data_allocation_dask-cuda.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Objective: predict 28 days, from d_1942 to d_1969"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table style=\"border: 2px solid white;\">\n",
       "<tr>\n",
       "<td style=\"vertical-align: top; border: 0px solid white\">\n",
       "<h3 style=\"text-align: left;\">Client</h3>\n",
       "<ul style=\"text-align: left; list-style: none; margin: 0; padding: 0;\">\n",
       "  <li><b>Scheduler: </b>tcp://127.0.0.1:41765</li>\n",
       "  <li><b>Dashboard: </b><a href='http://127.0.0.1:8787/status' target='_blank'>http://127.0.0.1:8787/status</a></li>\n",
       "</ul>\n",
       "</td>\n",
       "<td style=\"vertical-align: top; border: 0px solid white\">\n",
       "<h3 style=\"text-align: left;\">Cluster</h3>\n",
       "<ul style=\"text-align: left; list-style:none; margin: 0; padding: 0;\">\n",
       "  <li><b>Workers: </b>1</li>\n",
       "  <li><b>Cores: </b>1</li>\n",
       "  <li><b>Memory: </b>16.66 GB</li>\n",
       "</ul>\n",
       "</td>\n",
       "</tr>\n",
       "</table>"
      ],
      "text/plain": [
       "<Client: 'tcp://127.0.0.1:41765' processes=1 threads=1, memory=16.66 GB>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "import gc\n",
    "\n",
    "import dask\n",
    "import dask.dataframe as dd\n",
    "from dask_cuda import LocalCUDACluster\n",
    "from dask.distributed import Client\n",
    "\n",
    "# Create a Dask Cluster with one worker per GPU\n",
    "cluster = LocalCUDACluster()\n",
    "client = Client(cluster)\n",
    "client"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'temporary-directory': None,\n",
       " 'dataframe': {'shuffle-compression': None},\n",
       " 'array': {'svg': {'size': 120},\n",
       "  'chunk-size': '128MiB',\n",
       "  'rechunk-threshold': 4},\n",
       " 'optimization': {'fuse': {'active': True,\n",
       "   'ave-width': 1,\n",
       "   'max-width': None,\n",
       "   'max-height': inf,\n",
       "   'max-depth-new-edges': None,\n",
       "   'subgraphs': None,\n",
       "   'rename-keys': True}},\n",
       " 'distributed': {'version': 2,\n",
       "  'scheduler': {'allowed-failures': 3,\n",
       "   'bandwidth': 100000000,\n",
       "   'blocked-handlers': [],\n",
       "   'default-data-size': '1kiB',\n",
       "   'events-cleanup-delay': '1h',\n",
       "   'idle-timeout': None,\n",
       "   'transition-log-length': 100000,\n",
       "   'work-stealing': True,\n",
       "   'work-stealing-interval': '100ms',\n",
       "   'worker-ttl': None,\n",
       "   'pickle': True,\n",
       "   'preload': [],\n",
       "   'preload-argv': [],\n",
       "   'unknown-task-duration': '500ms',\n",
       "   'default-task-durations': {'rechunk-split': '1us', 'shuffle-split': '1us'},\n",
       "   'validate': False,\n",
       "   'dashboard': {'status': {'task-stream-length': 1000},\n",
       "    'tasks': {'task-stream-length': 100000},\n",
       "    'tls': {'ca-file': None, 'key': None, 'cert': None},\n",
       "    'bokeh-application': {'allow_websocket_origin': ['*'],\n",
       "     'keep_alive_milliseconds': 500,\n",
       "     'check_unused_sessions_milliseconds': 500}},\n",
       "   'locks': {'lease-validation-interval': '10s', 'lease-timeout': '30s'},\n",
       "   'http': {'routes': ['distributed.http.scheduler.prometheus',\n",
       "     'distributed.http.scheduler.info',\n",
       "     'distributed.http.scheduler.json',\n",
       "     'distributed.http.health',\n",
       "     'distributed.http.proxy',\n",
       "     'distributed.http.statics']}},\n",
       "  'worker': {'blocked-handlers': [],\n",
       "   'multiprocessing-method': 'spawn',\n",
       "   'use-file-locking': True,\n",
       "   'connections': {'outgoing': 50, 'incoming': 10},\n",
       "   'preload': [],\n",
       "   'preload-argv': [],\n",
       "   'daemon': True,\n",
       "   'validate': False,\n",
       "   'lifetime': {'duration': None, 'stagger': '0 seconds', 'restart': False},\n",
       "   'profile': {'interval': '10ms', 'cycle': '1000ms', 'low-level': False},\n",
       "   'memory': {'target': 0.6, 'spill': 0.7, 'pause': 0.8, 'terminate': 0.95},\n",
       "   'http': {'routes': ['distributed.http.worker.prometheus',\n",
       "     'distributed.http.health',\n",
       "     'distributed.http.statics']}},\n",
