nihIAAJawJwwAddj5vNxa3/ZKbWBPGAAASwhhAAAsIYQBALCEEAYAwBJCGAAASwhhAAAsIYQBALCE\nEAYAwBJCGAAASwhhAAAsIYQBALCEEAYAwBJCGAAASwhhAAAsIYQBALCEEAYAwBJCGAAASwhhAAAs\nIYQBALCEEAYAwBJCGAAASwhhAAAsIYQBALDE5bSCx+PRjBkztGPHDoWEhGj27Nlq166dd/nLL7+s\n5cuXq2XLlpKkmTNnqn379jU3MQAAdYRjCGdkZKikpERpaWnKyclRamqqFi9e7F2em5urp556Stdf\nf32NDgoAQF3jGMLZ2dmKjY2VJEVFRSk3N7fc8q+//lpLlixRfn6+unfvrgceeKBmJgUAoI5xDOGC\nggK53W7v7eDgYJWWlsrlOl3av39/jRgxQm63WxMmTFBmZqZ69OhR4eOFhTWVyxXs03Dh4aE+rVdd\ndfWlZyDNaqNnIM1qo2cgzWqjZyDNaqNnIM1a1Vp/+jiGsNvtVmFhofe2x+PxBrAxRmPGjFFo6OnG\n3bp107Zt2yoN4SNHinweLj//hM/rVkddfekZSLPa6BlIs9roGUiz2ugZSLPa6BlIs1alNjw8tNJ1\nKwpox6Ojo6OjtW7dOklSTk6OIiMjvcsKCgo0YMAAFRYWyhij9evX894wAAA+ctwTjouLU1ZWlhIS\nEmSMUUpKitLT01VUVKT4+Hg9/PDDGj16tEJCQtS5c2d169atNuYGACDgOYZwUFCQkpOTy90XERHh\n/fqOO+7QHXfcUf2TAQBQx3GxDgAALCGEAQCwhBAGAMASx/eEAQDA2e5OXVPhshcn9/TpMdgTBgDA\nEkIYAABLCGEAACwhhAEAsIQQBgDAEkIYAABLCGEAACwhhAEAsIQQBgDAEkIYAABLCGEAACwhhAEA\nsIQQBgDAEkIYAABLCGEAACwhhAEAsIQQBgDAEkIYAABLCGEAACwhhAEAsIQQBgDAEkIYAABLHEPY\n4/Fo+vTpio+PV2JiovLy8s653hNPPKGnn3662gcEAKCucgzhjIwMlZSUKC0tTUlJSUpNTT1rnTfe\neEP//Oc/a2RAAADqKscQzs7OVmxsrCQpKipKubm55ZZ/9dVX2rx5s+Lj42tmQgAA6ijHEC4oKJDb\n7fbeDg4OVmlpqSTphx9+0KJFizR9+vSamxAAgDrK5bSC2+1WYWGh97bH45HLdbrsww8/1JEjR3T/\n/fcrPz9fp06dUvv27TV48OAKHy8srKlcrmCfhgsPD/Vpveqqqy89A2lWGz0DaVYbPQNpVhs9A2lW\nGz0Dadba6OkYwtHR0crMzFS/fv2Uk5OjyMhI77LRo0dr9OjRkqSVK1dq9+7dlQawJB05UuTTYJKU\nn3/C53Wro66+9AykWW30DKRZbfQMpFlt9AykWW30DKRZq7NnRaHsGMJxcXHKyspSQkKCjDFKSUlR\nenq6ioqKeB8YAIDz4BjCQUFBSk5OLndfRETEWes57QEDAIDyuFgHAACWEMIAAFhCCAMAYAkhDACA\nJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMA\nYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWEMIA\nAFhCCAMAYIljCHs8Hk2fPl3x8fFKTExUXl5eueX/+Mc/NGTIEN1111165ZVXamxQAADqGscQzsjI\nUElJidLS0pSUlKTU1FTvsrKyMj3zzDN6+eWXlZaWptdee02HDx+u0YEBAKgrXE4rZGdnKzY2VpIU\nFRWl3Nxc77Lg4GCtWrVKLpdLhw4dksfjUUhISM1NCwBAHeK4J1xQUCC32+29HRwcrNLSUu9tl8ul\njz76SLfffrtiYmLUpEmTmpkUAIA6xnFP2O12q7Cw0Hvb4/HI5Spf1qdPH/