wrangle_act.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FM5nYRDYint-"
   },
   "source": [
    "<h1> Project: Wrangle and Analyze Data</h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CpauUbJ89tfD"
   },
   "source": [
    "### Table of Contents\n",
    "\n",
    "- Gathering Data\n",
    "- Assessing Data\n",
    "- Assessment Summary\n",
    "- Cleaning Data\n",
    "- Storing Data\n",
    "- Analysis and Visualizations\n",
    "- References"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "NAAZ9Ir-r-61"
   },
   "source": [
    "### Project Details\n",
    "\n",
    "Tasks in this project are as follows:\n",
    "<ul>\n",
    "    <li>Data wrangling, which consists of:</li>\n",
    "        <ul>\n",
    "          <li>Gathering Data</li>\n",
    "           <li>Assessing data</li>\n",
    "           <li>Cleaning data</li>\n",
    "         </ul>\n",
    "    <li>Storing, analyzing, and visualizing your wrangled data</li>\n",
    "    <li>Reporting on\n",
    "    <ol>\n",
    "          <li>your data wrangling efforts and </li>\n",
    "           <li>your data analyses and visualizations</li>\n",
    "         </ol> \n",
    "</ul>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "M1S6LZC--mYc"
   },
   "source": [
    "## Import Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {
    "collapsed": true,
    "id": "THdDDAT_hr-j"
   },
   "outputs": [],
   "source": [
    "# Import the necessary libraries\n",
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import requests\n",
    "import os\n",
    "import tweepy\n",
    "from tweepy import OAuthHandler\n",
    "import json\n",
    "from timeit import default_timer as timer\n",
    "import time\n",
    "import re\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "% matplotlib inline\n",
    "import warnings\n",
    "from IPython.display import Image\n",
    "from functools import reduce\n",
    "import datetime"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qYCX0cYS-Yld"
   },
   "source": [
    "## Gathering Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "-x1m892yh53C",
    "outputId": "4a1de2a8-686b-4344-b8ed-5192d4ffd996"
   },
   "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>tweet_id</th>\n",
       "      <th>in_reply_to_status_id</th>\n",
       "      <th>in_reply_to_user_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>source</th>\n",
       "      <th>text</th>\n",
       "      <th>retweeted_status_id</th>\n",
       "      <th>retweeted_status_user_id</th>\n",
       "      <th>retweeted_status_timestamp</th>\n",
       "      <th>expanded_urls</th>\n",
       "      <th>rating_numerator</th>\n",
       "      <th>rating_denominator</th>\n",
       "      <th>name</th>\n",
       "      <th>doggo</th>\n",
       "      <th>floofer</th>\n",
       "      <th>pupper</th>\n",
       "      <th>puppo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-08-01 16:23:56 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Phineas. He's a mystical boy. Only eve...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/892420643...</td>\n",
       "      <td>13</td>\n",
       "      <td>10</td>\n",
       "      <td>Phineas</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-08-01 00:17:27 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Tilly. She's just checking pup on you....</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/892177421...</td>\n",
       "      <td>13</td>\n",
       "      <td>10</td>\n",
       "      <td>Tilly</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-07-31 00:18:03 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Archie. He is a rare Norwegian Pouncin...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/891815181...</td>\n",
       "      <td>12</td>\n",
       "      <td>10</td>\n",
       "      <td>Archie</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-07-30 15:58:51 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Darla. She commenced a snooze mid meal...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/891689557...</td>\n",
       "      <td>13</td>\n",
       "      <td>10</td>\n",
       "      <td>Darla</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-07-29 16:00:24 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Franklin. He would like you to stop ca...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/891327558...</td>\n",
       "      <td>12</td>\n",
       "      <td>10</td>\n",
       "      <td>Franklin</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             tweet_id  in_reply_to_status_id  ...  pupper puppo\n",
       "0  892420643555336193                    NaN  ...    None  None\n",
       "1  892177421306343426                    NaN  ...    None  None\n",
       "2  891815181378084864                    NaN  ...    None  None\n",
       "3  891689557279858688                    NaN  ...    None  None\n",
       "4  891327558926688256                    NaN  ...    None  None\n",
       "\n",
       "[5 rows x 17 columns]"
      ]
     },
     "execution_count": 244,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read-in the twitter-archive-enhanced-2 file\n",
    "df_twit_arch = pd.read_csv(\"twitter-archive-enhanced-2.csv\")\n",
    "# display the first five rows\n",
    "df_twit_arch.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {
    "collapsed": true,
    "id": "gDNlLwyq_5Wx"
   },
   "outputs": [],
   "source": [
    "# if not already in the folder, Download and save the image prediction file using the link provided\n",
    "if not os.path.exists('image_predictions.tsv'):\n",