       "  'nanny': {'preload': [], 'preload-argv': []},\n",
       "  'client': {'heartbeat': '5s', 'scheduler-info-interval': '2s'},\n",
       "  'deploy': {'lost-worker-timeout': '15s', 'cluster-repr-interval': '500ms'},\n",
       "  'adaptive': {'interval': '1s',\n",
       "   'target-duration': '5s',\n",
       "   'minimum': 0,\n",
       "   'maximum': inf,\n",
       "   'wait-count': 3},\n",
       "  'comm': {'retry': {'count': 0, 'delay': {'min': '1s', 'max': '20s'}},\n",
       "   'compression': 'auto',\n",
       "   'offload': '10MiB',\n",
       "   'default-scheme': 'tcp',\n",
       "   'socket-backlog': 2048,\n",
       "   'recent-messages-log-length': 0,\n",
       "   'zstd': {'level': 3, 'threads': 0},\n",
       "   'timeouts': {'connect': '10s', 'tcp': '30s'},\n",
       "   'require-encryption': None,\n",
       "   'tls': {'ciphers': None,\n",
       "    'ca-file': None,\n",
       "    'scheduler': {'cert': None, 'key': None},\n",
       "    'worker': {'key': None, 'cert': None},\n",
       "    'client': {'key': None, 'cert': None}}},\n",
       "  'dashboard': {'link': '{scheme}://{host}:{port}/status',\n",
       "   'export-tool': False,\n",
       "   'graph-max-items': 5000},\n",
       "  'admin': {'tick': {'interval': '20ms', 'limit': '3s'},\n",
       "   'max-error-length': 10000,\n",
       "   'log-length': 10000,\n",
       "   'log-format': '%(name)s - %(levelname)s - %(message)s',\n",
       "   'pdb-on-err': False}},\n",
       " 'rmm': {'pool-size': None},\n",
       " 'ucx': {'tcp': None,\n",
       "  'nvlink': None,\n",
       "  'infiniband': None,\n",
       "  'rdmacm': None,\n",
       "  'cuda_copy': None,\n",
       "  'net-devices': None},\n",
       " 'scheduler': 'dask.distributed',\n",
       " 'shuffle': 'tasks'}"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dask.config.config"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load & Check Datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "30490 , 1947\n"
     ]
    },
    {
     "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>id</th>\n",
       "      <th>item_id</th>\n",
       "      <th>dept_id</th>\n",
       "      <th>cat_id</th>\n",
       "      <th>store_id</th>\n",
       "      <th>state_id</th>\n",
       "      <th>d_1</th>\n",
       "      <th>d_2</th>\n",
       "      <th>d_3</th>\n",
       "      <th>d_4</th>\n",
       "      <th>...</th>\n",
       "      <th>d_1932</th>\n",
       "      <th>d_1933</th>\n",
       "      <th>d_1934</th>\n",
       "      <th>d_1935</th>\n",
       "      <th>d_1936</th>\n",
       "      <th>d_1937</th>\n",
       "      <th>d_1938</th>\n",
       "      <th>d_1939</th>\n",
       "      <th>d_1940</th>\n",
       "      <th>d_1941</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>HOBBIES_1_001_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_001</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>2</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>HOBBIES_1_002_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_002</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>HOBBIES_1_003_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_003</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>HOBBIES_1_004_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_004</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>HOBBIES_1_005_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_005</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 1947 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                              id        item_id    dept_id   cat_id store_id  \\\n",
       "0  HOBBIES_1_001_CA_1_evaluation  HOBBIES_1_001  HOBBIES_1  HOBBIES     CA_1   \n",
       "1  HOBBIES_1_002_CA_1_evaluation  HOBBIES_1_002  HOBBIES_1  HOBBIES     CA_1   \n",
       "2  HOBBIES_1_003_CA_1_evaluation  HOBBIES_1_003  HOBBIES_1  HOBBIES     CA_1   \n",
       "3  HOBBIES_1_004_CA_1_evaluation  HOBBIES_1_004  HOBBIES_1  HOBBIES     CA_1   \n",
       "4  HOBBIES_1_005_CA_1_evaluation  HOBBIES_1_005  HOBBIES_1  HOBBIES     CA_1   \n",
       "\n",
       "  state_id  d_1  d_2  d_3  d_4  ...  d_1932  d_1933  d_1934  d_1935  d_1936  \\\n",
       "0       CA    0    0    0    0  ...       2       4       0       0       0   \n",
       "1       CA    0    0    0    0  ...       0       1       2       1       1   \n",
       "2       CA    0    0    0    0  ...       1       0       2       0       0   \n",
