Jenny_Problem1_Graphs.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = pd.read_csv(\"train.csv\")\n",
    "#val = pd.read_csv(\"validation.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "train.insert(0, 'Imps', 1)\n",
    "train['clickcost'] = train['click']*train['payprice']\n",
    "train['OS'],train['browser'] = train['useragent'].str.split('_', 1).str\n",
    "#train['slot_size']= train['slotwidth']*train['slotheight']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0          200\n",
       "1          250\n",
       "2          250\n",
       "3          600\n",
       "4           90\n",
       "5          280\n",
       "6           90\n",
       "7          600\n",
       "8           90\n",
       "9           90\n",
       "10         250\n",
       "11          90\n",
       "12         280\n",
       "13          90\n",
       "14         250\n",
       "15          90\n",
       "16         280\n",
       "17         280\n",
       "18          90\n",
       "19          90\n",
       "20          90\n",
       "21         600\n",
       "22         250\n",
       "23          90\n",
       "24          90\n",
       "25         600\n",
       "26         250\n",
       "27         250\n",
       "28         250\n",
       "29         250\n",
       "          ... \n",
       "2430951     90\n",
       "2430952    600\n",
       "2430953    280\n",
       "2430954    250\n",
       "2430955    280\n",
       "2430956     60\n",
       "2430957    250\n",
       "2430958     90\n",
       "2430959     90\n",
       "2430960    250\n",
       "2430961     90\n",
       "2430962    250\n",
       "2430963     90\n",
       "2430964    250\n",
       "2430965    250\n",
       "2430966    280\n",
       "2430967    250\n",
       "2430968     90\n",
       "2430969    250\n",
       "2430970     90\n",
       "2430971    250\n",
       "2430972     90\n",
       "2430973     90\n",
       "2430974    280\n",
       "2430975     90\n",
       "2430976    600\n",
       "2430977    280\n",
       "2430978    250\n",
       "2430979     90\n",
       "2430980    280\n",
       "Name: slotheight, Length: 2430981, dtype: object"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.slot_width = train.slotwidth.astype(str)\n",
    "train.slot_width\n",
    "train.slot_height = train.slotheight.astype(str)\n",
    "train.slot_height\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "slot_size = train.slot_width + '*' + train.slot_height\n",
    "train['Slot__Size'] = slot_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Imps</th>\n",
       "      <th>click</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "      <th>bidid</th>\n",
       "      <th>userid</th>\n",
       "      <th>useragent</th>\n",
       "      <th>IP</th>\n",
       "      <th>region</th>\n",
       "      <th>city</th>\n",
       "      <th>...</th>\n",
       "      <th>creative</th>\n",
       "      <th>bidprice</th>\n",
       "      <th>payprice</th>\n",
       "      <th>keypage</th>\n",
       "      <th>advertiser</th>\n",
       "      <th>usertag</th>\n",
       "      <th>clickcost</th>\n",
       "      <th>OS</th>\n",
       "      <th>browser</th>\n",
       "      <th>Slot__Size</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>22</td>\n",
       "      <td>b7bea80521fdecd95d2d761a38c91c3f09618066</td>\n",
       "      <td>2e880fb7d690cf7377b2e42e701728e3f3c0e4c1</td>\n",
       "      <td>windows_ie</td>\n",
       "      <td>125.37.175.*</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>...</td>\n",
       "      <td>a4f763f78ef3eedfe614263b94a8924e</td>\n",
       "      <td>238</td>\n",
       "      <td>5</td>\n",
       "      <td>0f951a030abdaedd733ee8d114ce2944</td>\n",
       "      <td>3427</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>ie</td>\n",
       "      <td>200*200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>20</td>\n",
       "      <td>4f51205475678f5a124bc76b2c54163bf8eaa7eb</td>\n",
       "      <td>3a1fe01360ff8100e7d006b83b77a3e4c01d928c</td>\n",
       "      <td>windows_chrome</td>\n",
       "      <td>171.36.92.*</td>\n",
       "      <td>238</td>\n",
       "      <td>239</td>\n",
       "      <td>...</td>\n",
       "      <td>10722</td>\n",
       "      <td>294</td>\n",
       "      <td>23</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2821</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>chrome</td>\n",
       "      <td>300*250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>13</td>\n",
       "      <td>b604e3fd054a658ab7ced4285ebf2ef54d2bd890</td>\n",
       "      <td>801d18a056b6fe6b06a794aef17fb0d6daff2414</td>\n",