    "  url = 'https://d17h27t6h515a5.cloudfront.net/topher/2017/August/599fd2ad_image-predictions/image-predictions.tsv'\n",
    "\n",
    "  resp = requests.get(url)\n",
    "  # write file\n",
    "  with open('image_predictions.tsv', 'wb') as file:\n",
    "      file.write(resp.content)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "BFRTVAMDLvcX",
    "outputId": "4b1f56d7-5b12-4fc5-af88-ca69a351b035"
   },
   "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>tweet_id</th>\n",
       "      <th>jpg_url</th>\n",
       "      <th>img_num</th>\n",
       "      <th>p1</th>\n",
       "      <th>p1_conf</th>\n",
       "      <th>p1_dog</th>\n",
       "      <th>p2</th>\n",
       "      <th>p2_conf</th>\n",
       "      <th>p2_dog</th>\n",
       "      <th>p3</th>\n",
       "      <th>p3_conf</th>\n",
       "      <th>p3_dog</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>666020888022790149</td>\n",
       "      <td>https://pbs.twimg.com/media/CT4udn0WwAA0aMy.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>Welsh_springer_spaniel</td>\n",
       "      <td>0.465074</td>\n",
       "      <td>True</td>\n",
       "      <td>collie</td>\n",
       "      <td>0.156665</td>\n",
       "      <td>True</td>\n",
       "      <td>Shetland_sheepdog</td>\n",
       "      <td>0.061428</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>666029285002620928</td>\n",
       "      <td>https://pbs.twimg.com/media/CT42GRgUYAA5iDo.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>redbone</td>\n",
       "      <td>0.506826</td>\n",
       "      <td>True</td>\n",
       "      <td>miniature_pinscher</td>\n",
       "      <td>0.074192</td>\n",
       "      <td>True</td>\n",
       "      <td>Rhodesian_ridgeback</td>\n",
       "      <td>0.072010</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>666033412701032449</td>\n",
       "      <td>https://pbs.twimg.com/media/CT4521TWwAEvMyu.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>German_shepherd</td>\n",
       "      <td>0.596461</td>\n",
       "      <td>True</td>\n",
       "      <td>malinois</td>\n",
       "      <td>0.138584</td>\n",
       "      <td>True</td>\n",
       "      <td>bloodhound</td>\n",
       "      <td>0.116197</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>666044226329800704</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5Dr8HUEAA-lEu.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>Rhodesian_ridgeback</td>\n",
       "      <td>0.408143</td>\n",
       "      <td>True</td>\n",
       "      <td>redbone</td>\n",
       "      <td>0.360687</td>\n",
       "      <td>True</td>\n",
       "      <td>miniature_pinscher</td>\n",
       "      <td>0.222752</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>666049248165822465</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5IQmsXIAAKY4A.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>miniature_pinscher</td>\n",
       "      <td>0.560311</td>\n",
       "      <td>True</td>\n",
       "      <td>Rottweiler</td>\n",
       "      <td>0.243682</td>\n",
       "      <td>True</td>\n",
       "      <td>Doberman</td>\n",
       "      <td>0.154629</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             tweet_id  ... p3_dog\n",
       "0  666020888022790149  ...   True\n",
       "1  666029285002620928  ...   True\n",
       "2  666033412701032449  ...   True\n",
       "3  666044226329800704  ...   True\n",
       "4  666049248165822465  ...   True\n",
       "\n",
       "[5 rows x 12 columns]"
      ]
     },
     "execution_count": 246,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read-in the image_predictions file\n",
    "df_img_pred = pd.read_csv('image_predictions.tsv', sep='\\t')\n",
    "# display the first five rows\n",
    "df_img_pred.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "slOk4WaTbZqt",
    "outputId": "b57ed4b2-5d99-4d64-cf29-5987d7008af2"
   },
   "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>created_at</th>\n",
       "      <th>id</th>\n",
       "      <th>id_str</th>\n",
       "      <th>full_text</th>\n",
       "      <th>truncated</th>\n",
       "      <th>display_text_range</th>\n",
       "      <th>entities</th>\n",
       "      <th>extended_entities</th>\n",
       "      <th>source</th>\n",
       "      <th>in_reply_to_status_id</th>\n",
       "      <th>in_reply_to_status_id_str</th>\n",
       "      <th>in_reply_to_user_id</th>\n",
       "      <th>in_reply_to_user_id_str</th>\n",
       "      <th>in_reply_to_screen_name</th>\n",
       "      <th>user</th>\n",
       "      <th>geo</th>\n",
       "      <th>coordinates</th>\n",
       "      <th>place</th>\n",
       "      <th>contributors</th>\n",
       "      <th>is_quote_status</th>\n",
       "      <th>retweet_count</th>\n",
       "      <th>favorite_count</th>\n",
       "      <th>favorited</th>\n",
       "      <th>retweeted</th>\n",
       "      <th>possibly_sensitive</th>\n",
       "      <th>possibly_sensitive_appealable</th>\n",
       "      <th>lang</th>\n",
       "      <th>retweeted_status</th>\n",
       "      <th>quoted_status_id</th>\n",
       "      <th>quoted_status_id_str</th>\n",
       "      <th>quoted_status_permalink</th>\n",
       "      <th>quoted_status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2017-08-01 16:23:56+00:00</td>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>892420643555336192</td>\n",
       "      <td>This is Phineas. He's a mystical boy. Only eve...</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 85]</td>\n",
       "      <td>{'hashtags': [], 'symbols': [], 'user_mentions...</td>\n",