       "3       CA    0    0    0    0  ...       1       1       0       4       0   \n",
       "4       CA    0    0    0    0  ...       0       0       0       2       1   \n",
       "\n",
       "   d_1937  d_1938  d_1939  d_1940  d_1941  \n",
       "0       0       3       3       0       1  \n",
       "1       0       0       0       0       0  \n",
       "2       0       2       3       0       1  \n",
       "3       1       3       0       2       6  \n",
       "4       0       0       2       1       0  \n",
       "\n",
       "[5 rows x 1947 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sales = dd.read_csv(\"INPUT/sales_train_evaluation.csv\")\n",
    "print(len(sales),\",\", len(sales.columns))\n",
    "sales.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#as they ask us to predict 1942-1969 days we need this days in sales\n",
    "for d in range(1942,1970):\n",
    "    sales[\"d_\" + str(d)] = int(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sales.isnull().sum().compute().any()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1969 , 14\n"
     ]
    },
    {
     "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>date</th>\n",
       "      <th>wm_yr_wk</th>\n",
       "      <th>weekday</th>\n",
       "      <th>wday</th>\n",
       "      <th>month</th>\n",
       "      <th>year</th>\n",
       "      <th>d</th>\n",
       "      <th>event_name_1</th>\n",
       "      <th>event_type_1</th>\n",
       "      <th>event_name_2</th>\n",
       "      <th>event_type_2</th>\n",
       "      <th>snap_CA</th>\n",
       "      <th>snap_TX</th>\n",
       "      <th>snap_WI</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2011-01-29</td>\n",
       "      <td>11101</td>\n",
       "      <td>Saturday</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>2011</td>\n",
       "      <td>d_1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2011-01-30</td>\n",
       "      <td>11101</td>\n",
       "      <td>Sunday</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>2011</td>\n",
       "      <td>d_2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2011-01-31</td>\n",
       "      <td>11101</td>\n",
       "      <td>Monday</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>2011</td>\n",
       "      <td>d_3</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2011-02-01</td>\n",
       "      <td>11101</td>\n",
       "      <td>Tuesday</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>2011</td>\n",
       "      <td>d_4</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2011-02-02</td>\n",
       "      <td>11101</td>\n",
       "      <td>Wednesday</td>\n",
       "      <td>5</td>\n",
       "      <td>2</td>\n",
       "      <td>2011</td>\n",
       "      <td>d_5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date  wm_yr_wk    weekday  wday  month  year    d event_name_1  \\\n",
       "0  2011-01-29     11101   Saturday     1      1  2011  d_1          NaN   \n",
       "1  2011-01-30     11101     Sunday     2      1  2011  d_2          NaN   \n",
       "2  2011-01-31     11101     Monday     3      1  2011  d_3          NaN   \n",
       "3  2011-02-01     11101    Tuesday     4      2  2011  d_4          NaN   \n",
       "4  2011-02-02     11101  Wednesday     5      2  2011  d_5          NaN   \n",
       "\n",
       "  event_type_1 event_name_2 event_type_2  snap_CA  snap_TX  snap_WI  \n",
       "0          NaN          NaN          NaN        0        0        0  \n",
       "1          NaN          NaN          NaN        0        0        0  \n",
       "2          NaN          NaN          NaN        0        0        0  \n",
       "3          NaN          NaN          NaN        1        1        0  \n",
       "4          NaN          NaN          NaN        1        0        1  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "calendar = dd.read_csv(\"INPUT/calendar.csv\")\n",
    "print(len(calendar),\",\", len(calendar.columns))\n",
    "calendar.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "date               0\n",
       "wm_yr_wk           0\n",
       "weekday            0\n",
       "wday               0\n",
       "month              0\n",
       "year               0\n",
       "d                  0\n",
       "event_name_1    1807\n",
       "event_type_1    1807\n",
       "event_name_2    1964\n",