       "      <td>windows_ie</td>\n",
       "      <td>59.46.106.*</td>\n",
       "      <td>40</td>\n",
       "      <td>41</td>\n",
       "      <td>...</td>\n",
       "      <td>798b2d49952d77f1eace9f23c210d0b5</td>\n",
       "      <td>238</td>\n",
       "      <td>24</td>\n",
       "      <td>0f951a030abdaedd733ee8d114ce2944</td>\n",
       "      <td>3427</td>\n",
       "      <td>10052,10006,13866,10110</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>ie</td>\n",
       "      <td>250*250</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>23</td>\n",
       "      <td>0348beeae93e561584c3b50fc9e7746a33048ad7</td>\n",
       "      <td>0d6eaf2259699990e38a1fc5116f112070b9ecdc</td>\n",
       "      <td>windows_ie</td>\n",
       "      <td>114.250.226.*</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>cb7c76e7784031272e37af8e7e9b062c</td>\n",
       "      <td>300</td>\n",
       "      <td>25</td>\n",
       "      <td>bebefa5efe83beee17a3d245e7c5085b</td>\n",
       "      <td>1458</td>\n",
       "      <td>13866,10063,10111</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>ie</td>\n",
       "      <td>160*600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>268149c1789bce2bc9798ffd97ec431219bafeb3</td>\n",
       "      <td>a239d9bb642460d974ba67f85e63b8d3e214da0e</td>\n",
       "      <td>windows_ie</td>\n",
       "      <td>183.63.192.*</td>\n",
       "      <td>216</td>\n",
       "      <td>233</td>\n",
       "      <td>...</td>\n",
       "      <td>7330</td>\n",
       "      <td>277</td>\n",
       "      <td>133</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2259</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>ie</td>\n",
       "      <td>728*90</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Imps  click  weekday  hour                                     bidid  \\\n",
       "0     1      0        5    22  b7bea80521fdecd95d2d761a38c91c3f09618066   \n",
       "1     1      0        1    20  4f51205475678f5a124bc76b2c54163bf8eaa7eb   \n",
       "2     1      0        3    13  b604e3fd054a658ab7ced4285ebf2ef54d2bd890   \n",
       "3     1      0        6    23  0348beeae93e561584c3b50fc9e7746a33048ad7   \n",
       "4     1      0        5     6  268149c1789bce2bc9798ffd97ec431219bafeb3   \n",
       "\n",
       "                                     userid       useragent             IP  \\\n",
       "0  2e880fb7d690cf7377b2e42e701728e3f3c0e4c1      windows_ie   125.37.175.*   \n",
       "1  3a1fe01360ff8100e7d006b83b77a3e4c01d928c  windows_chrome    171.36.92.*   \n",
       "2  801d18a056b6fe6b06a794aef17fb0d6daff2414      windows_ie    59.46.106.*   \n",
       "3  0d6eaf2259699990e38a1fc5116f112070b9ecdc      windows_ie  114.250.226.*   \n",
       "4  a239d9bb642460d974ba67f85e63b8d3e214da0e      windows_ie   183.63.192.*   \n",
       "\n",
       "   region  city     ...                              creative bidprice  \\\n",
       "0       2     2     ...      a4f763f78ef3eedfe614263b94a8924e      238   \n",
       "1     238   239     ...                                 10722      294   \n",
       "2      40    41     ...      798b2d49952d77f1eace9f23c210d0b5      238   \n",
       "3       1     1     ...      cb7c76e7784031272e37af8e7e9b062c      300   \n",
       "4     216   233     ...                                  7330      277   \n",
       "\n",
       "  payprice                           keypage advertiser  \\\n",
       "0        5  0f951a030abdaedd733ee8d114ce2944       3427   \n",
       "1       23                               NaN       2821   \n",
       "2       24  0f951a030abdaedd733ee8d114ce2944       3427   \n",
       "3       25  bebefa5efe83beee17a3d245e7c5085b       1458   \n",
       "4      133                               NaN       2259   \n",
       "\n",
       "                   usertag  clickcost       OS browser  Slot__Size  \n",
       "0                      NaN          0  windows      ie     200*200  \n",
       "1                      NaN          0  windows  chrome     300*250  \n",
       "2  10052,10006,13866,10110          0  windows      ie     250*250  \n",
       "3        13866,10063,10111          0  windows      ie     160*600  \n",
       "4                      NaN          0  windows      ie      728*90  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "train['CTR%'] = train['click']/train['Imps']*100\n",
    "train['CPM'] = train['payprice']/(train['Imps'])\n",
    "train['eCPC'] = train['payprice']/train['click']/1000\n",
    "train['eCost'] = train['clickcost']/(train['payprice'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0            2\n",
       "1          239\n",
       "2           41\n",