Xu3VuTJ0/W22+/rSFD\nhlT4eGFhTeVyBfs0XHh4qE/rVVddfekZSLPa6BlIs9roGUiz2ugZSLPa6BlIs9ZGT8cQjo6OVmZm\npvr166ecnBxFRkZ6lxUUFGjcuHF68cUXFRISoiZNmigoqPKd6yNHinwaTJLy80/4vG511NWXnoE0\nq42egTSrjZ6BNKuNnoE0q42egTRrdfasKJQdQzguLk5ZWVlKSEiQMUYpKSlKT09XUVGR4uPjNXDg\nQI0cOVIul0sdO3bUoEGD/BoYAID6xjGEg4KClJycXO6+iIgI79fx8fGKj4+v/skAAKjjuFgHAACW\nEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACA\nJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMAYAkhDACAJYQwAACWEMIAAFhCCAMA\nYAkhDACAJYQwAACWEMIAAFhCCAMAYInLaQWPx6MZM2Zox44dCgkJ0ezZs9WuXTvv8vfee0+vvPKK\ngoODFRkZqRkzZigoiGwHAMCJY1pmZGSopKREaWlpSkpKUmpqqnfZqVOn9Pzzz+vVV1/VG2+8oYKC\nAmVmZtbowAAA1BWOIZydna3Y2FhJUlRUlHJzc73LQkJC9MYbb6hJkyaSpNLSUjVq1KiGRgUAoG5x\nfDm6oKBAbrfbezs4OFilpaVyuVwKCgrSJZdcIklaunSpioqK1KVLl0ofLyysqVyuYJ+GCw8P9Wm9\n6qqrLz0DaVYbPQNpVhs9A2lWGz0DaVYbPQNp1tro6RjCbrdbhYWF3tsej0cul6vc7Xnz5mnPnj1a\nsGCBGjRoUOnjHTlS5NNgkpSff8Lndaujrr70DKRZbfQMpFlt9AykWW30DKRZbfQMpFmrs2dFoez4\ncnR0dLTWrVsnScrJyVFkZGS55dOnT1dxcbFeeOEF78vSAADAmeOecFxcnLKyspSQkCBjjFJSUpSe\nnq6ioiJdf/31evPNN3XLLbdozJgxkqTRo0crLi6uxgcHACDQOYZwUFCQkpOTy90XERHh/Xr79u3V\nPxUAAPUAJ/QCAGAJIQwAgCWEMAAAlhDCAABYQggDAGAJIQwAgCWEMAAAlhDCAABYQggDAGAJIQwA\ngCWEMAAAlhDCAABYQggDAGAJIQwAgCWEMAAAlhDCAABYQggDAGAJIQwAgCWEMAAAlhDCAABYQggD\nAGAJIQwAgCWEMAAAlhDCAABYQggDAGAJIQwAgCWOIezxeDR9+nTFx8crMTFReXl5Z61z8uRJJSQk\naNeuXTUyJAAAdZFjCGdkZKikpERpaWlKSkpSampqueVbt27VyJEjtW/fvhobEgCAusgxhLOzsxUb\nGytJioqKUm5ubrnlJSUlWrRokdq3b18zEwIAUEe5nFYoKCiQ2+323g4ODlZpaalcrtOlnTp1qrnp\nAACowxxD2O12q7Cw0Hvb4/F4A9gfYWFN5XIF+7RueHioXz38rasvPQNpVhs9A2lWGz0DaVYbPQNp\nVhs9A2nW2ujpmKbR0dHKzMxUv379lJOTo8jISL8GOuPIkSKf183PP+FXD3/r6kvPQJrVRs9AmtVG\nz0Ca1UbPQJrVRs9AmrU6e1YUyo4hHBcXp6ysLCUkJMgYo5SUFKWnp6uoqEjx8fF+DQcAAHwI4aCg\nICUnJ5e7LyIi4qz1li5dWn1TAQBQD3CxDgAALCGEAQCwhBAGAMASQhgAAEsIYQAALCGEAQCwhBAG\nAMASQhgAAEsIYQAALCGEAQCwhBAGAMASQhgAAEsIYQAALCGEAQCwhBAGAMASQhgAAEsIYQAALCGE\nAQCwhBAGAMASQhgAAEsIYQAALCGEAQCwhBAGAMASQhgAAEsIYQAALCGEAQCwhBAGAMASxxD2eDya\nPn264uPjlZiYqLy8vHLL16xZoyFDhig+Pl7Lli2rsUEBAKhrHEM4IyNDJSUlSktLU1JSklJTU73L\nfvzxR82dO1cvvviili5dqrS0NB08eLBGBwYAoK5wDOHs7GzFxsZKkqKiopSbm+tdtmvXLrVt21Yt\nWrRQSEiIOnXqpA0bNtTctAAA1CENjDGmshWmTp2qPn36qFu3bpKk7t27KyMjQy6XSxs3btTf/vY3\nPf/885Kk+fPn64orrtDQoUNrfnIAAAKc456w2+1WYWGh97bH45HL5TrnssLCQoWGhtbAmAAA1D2O\nIRwdHa1169ZJknJychQZGeldFhERoby8PB09elQlJSXauHGjbr755pqbFgCAOsTx5WiPx6MZM2bo\nn//8p4wxSklJ0bZt21RUVKT4+HitWbNGixYtkjFGQ4YM0ciRI2trdgAAAppjCAMAgJrBxToAALCE\nEAYAwBJCGAAASwhhAAAscdlsfvjwYW3YsEEnTpxQ8+bNFRUVpUsvvfSC7elvbW3X1ZdZgerg8XgU\nFOTf/oi/tbVdV196BtKsZ1g7Onr58uVKS0tTp06d1KxZMxUWFmrDhg0aOnSohg8f7ljvz4b7fHr6\nW1vbdfVl1p+rL09aNmrVU7tv3z7NnTtXubm5crlc8ng8ioyM1JQpU3T11VfXSG1t19WXnoE06zkZ\nS+Lj401JSUm5+4qLi83gwYMda5ctW2aGDBliUlJSzPz5801KSoq58847zWuvvVZjPf2tre26+jKr\nMcZ8++23Zvz48SY2Ntb06NHDdOvWzdx3331m9+7dNVJXX3oG0qz+1iYmJpqcnJxy923atMnEx8c7\n9vO3trbr6kvPQJr1XKy9HF1aWqri4mI1bNjQe9+pU6fUoEEDx9oVK1bo9ddfL1dbUlKi4cOHV7r3\ndD49/a2t7br6Mqt0+rrmSUlJuummm7z35eTkaMqUKXrjjTeqva6+9AykWf2tLSkpKbe+dPoDanzh\nb21t19WXnoE067lYC+EHH3xQgwcPVrt27RQaGqqCggLl5eVpypQpjrX+brjPp6e/tbVdV19mlerP\nk5aNWvXXduzYUVOmTFFsbKxCQ0NVWFioTz75RB07dnTs529tbdfVl56BNOu5WL1iVmlpqXbt2qWC\nggK53W5FRER4PxyiMmvWrFFqauo5N9zdu3evkZ7nU1vbdfVl1ieffFIlJSVnPRFCQkI0c+bMaq+r\nLz0DaVZ/a40xysjIUHZ2tvfvLjo6WnFxcY7/mfe3trbr6kvPQJr1XC64y1YuX77cp49CPJ+Nvr89\nq7O2tuts9KzpWevLk5aNWs30PJcDBw7osssuq1LN+dbWdl196Rkws1b5XeQaUFZW5v06PT3d78dZ\ntmyZX3X+9Dx58qQpLi6ucu3Bgwer3LOsrMwcOHDAlJWV+TXroUOHjMfjcaw9ceLEOe+vas/i4mJz\n8uTJKtd5PB6/e/7Ud999V6t19aVnIM3qb+1jjz3mdz9/a2u7rr70DJRZrYXw+RwNWRGnDffq1atN\n9+7dTe/evc3777/vvT8xMdHxsXfu3GnGjx9vJk+ebLKyskyPHj1Mr169zOrVqyut2717d7l/Q4cO\n9X5dmSlTphhjjMnJyTG9e/c2d911l+nbt6/ZtGmT46xvvvmmWbBggcnNzTW33XabGTRokOnTp4/5\n7LPPKq278cYb/fqPzO7du83vfvc7M3HiRLNp0ybTp08f07NnT/Pee+851ubl5Zm7777bdO/e3Vx7\n7bVm6NChZuLEieaHH36o8hxn1PUnra2egTTr+dYCtcVaCFfnId6+Gjp0qDl69Kg5fPiwSUxMNCtX\nrjTGGDNq1CjH2hEjRpj169eblStXmk6dOpmDBw+aEydOOM7brVs3c9ttt5nExEQzatQoc8stt5hR\no0Y5Bv+Z5WPGjDF79uwxxhhz4MABM3LkSMdZBw8ebAoLC83o0aO9YX/gwAHH036GDRtmZs6caRIT\nE8369esd+5wxcuRIk5WVZT788EMTExNjDhw4YAoLC82wYcMca++++27vjJs2bTJPP/202bp1q7nv\nvvt87g/44+OPPzbJycnm0UcfNbNmzTKrVq0q92pMdTt06JCZO3euefbZZ83hw4e99y9YsKDSurKy\nMvPxxx+bL774whw9etRMmjTJTJkyxeTn51d5hpSUFJ/WW7VqlTHGmMLCQpOammrGjh1r5s2bZwoK\nChxrv/32W7N27Vpz8uRJM3/+fHP//febP/7xj+b48eOV1k2cONH7SmFVZWZmmk8//dQUFxebmTNn\nmqSkJPPvf//bp9p3333XPPnkk+axxx4zc+fONZ988olPddX192Pt6OjzORoyMTFRP/74Y7n7jDFq\n0KBBpaczNGzYUC1atJAkvfDCCxozZowuv/xyn9438ng8iomJkSStX79eF198sSQ5vg+9YsUKPfnk\nkxo+fLi6dOmixMRELV261LHfGcHBwbrqqqskSa1atZLH43GsadiwoZo2bapmzZqpTZs23lqn77NR\no0aaPn26tm7dqiVLlmjWrFm69dZb1aZNG40ePbrCutLSUv3617+WMUbPPvusWrVqJcn5ZyNJBQUF\n3pPbo6KiNG/ePCUlJen48eOOtZKUkZGhzz//3HvRlk6dOul//ud/qvxeoK8OHz6sJUuWqFGjRho7\ndqzCwsIkSQsXLtSECRMqrPN4PFqzZo1CQ0P1X//1X5o7d66CgoI0ceJEXXLJJVWaYe7cuT4dPf7B\nBx+ob9++Kioq0oIFC7R9+3Zdd911Gj9+vJo1a1Zh3b59+7R792796le/0pIlS/T111/rmmuu0bhx\n4xQaGlppz6SkJD3++OPe50dVrF27Vi6XSzExMUpNTdXx48c1ceJEXXHFFY616enpys7O1smTJxUW\nFqZf//rX6tq1a4Xrz5w5Ux6PR127dvVeJGbdunX67LPPNGfOnEp7paWlVbgsPj6+wmWPPfaY4uLi\nVFpaqlGjRmnJkiVq3bq1vvzyy0r7TZ06VZKUn5+vo0ePKj4+Xs2aNdO0adP05z//udLahIQE79fG\nGO3atUubN2+WpEq3la+//rr69u2rOXPmqE2bNpo2bZo+//xzTZ8+Xc8880ylPSdNmqSHHnpIc+bM\n0WWXXaY//OEP2rBhg5KSkrRkyZIK6zZt2qR7771Xo0aN0uDBg31+Dk+dOlXFxcUqLCzUggULNGjQ\nILVq1UpPPPGE/vrXv1ZaO3v2bIWGhqpnz57KzMyU2+3WunXr9NVXX+kPf/hDhXXn8/fzc9ZC+HwO\n8X7kkUc0bdo0LVq0SMHBwT73bN26tebOnauHHnpIbrdbCxcu1D333OPTBv/qq6/W1KlTNWvWLKWm\npkqSliyOT7yPAAAOTElEQVRZ4rgBvfjii/X888/rqaee0tatW32etaCgQIMHD1ZRUZGWL1+uQYMG\nKTU11acNUs+ePTV+/HhFRkbqgQceUGxsrD799FPdeuutldaZ/+8YvRtuuEELFizQiRMntGHDBu3Z\ns6fSutatW+vhhx9WWVmZmjVrpueee05ut1vh4eGOs1555ZWaPn26unbtqrVr1+r666/X2rVr1aRJ\nE8daf58I/m5EpfqxIfV3IyoFzoZ0586d+tvf/lbuvl69epX7WVdk9+7dyszM1KBBg3z63s4oKSnx\n/n394he/0IMPPqilS5d6n3cVycvL02uvvaaSkhINHDjQe8BiZX/HZ4wcOVIrVqzQ1KlT1aRJEyUl\nJTmG6M97n3kuRURE6KOPPnKsCQ4O1q9+9Sv9+c9/1qxZsySd/n4/+OCDSutat26tRYsW6U9/+pMG\nDRqkAQMGqGvXrmrTpo3cbneFdXv37tXf//53GWPUv39/jRw5UpL0yiuvOM