       "event_type_2    1964\n",
       "snap_CA            0\n",
       "snap_TX            0\n",
       "snap_WI            0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "calendar.isnull().sum().compute()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "calendar = calendar.fillna(\"no event\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "6841121 , 4\n"
     ]
    },
    {
     "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>store_id</th>\n",
       "      <th>item_id</th>\n",
       "      <th>wm_yr_wk</th>\n",
       "      <th>sell_price</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CA_1</td>\n",
       "      <td>HOBBIES_1_001</td>\n",
       "      <td>11325</td>\n",
       "      <td>9.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>CA_1</td>\n",
       "      <td>HOBBIES_1_001</td>\n",
       "      <td>11326</td>\n",
       "      <td>9.58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CA_1</td>\n",
       "      <td>HOBBIES_1_001</td>\n",
       "      <td>11327</td>\n",
       "      <td>8.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>CA_1</td>\n",
       "      <td>HOBBIES_1_001</td>\n",
       "      <td>11328</td>\n",
       "      <td>8.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>CA_1</td>\n",
       "      <td>HOBBIES_1_001</td>\n",
       "      <td>11329</td>\n",
       "      <td>8.26</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  store_id        item_id  wm_yr_wk  sell_price\n",
       "0     CA_1  HOBBIES_1_001     11325        9.58\n",
       "1     CA_1  HOBBIES_1_001     11326        9.58\n",
       "2     CA_1  HOBBIES_1_001     11327        8.26\n",
       "3     CA_1  HOBBIES_1_001     11328        8.26\n",
       "4     CA_1  HOBBIES_1_001     11329        8.26"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "price = dd.read_csv(\"INPUT/sell_prices.csv\")\n",
    "print(len(price),\",\", len(price.columns))\n",
    "price.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "store_id      0\n",
       "item_id       0\n",
       "wm_yr_wk      0\n",
       "sell_price    0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "price.isnull().sum().compute()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Downcasting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dask.dataframe.core.DataFrame'>\n",
      "Columns: 1975 entries, id to d_1969\n",
      "dtypes: object(6), int64(1969)"
     ]
    }
   ],
   "source": [
    "sales.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dask.dataframe.core.DataFrame'>\n",
      "Columns: 14 entries, date to snap_WI\n",
      "dtypes: object(7), int64(7)"
     ]
    }
   ],
   "source": [
    "calendar.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dask.dataframe.core.DataFrame'>\n",
      "Columns: 4 entries, store_id to sell_price\n",
      "dtypes: object(2), float64(1), int64(1)"
     ]
    }
   ],
   "source": [
    "price.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "sales = sales.compute()\n",
    "calendar = calendar.compute()\n",
    "price = price.compute()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Memory in: 470.04 MB\n",
      "Starting integers.\n",
      "Starting objects.\n",
      "Memory out: 68.26 MB\n",
      "Reduction: 85.5%\n",
      "Memory in: 0.96 MB\n",
      "Starting integers.\n",
      "Starting objects.\n",
      "Memory out: 0.27 MB\n",
      "Reduction: 71.3%\n",
      "Memory in: 1009.71 MB\n",
      "Starting integers.\n",
      "Starting objects.\n",
      "Memory out: 137.29 MB\n",
      "Reduction: 86.4%\n"
     ]
    }
   ],
   "source": [
    "from skmem import MemReducer\n",
    "\n",
    "mr = MemReducer()\n",
    "\n",
    "sales = mr.fit_transform(sales)\n",
    "calendar = mr.fit_transform(calendar)\n",
    "price = mr.fit_transform(price)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "sales = dd.from_pandas(sales, npartitions=8)\n",
    "calendar = dd.from_pandas(calendar, npartitions=8)\n",
    "price = dd.from_pandas(price, npartitions=8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dask.dataframe.core.DataFrame'>\n",
      "Columns: 1975 entries, id to d_1969\n",
      "dtypes: category(1), category(1), category(1), category(1), category(1), object(1), uint16(269), uint8(1700)"
     ]
    }
   ],
   "source": [
    "sales.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dask.dataframe.core.DataFrame'>\n",