       "3            1\n",
       "4          233\n",
       "5          277\n",
       "6           89\n",
       "7           19\n",
       "8          342\n",
       "9            2\n",
       "10         309\n",
       "11         219\n",
       "12         301\n",
       "13          64\n",
       "14          79\n",
       "15           1\n",
       "16         387\n",
       "17         278\n",
       "18          32\n",
       "19         279\n",
       "20          91\n",
       "21          97\n",
       "22         219\n",
       "23          79\n",
       "24          89\n",
       "25         193\n",
       "26         217\n",
       "27         166\n",
       "28         145\n",
       "29         288\n",
       "          ... \n",
       "2430951     27\n",
       "2430952    289\n",
       "2430953     79\n",
       "2430954    190\n",
       "2430955     41\n",
       "2430956    334\n",
       "2430957    354\n",
       "2430958    187\n",
       "2430959    160\n",
       "2430960    315\n",
       "2430961    309\n",
       "2430962    276\n",
       "2430963     32\n",
       "2430964    334\n",
       "2430965    217\n",
       "2430966    161\n",
       "2430967    219\n",
       "2430968    343\n",
       "2430969     24\n",
       "2430970    217\n",
       "2430971    164\n",
       "2430972     57\n",
       "2430973    397\n",
       "2430974     73\n",
       "2430975      4\n",
       "2430976     79\n",
       "2430977    334\n",
       "2430978    308\n",
       "2430979      1\n",
       "2430980     99\n",
       "Name: city, Length: 2430981, dtype: object"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train['region'].apply(str)\n",
    "train['city'].apply(str)\n",
    "#df['A'].apply(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Imps</th>\n",
       "      <th>click</th>\n",
       "      <th>weekday</th>\n",
       "      <th>hour</th>\n",
       "      <th>region</th>\n",
       "      <th>city</th>\n",
       "      <th>adexchange</th>\n",
       "      <th>slotwidth</th>\n",
       "      <th>slotheight</th>\n",
       "      <th>slotprice</th>\n",
       "      <th>...</th>\n",
       "      <th>payprice</th>\n",
       "      <th>advertiser</th>\n",
       "      <th>clickcost</th>\n",
       "      <th>OS</th>\n",
       "      <th>browser</th>\n",
       "      <th>Slot__Size</th>\n",
       "      <th>CTR%</th>\n",
       "      <th>CPM</th>\n",
       "      <th>eCPC</th>\n",
       "      <th>eCost</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>22</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2.0</td>\n",
       "      <td>200</td>\n",
       "      <td>200</td>\n",
       "      <td>5</td>\n",
       "      <td>...</td>\n",
       "      <td>5</td>\n",
       "      <td>3427</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>ie</td>\n",
       "      <td>200*200</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>inf</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>20</td>\n",
       "      <td>238</td>\n",
       "      <td>239</td>\n",
       "      <td>1.0</td>\n",
       "      <td>300</td>\n",
       "      <td>250</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>23</td>\n",
       "      <td>2821</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>chrome</td>\n",
       "      <td>300*250</td>\n",
       "      <td>0.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>inf</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>13</td>\n",
       "      <td>40</td>\n",
       "      <td>41</td>\n",
       "      <td>2.0</td>\n",
       "      <td>250</td>\n",
       "      <td>250</td>\n",
       "      <td>5</td>\n",
       "      <td>...</td>\n",
       "      <td>24</td>\n",
       "      <td>3427</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>ie</td>\n",
       "      <td>250*250</td>\n",
       "      <td>0.0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>inf</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>6</td>\n",
       "      <td>23</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>160</td>\n",
       "      <td>600</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>25</td>\n",
       "      <td>1458</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>ie</td>\n",
       "      <td>160*600</td>\n",
       "      <td>0.0</td>\n",
       "      <td>25.0</td>\n",
       "      <td>inf</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>216</td>\n",
       "      <td>233</td>\n",
       "      <td>2.0</td>\n",
       "      <td>728</td>\n",
       "      <td>90</td>\n",
       "      <td>133</td>\n",
       "      <td>...</td>\n",
       "      <td>133</td>\n",
       "      <td>2259</td>\n",