66fft2799B165d9Zvf\n/EYvvfSS49X6zufv5+eshfCMGTPOOqKxR48eiouLc6y96aabdPvtt2vHjh0+rX9GSkqK3n33Xe+G\n4fLLL9err76qv/zlL461s2fP1po1a8pdEq9Vq1ZKTEx0rHW5XJo6dapWrlzp+IQ7Y+XKlSopKdH2\n7dvVuHFjNWjQQJGRkbrrrrsca++//359+eWX+uyzz3TFFVfo0KFDSkxMdDx9a/DgweVun9mwOXnq\nqaf0ySef6KqrrlKzZs308ssvq3HjxkpJSXGsnTt3rpYvX66srCzdeOONGjJkiLZu3apnn33Wsdbf\nJ4K/G1GpfmxI/d2ISoGzIfV4PNq4caNuueUW730bNmwod+2BikyZMkW7d+9W165ddeONNzquf0ZZ\nWZl27Nihjh07Kjo6Wg888IDGjx+voqIix9rs7Gx16tRJL730kqTTv9OSkhLHuoEDByoiIkLz5s3T\n5MmT1ahRI7Vu3dqxbu/evXr55ZcVHBysbdu26dprr9XWrVvPegXyXEJDQ/Xhhx+qW7duevvtt9Wj\nRw998sknjv+xbtCggZo3b65p06bp8OHD+vDDD/XCCy9o7969Sk9Pr7CutLRU69at09GjR3Xo0CHt\n2rVLbrdbpaWljrMWFxdr8+bNuummm7Rx40YFBwfr2LFjOnnyZKV15/r7+fLLL336+zlLlV/ABi4Q\nw4cPNxs2bCh335dffunTe/z33nuv2bx5c5V7jhgxwmzfvt17+/333zcjRowwd9xxh+OsGzduNMYY\n73tVe/fuNQkJCT71/frrr819991ndu3a5dOBhMYYExsba1566SUzZswY8/XXXxtjjNmyZYtjz/Hj\nx5sPPvjAvPTSS+att94yR48eNe+88475zW9+49jzp7MdOnTI/P3vfzcTJkwwAwYMqLRu2LBhZt26\ndeadd94xMTEx5l//+pc5cOCATz+fu+66y3t8yYYNG8w999xjjh49am6//fYKa/Ly8sy4ceNM165d\nTWxsrOnWrZsZN25cud9tZQ4fPmz2799f7r7i4uJKa7755hszatSocu97vv322yYmJqbSun/961/m\nt7/9bbn7xo0bZ7766iufZjXGmCNHjpgHH3zQ+3twmnXbtm1m+fLlZsaMGWblypXm+PHjZujQoT71\nPHTokJk8ebLp06ePue6660yXLl3M73//e/Of//yn0rqHH374nPefOnWq0rpvvvnGTJgwwSxatMi8\n9957pnPnzqZv377e51tlvv76azNkyBDTpUsXk5CQYHbv3m1eeuklk5mZWWndz/9+YmJizN133232\n7t3r2PPnCGEErDNPhNjYWPPf//3fpmvXrmbcuHHeA9kqc+jQIbNv374q99y2bZsZNWpUuYNifNmQ\n7ty50zz44IPlDtyo6ob08OHD5Takvsy6bNky8+STT5bbkJ4J5IpUtBH15UCXijakvsz629/+1ixc\nuLDKG9Lc3FwzePDgszaka9asqbDmzJkSvXr1KncUvy//wfH3LIuK6kaMGFEj/X5em56e7v2PZ1Vm\n9ffnU9WfbXX8XGtr1p+eLfPZZ5+Z7t27m549e1b6N1cRQhioBj89172m+2zZsqVWetVlQ4cONceO\nHfPrTAl/z7Lwt6etWf09k6S2e57vz8ef2jNny7z11ltVOlvmXKx+njBwPs51lPwZlR2w5O/R9dXd\n05e685m3Pszqb23Dhg3VvHlzSVU/U8Lfsyz87WlrVn/PJKntnuf78/Gn9qdny3zxxRc+ny1zTlWO\nbeACkZOTYwYMGGDy8vLM/v37y/2ribr60jOQZvW39tFHHzUpKSmmsLDQGGPMf/7zH9O3b1/TpUsX\nx37+1tZ2XX3paWPWKVOmmMcff7zcK2B/+ctfzEMPPeTY8+eCZ8yYMaPq0Q3Yd9lll6moqEilpaWK\niopS8+bNvf9qoq6+9AykWf2t7dGjhw4dOqQOHTqoYcOGCg0N1W233aZjx45Ven7x+dTWdl196Wlr\nVun02QZn7N+/Xw888ECVj5C+4D7AAQCA+iLIeRUAAFATCGEAACwhhIFzmDx5slauXFnrfefPn6/V\nq1fXet+KrF69WvPnz7fWf+XKlZo8ebK1/kBN4xQl4ALy0EMP2R6hnF69eqlXr162xwDqLEIY0Olz\nWVNTU7V27VpdeumlKisrU0xMjJ577jl9/vnnOnbsmMLCwrRgwQKtXbtWX3zxhff6zQsXLlRISIhu\nuOEGzZs3T5LUokULPfPMM2rZsuU5+/344496/PHHtXPnTknSiBEjNGzYME2ePFkxMTGKiYnRhAkT\n1KFDB33zzTe6+OKLNX/+fF100UVKT0/X4sWL1aBBA91www2aNWuWSkpKlJycrJ07d6qsrEz33Xef\nBgwYUOH3u379ei1YsEAul0vfffedbrzxRs2ZM0c//PCD7r33XoWFhalRo0YaNGiQvvzyS6Wmpion\nJ0dz5sxRcXGxwsLClJycrHbt2ikxMVEtWrTQzp079fzzz+sXv/jFOXv6Wr9jxw4tXrxYbrdbrVu3\nVtOmTSVJW7Zs0dy5c3Xq1CmFhYVp5syZatOmjc/9gQtSlU9qAuqgDz74wIwaNcqUlJSYQ4cOmS5d\nuphly5aZCRMmeM8FfPTRR81f//pXU1BQYDp37mwKCgqMx+Mxffr0MQcOHDCjRo3yXhbwlVdeMZ9+\n+mmF/davX+/9zOTDhw+bSZMmGWOMmTRpklmxYoXZt2+f6dixo/cSkxMmTDCvvvqqOXDggOncubP5\n7rvvjDHGPPLII+bjjz828+bNM6+88ooxxpgTJ06Y/v37m2+//bbC/l988YW54YYbzK5du4zH4zG/\n+93vzIsvvmj27dtnIiMjvZf0XLFihZk0aZIpLi42PXr08H5/q1at8n4+9ahRo8yf/vSnSn++vtYf\nOHDAdOnSxeTn55sff/zR3H333d7+AwcO9F46c926dWbMmDE+9wcuVOwJAzr9CSh9+vRRw4YN1bJl\nS3Xt2lXBwcGaNGmSli9frj179ignJ0dt27ZVs2bN1K1bN3300Udq06aN2rRpo1atWqlXr16aMGGC\nevfurV69eqlLly4V9uvQoYP27Nmje+65R127dtUjjzxy1joXX3yxrr32Wu/6x44d06ZNmxQdHa3L\nLrtMkrx73i+88IJOnTqlFStWSJKKioq0c+dO7+dJn8svf/lLtW/fXpJ0++23a9myZYqLi9PFF1+s\nK6+8sty6e/fuVfPmzb2fGtS3b19Nnz5dJ06ckCTHTxPytX7Tpk26+eabvR8ROnDgQH3xxRfau3ev\n9u3bp/Hjx3sfs6CgwPt1VT7NCLiQEMKATn+Mmsfj8d52uVw6evSo7rnnHo0dO1a33XabgoKCvB9Z\nOGTIEC1evFhXXnml9yMgx44dqx49eigzM1Pz5s3Tli1byoXGT4WFhen9999XVlaWPvnkE9155516\n//33y63TqFGjcvMZY866LN7hw4clnb6M3rx583TddddJkg4ePOi9BGBFfvpZ3MYY7+3GjRufte5P\nfzY/rSkrK6uwxp/6c/0eztRfeeWVeueddySd/ljAgwcPetdz6g9cqDg6GpDUuXNnffjhhyopKdGx\nY8f06aefqkGDBoqJidHw4cN1zTXXKCsryxsat9xyiw4cOKD169erd+/ekqShQ4eqsLBQY8eO1dix\nY7Vt27YK+61evVqPPPKIunfvrmnTpqlp06b67rvvHOe84YYbtHnzZuXn50s6/RnZq1ev1q233qrX\nX39dkvTDDz9o0KBBjo+XnZ2t77//Xh6PR2+//XalVwhq3769jh49qi1btkiSVq1apSuuuEIXXXSR\n48xVqe/UqZM2b97snWvVqlXe