      "Columns: 14 entries, date to snap_WI\n",
      "dtypes: category(1), category(1), category(1), category(1), category(1), datetime64[ns](1), object(1), uint16(2), uint8(5)"
     ]
    }
   ],
   "source": [
    "#we will change date to datetime, otherwise pyplot will take it as a string a won't plot it as time series\n",
    "calendar.date = dd.to_datetime(calendar.date)\n",
    "\n",
    "calendar.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dask.dataframe.core.DataFrame'>\n",
      "Columns: 4 entries, store_id to sell_price\n",
      "dtypes: category(1), category(1), float64(1), uint16(1)"
     ]
    }
   ],
   "source": [
    "price.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Melting & Merging "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60034810 , 8\n"
     ]
    },
    {
     "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>id</th>\n",
       "      <th>item_id</th>\n",
       "      <th>dept_id</th>\n",
       "      <th>cat_id</th>\n",
       "      <th>store_id</th>\n",
       "      <th>state_id</th>\n",
       "      <th>d</th>\n",
       "      <th>sales</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>HOBBIES_1_001_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_001</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>d_1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>HOUSEHOLD_1_220_TX_2_evaluation</td>\n",
       "      <td>HOUSEHOLD_1_220</td>\n",
       "      <td>HOUSEHOLD_1</td>\n",
       "      <td>HOUSEHOLD</td>\n",
       "      <td>TX_2</td>\n",
       "      <td>TX</td>\n",
       "      <td>d_1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>HOUSEHOLD_1_221_TX_2_evaluation</td>\n",
       "      <td>HOUSEHOLD_1_221</td>\n",
       "      <td>HOUSEHOLD_1</td>\n",
       "      <td>HOUSEHOLD</td>\n",
       "      <td>TX_2</td>\n",
       "      <td>TX</td>\n",
       "      <td>d_1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>HOBBIES_1_002_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_002</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>d_1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>HOBBIES_1_003_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_003</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>d_1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                id          item_id      dept_id     cat_id  \\\n",
       "0    HOBBIES_1_001_CA_1_evaluation    HOBBIES_1_001    HOBBIES_1    HOBBIES   \n",
       "1  HOUSEHOLD_1_220_TX_2_evaluation  HOUSEHOLD_1_220  HOUSEHOLD_1  HOUSEHOLD   \n",
       "2  HOUSEHOLD_1_221_TX_2_evaluation  HOUSEHOLD_1_221  HOUSEHOLD_1  HOUSEHOLD   \n",
       "3    HOBBIES_1_002_CA_1_evaluation    HOBBIES_1_002    HOBBIES_1    HOBBIES   \n",
       "4    HOBBIES_1_003_CA_1_evaluation    HOBBIES_1_003    HOBBIES_1    HOBBIES   \n",
       "\n",
       "  store_id state_id    d  sales  \n",
       "0     CA_1       CA  d_1      0  \n",
       "1     TX_2       TX  d_1      0  \n",
       "2     TX_2       TX  d_1      0  \n",
       "3     CA_1       CA  d_1      0  \n",
       "4     CA_1       CA  d_1      0  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#melt will take all id_vars, and the rest will be transpossed in 1 column, which in this case is perfect\n",
    "sales = dd.melt(sales, id_vars = ['id', 'item_id', 'dept_id', 'cat_id', 'store_id', \n",
    "                                           'state_id'], var_name = 'd', value_name = 'sales')\n",
    "print(len(sales),\",\",len(sales.columns))\n",
    "sales.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "sales_cal = dd.merge(sales, calendar, on=\"d\", how=\"left\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = dd.merge(sales_cal, price, on=[\"store_id\", \"item_id\", \"wm_yr_wk\"], how=\"left\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Garbage Collector collected: 104 elements\n"
     ]
    }
   ],
   "source": [
    "#freeing up space\n",
    "del sales, calendar, price, sales_cal\n",
    "collected = gc.collect()\n",
    "print(\"Garbage Collector collected:\", collected, \"elements\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "60034810 , 22\n",
      "CPU times: user 3.05 s, sys: 392 ms, total: 3.44 s\n",
      "Wall time: 2min 27s\n"
     ]
    },
    {
     "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>id</th>\n",
       "      <th>item_id</th>\n",
       "      <th>dept_id</th>\n",
       "      <th>cat_id</th>\n",
       "      <th>store_id</th>\n",
       "      <th>state_id</th>\n",
       "      <th>d</th>\n",
       "      <th>sales</th>\n",
       "      <th>date</th>\n",
       "      <th>wm_yr_wk</th>\n",
       "      <th>...</th>\n",
       "      <th>month</th>\n",
       "      <th>year</th>\n",
       "      <th>event_name_1</th>\n",
       "      <th>event_type_1</th>\n",
       "      <th>event_name_2</th>\n",
       "      <th>event_type_2</th>\n",
       "      <th>snap_CA</th>\n",
       "      <th>snap_TX</th>\n",
       "      <th>snap_WI</th>\n",
       "      <th>sell_price</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>HOUSEHOLD_1_223_TX_2_evaluation</td>\n",
       "      <td>HOUSEHOLD_1_223</td>\n",
       "      <td>HOUSEHOLD_1</td>\n",
       "      <td>HOUSEHOLD</td>\n",
       "      <td>TX_2</td>\n",
       "      <td>TX</td>\n",
       "      <td>d_2</td>\n",
       "      <td>0</td>\n",
       "      <td>2011-01-30</td>\n",
       "      <td>11101</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2011</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>HOUSEHOLD_1_224_TX_2_evaluation</td>\n",
       "      <td>HOUSEHOLD_1_224</td>\n",
       "      <td>HOUSEHOLD_1</td>\n",
       "      <td>HOUSEHOLD</td>\n",
       "      <td>TX_2</td>\n",
       "      <td>TX</td>\n",
       "      <td>d_2</td>\n",
       "      <td>0</td>\n",
       "      <td>2011-01-30</td>\n",
       "      <td>11101</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2011</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>HOBBIES_1_008_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_008</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>d_2</td>\n",
       "      <td>15</td>\n",
       "      <td>2011-01-30</td>\n",
       "      <td>11101</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2011</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>HOBBIES_1_011_CA_1_evaluation</td>\n",
       "      <td>HOBBIES_1_011</td>\n",
       "      <td>HOBBIES_1</td>\n",
       "      <td>HOBBIES</td>\n",
       "      <td>CA_1</td>\n",
       "      <td>CA</td>\n",
       "      <td>d_2</td>\n",
       "      <td>0</td>\n",
       "      <td>2011-01-30</td>\n",
       "      <td>11101</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2011</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>HOUSEHOLD_1_235_TX_2_evaluation</td>\n",
       "      <td>HOUSEHOLD_1_235</td>\n",
       "      <td>HOUSEHOLD_1</td>\n",
       "      <td>HOUSEHOLD</td>\n",
       "      <td>TX_2</td>\n",
       "      <td>TX</td>\n",
       "      <td>d_2</td>\n",
       "      <td>2</td>\n",
       "      <td>2011-01-30</td>\n",
       "      <td>11101</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>2011</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>no event</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>3.28</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 22 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                id          item_id      dept_id     cat_id  \\\n",
       "0  HOUSEHOLD_1_223_TX_2_evaluation  HOUSEHOLD_1_223  HOUSEHOLD_1  HOUSEHOLD   \n",
       "1  HOUSEHOLD_1_224_TX_2_evaluation  HOUSEHOLD_1_224  HOUSEHOLD_1  HOUSEHOLD   \n",
       "2    HOBBIES_1_008_CA_1_evaluation    HOBBIES_1_008    HOBBIES_1    HOBBIES   \n",
       "3    HOBBIES_1_011_CA_1_evaluation    HOBBIES_1_011    HOBBIES_1    HOBBIES   \n",
       "4  HOUSEHOLD_1_235_TX_2_evaluation  HOUSEHOLD_1_235  HOUSEHOLD_1  HOUSEHOLD   \n",
       "\n",
       "  store_id state_id    d  sales       date  wm_yr_wk  ... month  year  \\\n",
       "0     TX_2       TX  d_2      0 2011-01-30     11101  ...     1  2011   \n",
       "1     TX_2       TX  d_2      0 2011-01-30     11101  ...     1  2011   \n",
       "2     CA_1       CA  d_2     15 2011-01-30     11101  ...     1  2011   \n",
       "3     CA_1       CA  d_2      0 2011-01-30     11101  ...     1  2011   \n",
       "4     TX_2       TX  d_2      2 2011-01-30     11101  ...     1  2011   \n",
       "\n",
       "   event_name_1  event_type_1 event_name_2 event_type_2 snap_CA snap_TX  \\\n",
       "0      no event      no event     no event     no event       0       0   \n",
       "1      no event      no event     no event     no event       0       0   \n",
       "2      no event      no event     no event     no event       0       0   \n",