       "      <td>0</td>\n",
       "      <td>windows</td>\n",
       "      <td>ie</td>\n",
       "      <td>728*90</td>\n",
       "      <td>0.0</td>\n",
       "      <td>133.0</td>\n",
       "      <td>inf</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 21 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Imps  click  weekday  hour  region  city  adexchange  slotwidth  \\\n",
       "0     1      0        5    22       2     2         2.0        200   \n",
       "1     1      0        1    20     238   239         1.0        300   \n",
       "2     1      0        3    13      40    41         2.0        250   \n",
       "3     1      0        6    23       1     1         1.0        160   \n",
       "4     1      0        5     6     216   233         2.0        728   \n",
       "\n",
       "   slotheight  slotprice  ...    payprice  advertiser  clickcost       OS  \\\n",
       "0         200          5  ...           5        3427          0  windows   \n",
       "1         250          0  ...          23        2821          0  windows   \n",
       "2         250          5  ...          24        3427          0  windows   \n",
       "3         600          0  ...          25        1458          0  windows   \n",
       "4          90        133  ...         133        2259          0  windows   \n",
       "\n",
       "  browser Slot__Size CTR%    CPM  eCPC  eCost  \n",
       "0      ie    200*200  0.0    5.0   inf    0.0  \n",
       "1  chrome    300*250  0.0   23.0   inf    0.0  \n",
       "2      ie    250*250  0.0   24.0   inf    0.0  \n",
       "3      ie    160*600  0.0   25.0   inf    0.0  \n",
       "4      ie     728*90  0.0  133.0   inf    0.0  \n",
       "\n",
       "[5 rows x 21 columns]"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_df = train.drop(['usertag','slotvisibility','slotformat','bidid','userid','useragent','IP','domain','url','urlid','slotid','creative','keypage'], axis=1)\n",
    "train_df.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_1458 = train_df[train_df['advertiser']==1458] ##\n",
    "df_2259 = train_df[train_df['advertiser']==2259] ###\n",
    "df_2997 = train_df[train_df['advertiser']==2997] ###\n",
    "df_3358 = train_df[train_df['advertiser']==3358]\n",
    "df_3386 = train_df[train_df['advertiser']==3386]\n",
    "df_3427 = train_df[train_df['advertiser']==3427]\n",
    "df_3476 = train_df[train_df['advertiser']==3476]\n",
    "df_2261 = train_df[train_df['advertiser']==2261]\n",
    "df_2821 = train_df[train_df['advertiser']==2821]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "week_df = train_df.groupby(['advertiser','weekday']).mean()\n",
    "hour_df = train_df.groupby(['advertiser','hour']).mean()\n",
    "OS_df = train_df.groupby(['advertiser','OS']).mean()\n",
    "browser_df = train_df.groupby(['advertiser','browser']).mean()\n",
    "region_df = train_df.groupby(['advertiser','region']).mean()\n",
    "city_df = train_df.groupby(['advertiser','city']).mean()\n",
    "slot_df = train_df.groupby(['advertiser','Slot__Size']).mean()\n",
    "adex_df = train_df.groupby(['advertiser','adexchange']).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [],
   "source": [
    "#x_features = ['weekday','hour','OS','browser','region','Slot__Size','adexchange']\n",
    "#y_features = ['CTR%','payprice','eCPC']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "x= 'browser'\n",
    "y= 'payprice'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>browser</th>\n",
       "      <th>payprice</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>chrome</td>\n",
       "      <td>71.541714</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>firefox</td>\n",
       "      <td>68.336451</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ie</td>\n",
       "      <td>67.841532</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>maxthon</td>\n",
       "      <td>71.364152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>opera</td>\n",
       "      <td>70.495614</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   browser   payprice\n",
       "0   chrome  71.541714\n",
       "1  firefox  68.336451\n",
       "2       ie  67.841532\n",
       "3  maxthon  71.364152\n",
       "4    opera  70.495614"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d1 = df_1458.groupby([x]).mean()\n",
    "d1 = d1[y]\n",
    "d_1 = pd.DataFrame(data = d1).reset_index()\n",
    "d_1.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>browser</th>\n",
       "      <th>payprice</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>chrome</td>\n",
       "      <td>95.927323</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>firefox</td>\n",