       "      <td>{'media': [{'id': 892420639486877696, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>7315</td>\n",
       "      <td>34860</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2017-08-01 00:17:27+00:00</td>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>892177421306343424</td>\n",
       "      <td>This is Tilly. She's just checking pup on you....</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 138]</td>\n",
       "      <td>{'hashtags': [], 'symbols': [], 'user_mentions...</td>\n",
       "      <td>{'media': [{'id': 892177413194625024, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>5465</td>\n",
       "      <td>30195</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2017-07-31 00:18:03+00:00</td>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>This is Archie. He is a rare Norwegian Pouncin...</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 121]</td>\n",
       "      <td>{'hashtags': [], 'symbols': [], 'user_mentions...</td>\n",
       "      <td>{'media': [{'id': 891815175371796480, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>3605</td>\n",
       "      <td>22733</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017-07-30 15:58:51+00:00</td>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>This is Darla. She commenced a snooze mid meal...</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 79]</td>\n",
       "      <td>{'hashtags': [], 'symbols': [], 'user_mentions...</td>\n",
       "      <td>{'media': [{'id': 891689552724799489, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>7505</td>\n",
       "      <td>38132</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2017-07-29 16:00:24+00:00</td>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>This is Franklin. He would like you to stop ca...</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 138]</td>\n",
       "      <td>{'hashtags': [{'text': 'BarkWeek', 'indices': ...</td>\n",
       "      <td>{'media': [{'id': 891327551943041024, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>8057</td>\n",
       "      <td>36394</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 created_at  ...  quoted_status\n",
       "0 2017-08-01 16:23:56+00:00  ...            NaN\n",
       "1 2017-08-01 00:17:27+00:00  ...            NaN\n",
       "2 2017-07-31 00:18:03+00:00  ...            NaN\n",
       "3 2017-07-30 15:58:51+00:00  ...            NaN\n",
       "4 2017-07-29 16:00:24+00:00  ...            NaN\n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 248,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read-in the tweet_json file\n",
    "with open('tweet_json.txt') as file:\n",
    "    df_tweet_json = pd.read_json(file, lines= True, encoding = 'utf-8')\n",
    "\n",
    "# display the first five rows\n",
    "df_tweet_json.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "oke6ugRxw7M3"
   },
   "source": [
    "## Assessing Data\n",
    "After gathering each of the above pieces of data, assess them visually and programmatically for quality and tidiness issues. Detect and document at least eight (8) quality issues and two (2) tidiness issue.<br>\n",
    "One of the key points for the analysis requires that only tweets with original ratings with images, no retweets, be retained.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "u4tm-69izgyW"
   },
   "source": [
    "#### assessing twitter-archive-enhanced dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "eo9cI0rXKNL9",
    "outputId": "68f1b912-1188-46bf-8a76-ac2a9b3367f3"
   },
   "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>tweet_id</th>\n",
       "      <th>in_reply_to_status_id</th>\n",
       "      <th>in_reply_to_user_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>source</th>\n",
       "      <th>text</th>\n",
       "      <th>retweeted_status_id</th>\n",
       "      <th>retweeted_status_user_id</th>\n",
       "      <th>retweeted_status_timestamp</th>\n",
       "      <th>expanded_urls</th>\n",
       "      <th>rating_numerator</th>\n",
       "      <th>rating_denominator</th>\n",
       "      <th>name</th>\n",
       "      <th>doggo</th>\n",
       "      <th>floofer</th>\n",
       "      <th>pupper</th>\n",
       "      <th>puppo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-08-01 16:23:56 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Phineas. He's a mystical boy. Only eve...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/892420643...</td>\n",
       "      <td>13</td>\n",
       "      <td>10</td>\n",
       "      <td>Phineas</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-08-01 00:17:27 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Tilly. She's just checking pup on you....</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/892177421...</td>\n",
       "      <td>13</td>\n",
       "      <td>10</td>\n",
       "      <td>Tilly</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-07-31 00:18:03 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Archie. He is a rare Norwegian Pouncin...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/891815181...</td>\n",
       "      <td>12</td>\n",
       "      <td>10</td>\n",
       "      <td>Archie</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-07-30 15:58:51 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Darla. She commenced a snooze mid meal...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/891689557...</td>\n",
       "      <td>13</td>\n",
       "      <td>10</td>\n",
       "      <td>Darla</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-07-29 16:00:24 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is Franklin. He would like you to stop ca...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/891327558...</td>\n",