       "3      no event      no event     no event     no event       0       0   \n",
       "4      no event      no event     no event     no event       0       0   \n",
       "\n",
       "   snap_WI  sell_price  \n",
       "0        0         NaN  \n",
       "1        0        8.97  \n",
       "2        0        0.46  \n",
       "3        0         NaN  \n",
       "4        0        3.28  \n",
       "\n",
       "[5 rows x 22 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(len(df),\",\",len(df.columns))\n",
    "%time df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'dask.dataframe.core.DataFrame'>\n",
      "Columns: 22 entries, id to sell_price\n",
      "dtypes: category(1), category(1), category(1), category(1), category(1), category(1), category(1), category(1), category(1), category(1), datetime64[ns](1), object(2), float64(1), uint16(3), uint8(5)"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Choosing Final Distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "FOODS        28294530\n",
       "HOUSEHOLD    20615430\n",
       "HOBBIES      11124850\n",
       "Name: cat_id, dtype: int64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat = df.cat_id.value_counts().compute()\n",
    "cat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,5))\n",
    "plt.pie(cat, labels=cat.keys(), autopct='%1.1f%%')\n",
    "plt.title(\"Categories Distribution\")\n",
    "plt.legend(loc=2)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 4 µs, sys: 0 ns, total: 4 µs\n",
      "Wall time: 8.11 µs\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "WI_3    6003481\n",
       "WI_2    6003481\n",
       "WI_1    6003481\n",
       "TX_3    6003481\n",
       "TX_2    6003481\n",
       "TX_1    6003481\n",
       "CA_4    6003481\n",
       "CA_3    6003481\n",
       "CA_2    6003481\n",
       "CA_1    6003481\n",
       "Name: store_id, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "store = df.store_id.value_counts().compute()\n",
    "%time store"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,5))\n",
    "plt.pie(store, labels=store.keys(), autopct='%1.1f%%')\n",
    "plt.title(\"Stores Distribution\")\n",
    "plt.legend(loc=2)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CA    24013924\n",
       "WI    18010443\n",
       "TX    18010443\n",
       "Name: state_id, dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "state = df.state_id.value_counts().compute()\n",
    "state"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,5))\n",
    "plt.pie(state, labels=state.keys(), autopct='%1.1f%%')\n",
    "plt.title(\"States Distribution\")\n",
    "plt.legend(loc=2)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Garbage Collector collected: 2385 elements\n"
     ]
    }
   ],
   "source": [
    "del cat, state\n",
    "collected = gc.collect()\n",
    "print(\"Garbage Collector collected:\", collected, \"elements\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will divide the final df based in the stores distribution, as it is the most equitable and also we will be able to work fine with the memory usage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving Dataset\n",
      "Saving Dataset\n",
      "Saving Dataset\n",
      "Saving Dataset\n",
      "Saving Dataset\n",
      "Saving Dataset\n",
      "Saving Dataset\n",
      "Saving Dataset\n",
      "Saving Dataset\n",
      "Saving Dataset\n",
      "df_WI_3 size: 60034810 , 22\n",
      "df_WI_2 size: 60034810 , 22\n",
      "df_WI_1 size: 60034810 , 22\n",
      "df_TX_3 size: 60034810 , 22\n",
      "df_TX_2 size: 60034810 , 22\n",
      "df_TX_1 size: 60034810 , 22\n",
      "df_CA_4 size: 60034810 , 22\n",
      "df_CA_3 size: 60034810 , 22\n",
      "df_CA_2 size: 60034810 , 22\n",
      "df_CA_1 size: 60034810 , 22\n"
     ]
    }
   ],
   "source": [
    "dfs_dict = {}\n",
    "for s in store.keys():\n",
    "    print(\"Saving Dataset\")\n",
    "    dfs_dict[\"df_\" + s] = df[df.store_id==s]\n",
    "\n",
    "for names, values in dfs_dict.items():\n",
    "    print(names, \"size:\", len(df),\",\",len(df.columns))\n",
    "    values.to_csv(f\"OUTPUT/d_{names}.csv\", single_file=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "435\n"
     ]
    }
   ],
   "source": [
    "del df\n",
    "collected = gc.collect()\n",
    "print(collected)"
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}