       "      <td>93.839416</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ie</td>\n",
       "      <td>91.380214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>maxthon</td>\n",
       "      <td>83.017182</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>opera</td>\n",
       "      <td>62.448598</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   browser   payprice\n",
       "0   chrome  95.927323\n",
       "1  firefox  93.839416\n",
       "2       ie  91.380214\n",
       "3  maxthon  83.017182\n",
       "4    opera  62.448598"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d2 = df_2259.groupby([x]).mean()\n",
    "d2 = d2[y]\n",
    "d_2 = pd.DataFrame(data = d2).reset_index()\n",
    "d_2.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>browser</th>\n",
       "      <th>payprice</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>chrome</td>\n",
       "      <td>64.784000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ie</td>\n",
       "      <td>66.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>maxthon</td>\n",
       "      <td>67.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>other</td>\n",
       "      <td>59.674762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>safari</td>\n",
       "      <td>62.860191</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   browser   payprice\n",
       "0   chrome  64.784000\n",
       "1       ie  66.000000\n",
       "2  maxthon  67.125000\n",
       "3    other  59.674762\n",
       "4   safari  62.860191"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d3 = df_2997.groupby([x]).mean()\n",
    "d3 = d3[y]\n",
    "d_3 = pd.DataFrame(data = d3).reset_index()\n",
    "d_3.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>browser</th>\n",
       "      <th>payprice</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>chrome</td>\n",
       "      <td>80.799593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>firefox</td>\n",
       "      <td>72.279466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ie</td>\n",
       "      <td>77.235134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>maxthon</td>\n",
       "      <td>72.796660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>opera</td>\n",
       "      <td>72.285143</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   browser   payprice\n",
       "0   chrome  80.799593\n",
       "1  firefox  72.279466\n",
       "2       ie  77.235134\n",
       "3  maxthon  72.796660\n",
       "4    opera  72.285143"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d_A = train_df.groupby([x]).mean()\n",
    "d_A = d_A[y]\n",
    "d_a = pd.DataFrame(data = d_A).reset_index()\n",
    "d_a.head(5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEJCAYAAACKWmBmAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x13bc98a90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, axes = plt.subplots(figsize=(6,4),sharex=False)\n",
    "\n",
    "d1.plot(label = \"1458\")\n",
    "d2.plot(label = \"2259\")\n",
    "d3.plot(label = \"2997\")\n",
    "d_A.plot(label = \"Avg\")\n",
    "\n",
    "#plt.title(\"graph\", fontsize=10)\n",
    "plt.legend()\n",
    "plt.ylabel(y, fontsize=8)\n",
    "plt.xlabel(x, fontsize=8)\n",
    "#plt.xticks(np.arange(0, 24, 1.0))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>browser</th>\n",
       "      <th>payprice_x</th>\n",
       "      <th>payprice_y</th>\n",
       "      <th>payprice</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>chrome</td>\n",
       "      <td>71.541714</td>\n",
       "      <td>95.927323</td>\n",
       "      <td>64.784000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>firefox</td>\n",
       "      <td>68.336451</td>\n",
       "      <td>93.839416</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ie</td>\n",
       "      <td>67.841532</td>\n",
       "      <td>91.380214</td>\n",
       "      <td>66.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>maxthon</td>\n",
       "      <td>71.364152</td>\n",
       "      <td>83.017182</td>\n",
       "      <td>67.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>opera</td>\n",
       "      <td>70.495614</td>\n",
       "      <td>62.448598</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>other</td>\n",
       "      <td>72.092585</td>\n",
       "      <td>88.544604</td>\n",
       "      <td>59.674762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>safari</td>\n",
       "      <td>67.216010</td>\n",
       "      <td>82.400225</td>\n",
       "      <td>62.860191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>sogou</td>\n",