+mPHjmnjxo2SpBUrVpzz1QMg0LAnDEjq3bu3tm7dqgEDBuiSSy5R\nRESETp06pe3bt2vgwIFq2LChOnbsqP3795erOXbsmEJCQiRJEydO1OTJk+VyudSoUSPNnDmzwn5d\nu3bVP/7xD/Xv31+NGjVSnz591LFjR8c5W7VqpalTp+qee+6Rx+NRVFSUBg8erJMnT2rGjBkaMGCA\nysrK9Oijj6pt27aVPtall16qxx57TN9//726dOmioUOHVhjcISEheu655zRr1iydPHlSLVq00HPP\nPec4b1XrL7nkEk2bNk1jx45VkyZNdM0113jr58+f7z2wy+1266mnnvK5P3Ch4rKVQBUZY/Tjjz9q\n7Nixmjp1qvcl4ECyfv16LVy4UEuXLrU9ClCvsScMVFF+fr769++voUOHVhrAp06dUnx8/DmX/f73\nv6/x8283btyoWbNmnXPZvffeWyM9ExMTdfz48bPuT0hI0PDhw2ukJxDI2BMGAMASDswCAMASQhgA\nAEsIYQAALCGEAQCwhBAGAMASQhgAAEv+HxWiAOpnccF+AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x13d38af50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#### Artículos agregados primero al carrito\n",
    "added_first = dfOrderTrain\n",
    "added_first = added_first[['product_id','add_to_cart_order']].groupby(['product_id','add_to_cart_order']).agg({'product_id':'count'}).rename(columns={'product_id':'product_count'})\n",
    "added_first['pct'] = added_first.groupby(level=0).apply(lambda x: x/float(x.sum()))\n",
    "added_first = added_first.reset_index(level=['product_id','add_to_cart_order'])\n",
    "added_first = added_first[added_first['add_to_cart_order']==1]\n",
    "added_first = added_first[added_first['product_count'] > 10].sort_values(by=['pct'],ascending=[False])\n",
    "added_first = pd.merge(added_first,dfProducts,how='left',on='product_id')\n",
    "print (added_first.head(20))\n",
    "\n",
    "#### Asociacón entre el tiempo de la úlitma orden y probabilidad de reorden\n",
    "\n",
    "reorder_assoc = pd.merge(dfOrder[['order_id','days_since_prior_order']],dfOrderTrain,how='left',on='order_id')\n",
    "reorder_assoc = reorder_assoc.dropna(subset=['product_id','days_since_prior_order'])\n",
    "reorder_assoc = reorder_assoc[['days_since_prior_order','reordered']].groupby('days_since_prior_order').mean()\n",
    "reorder_assoc.plot(kind='bar')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}