       "      <td>146.200000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>37.400000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>theworld</td>\n",
       "      <td>67.820000</td>\n",
       "      <td>88.177945</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    browser  payprice_x  payprice_y   payprice\n",
       "0    chrome   71.541714   95.927323  64.784000\n",
       "1   firefox   68.336451   93.839416        NaN\n",
       "2        ie   67.841532   91.380214  66.000000\n",
       "3   maxthon   71.364152   83.017182  67.125000\n",
       "4     opera   70.495614   62.448598        NaN\n",
       "5     other   72.092585   88.544604  59.674762\n",
       "6    safari   67.216010   82.400225  62.860191\n",
       "7     sogou  146.200000         NaN  37.400000\n",
       "8  theworld   67.820000   88.177945        NaN"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d_all = pd.merge(d_1, d_2, how='outer', on=[x, x])\n",
    "d_all = pd.merge(d_all, d_3, how='outer', on=[x, x])\n",
    "#d_all = pd.merge(d_all, d_a, how='outer', on=[x, x])\n",
    "d_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>browser</th>\n",
       "      <th>1458</th>\n",
       "      <th>2259</th>\n",
       "      <th>2997</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>chrome</td>\n",
       "      <td>71.541714</td>\n",
       "      <td>95.927323</td>\n",
       "      <td>64.784000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>firefox</td>\n",
       "      <td>68.336451</td>\n",
       "      <td>93.839416</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ie</td>\n",
       "      <td>67.841532</td>\n",
       "      <td>91.380214</td>\n",
       "      <td>66.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>maxthon</td>\n",
       "      <td>71.364152</td>\n",
       "      <td>83.017182</td>\n",
       "      <td>67.125000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>opera</td>\n",
       "      <td>70.495614</td>\n",
       "      <td>62.448598</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>other</td>\n",
       "      <td>72.092585</td>\n",
       "      <td>88.544604</td>\n",
       "      <td>59.674762</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>safari</td>\n",
       "      <td>67.216010</td>\n",
       "      <td>82.400225</td>\n",
       "      <td>62.860191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>sogou</td>\n",
       "      <td>146.200000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>37.400000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>theworld</td>\n",
       "      <td>67.820000</td>\n",
       "      <td>88.177945</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    browser        1458       2259       2997\n",
       "0    chrome   71.541714  95.927323  64.784000\n",
       "1   firefox   68.336451  93.839416   0.000000\n",
       "2        ie   67.841532  91.380214  66.000000\n",
       "3   maxthon   71.364152  83.017182  67.125000\n",
       "4     opera   70.495614  62.448598   0.000000\n",
       "5     other   72.092585  88.544604  59.674762\n",
       "6    safari   67.216010  82.400225  62.860191\n",
       "7     sogou  146.200000   0.000000  37.400000\n",
       "8  theworld   67.820000  88.177945   0.000000"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d_all= d_all.rename(columns={'payprice_x': '1458'})\n",
    "d_all= d_all.rename(columns={'payprice_y': '2259'})\n",
    "d_all= d_all.rename(columns={'payprice': '2997'})\n",
    "\n",
    "d_all= d_all.rename(columns={'CTR%_x': '1458'})\n",
    "d_all= d_all.rename(columns={'CTR%_y': '2259'})\n",
    "d_all= d_all.rename(columns={'CTR%': '2997'})\n",
    "\n",
    "d_all = d_all.fillna(0)\n",
    "#d_all = d_all.fillna(0)\n",
    "#d_all = d_all.fillna(0)\n",
    "\n",
    "\n",
    "d_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>browser</th>\n",
       "      <th>1458</th>\n",
       "      <th>2259</th>\n",
       "      <th>2997</th>\n",
       "      <th>Avg</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>chrome</td>\n",
       "      <td>71.541714</td>\n",
       "      <td>95.927323</td>\n",
       "      <td>64.784000</td>\n",
       "      <td>80.799593</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>firefox</td>\n",
       "      <td>68.336451</td>\n",
       "      <td>93.839416</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>72.279466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ie</td>\n",
       "      <td>67.841532</td>\n",
       "      <td>91.380214</td>\n",
       "      <td>66.000000</td>\n",
       "      <td>77.235134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>maxthon</td>\n",
       "      <td>71.364152</td>\n",
       "      <td>83.017182</td>\n",
       "      <td>67.125000</td>\n",
       "      <td>72.796660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>opera</td>\n",
       "      <td>70.495614</td>\n",
       "      <td>62.448598</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>72.285143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>other</td>\n",
       "      <td>72.092585</td>\n",
       "      <td>88.544604</td>\n",
       "      <td>59.674762</td>\n",
       "      <td>81.888711</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>safari</td>\n",
       "      <td>67.216010</td>\n",
       "      <td>82.400225</td>\n",
       "      <td>62.860191</td>\n",
       "      <td>72.098584</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>sogou</td>\n",
       "      <td>146.200000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>37.400000</td>\n",
       "      <td>88.419355</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>theworld</td>\n",
       "      <td>67.820000</td>\n",
       "      <td>88.177945</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>74.402424</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    browser        1458       2259       2997        Avg\n",
       "0    chrome   71.541714  95.927323  64.784000  80.799593\n",
       "1   firefox   68.336451  93.839416   0.000000  72.279466\n",
       "2        ie   67.841532  91.380214  66.000000  77.235134\n",
       "3   maxthon   71.364152  83.017182  67.125000  72.796660\n",
       "4     opera   70.495614  62.448598   0.000000  72.285143\n",
       "5     other   72.092585  88.544604  59.674762  81.888711\n",
       "6    safari   67.216010  82.400225  62.860191  72.098584\n",
       "7     sogou  146.200000   0.000000  37.400000  88.419355\n",
       "8  theworld   67.820000  88.177945   0.000000  74.402424"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "d_all = pd.merge(d_all, d_a, how='outer', on=[x, x])\n",
    "d_all= d_all.rename(columns={'CTR%': 'Avg'})\n",
    "d_all= d_all.rename(columns={'payprice': 'Avg'})\n",
    "d_all"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfcAAAEqCAYAAAAI1RnWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ns4JxcjczMyuY/w9viXgieZSmBQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1491a5048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, axes = plt.subplots(figsize=(8,4),sharex=False)\n",
    "\n",
    "plt.scatter(d_all[x],d_all['1458'], label = \"1458\", color=\"red\",marker = 's')\n",
    "plt.scatter(d_all[x],d_all['2259'],label = \"2259\", color=\"blue\",marker = 'v')\n",
    "plt.scatter(d_all[x],d_all['2997'], label = \"2997\", color=\"orange\",marker = 'x')\n",
    "plt.scatter(d_all[x],d_all['Avg'], label = \"Avg\", color=\"green\",marker = 'o')\n",
    "\n",
    "plt.legend()\n",
    "plt.ylabel(y, fontsize=8)\n",
    "plt.xlabel(x, fontsize=8)\n",
    "plt.xticks(d_all[x] ,rotation=60)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtAAAAEZCAYAAAC+WBJ9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIiIiIhmogBYRERERyeD/A7Tal4sZqCIIAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x171a69668>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, axes = plt.subplots(figsize=(12,4),sharex=False)\n",
    "\n",
    "newx = np.arange(len(d_all['region']))\n",
    "\n",
    "plt.scatter(newx,d_all['1458'], label = \"1458\", color=\"red\",marker = 's')\n",
    "plt.scatter(newx,d_all['2259'],label = \"2259\", color=\"blue\",marker = 'v')\n",
    "plt.scatter(newx,d_all['2997'], label = \"2997\", color=\"orange\",marker = 'x')\n",
    "\n",
    "plt.legend()\n",
    "plt.ylabel(y, fontsize=8)\n",
    "plt.xlabel(x, fontsize=8)\n",
    "\n",
    "plt.xticks(newx, d_all[x],fontsize=12,rotation=60)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "df2 = df[['click',f,'eCost']]\n",
    "df3 = df2.groupby([f]).sum()\n",
    "df3['eCPC'] = df3['eCost']/df3['click']\n",
    "df3['eCPC'].plot.bar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d3 = df_2259.groupby(['weekday']).mean()\n",
    "d3 = d3['payprice']\n",
    "d4 = df_2261.groupby(['weekday']).mean()\n",
    "d4 = d4['payprice']\n",
    "d5 = df_2821.groupby(['weekday']).mean()\n",
    "d5 = d5['payprice']\n",
    "d6 = df_2997.groupby(['weekday']).mean()\n",
    "d6 = d6['payprice']\n",
    "d7 = df_3386.groupby(['weekday']).mean()\n",
    "d7 = d7['payprice'\n",
    "d8 = df_2259.groupby(['weekday']).mean()\n",
    "d8 = d3['payprice']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "d4.plot(label = \"d4\")\n",
    "d5.plot(label = \"d5\")\n",
    "d6.plot(label = \"d6\")\n",
    "d7.plot(label = \"d7\")\n",
    "d8.plot(label = \"d8\")\n",
    "d9.plot(label = \"d9\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}