       "      <td>12</td>\n",
       "      <td>10</td>\n",
       "      <td>Franklin</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2351</th>\n",
       "      <td>666049248165822465</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-11-16 00:24:50 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>Here we have a 1949 1st generation vulpix. Enj...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/666049248...</td>\n",
       "      <td>5</td>\n",
       "      <td>10</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2352</th>\n",
       "      <td>666044226329800704</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-11-16 00:04:52 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is a purebred Piers Morgan. Loves to Netf...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/666044226...</td>\n",
       "      <td>6</td>\n",
       "      <td>10</td>\n",
       "      <td>a</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2353</th>\n",
       "      <td>666033412701032449</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-11-15 23:21:54 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>Here is a very happy pup. Big fan of well-main...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/666033412...</td>\n",
       "      <td>9</td>\n",
       "      <td>10</td>\n",
       "      <td>a</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2354</th>\n",
       "      <td>666029285002620928</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-11-15 23:05:30 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>This is a western brown Mitsubishi terrier. Up...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/666029285...</td>\n",
       "      <td>7</td>\n",
       "      <td>10</td>\n",
       "      <td>a</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2355</th>\n",
       "      <td>666020888022790149</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-11-15 22:32:08 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>Here we have a Japanese Irish Setter. Lost eye...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://twitter.com/dog_rates/status/666020888...</td>\n",
       "      <td>8</td>\n",
       "      <td>10</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2356 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                tweet_id  in_reply_to_status_id  ...  pupper puppo\n",
       "0     892420643555336193                    NaN  ...    None  None\n",
       "1     892177421306343426                    NaN  ...    None  None\n",
       "2     891815181378084864                    NaN  ...    None  None\n",
       "3     891689557279858688                    NaN  ...    None  None\n",
       "4     891327558926688256                    NaN  ...    None  None\n",
       "...                  ...                    ...  ...     ...   ...\n",
       "2351  666049248165822465                    NaN  ...    None  None\n",
       "2352  666044226329800704                    NaN  ...    None  None\n",
       "2353  666033412701032449                    NaN  ...    None  None\n",
       "2354  666029285002620928                    NaN  ...    None  None\n",
       "2355  666020888022790149                    NaN  ...    None  None\n",
       "\n",
       "[2356 rows x 17 columns]"
      ]
     },
     "execution_count": 249,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# visually asess the data\n",
    "df_twit_arch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "p6z1hpcHyMWO",
    "outputId": "4a63729e-d12a-4118-f10d-9c0ba51a9d57"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "None        745\n",
       "a            55\n",
       "Charlie      12\n",
       "Oliver       11\n",
       "Lucy         11\n",
       "           ... \n",
       "Schnozz       1\n",
       "Wesley        1\n",
       "Fletcher      1\n",
       "Bobby         1\n",
       "Andy          1\n",
       "Name: name, Length: 957, dtype: int64"
      ]
     },
     "execution_count": 250,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_twit_arch['name'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "YAqhKL_wrVJ8"
   },
   "source": [
    "Note missing values in the name column represented as None instead of NaN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "0ZPmD5eNXG_W",
    "outputId": "e553ec0f-6498-4d98-9845-1cb48bcd189c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tweet_id                        int64\n",
       "in_reply_to_status_id         float64\n",
       "in_reply_to_user_id           float64\n",
       "timestamp                      object\n",
       "source                         object\n",
       "text                           object\n",
       "retweeted_status_id           float64\n",
       "retweeted_status_user_id      float64\n",
       "retweeted_status_timestamp     object\n",
       "expanded_urls                  object\n",
       "rating_numerator                int64\n",
       "rating_denominator              int64\n",
       "name                           object\n",
       "doggo                          object\n",
       "floofer                        object\n",
       "pupper                         object\n",
       "puppo                          object\n",
       "dtype: object"
      ]
     },
     "execution_count": 251,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check columns data types\n",
    "df_twit_arch.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "QdThirmbXdfw"
   },
   "source": [
    "Note the timestamp to be of the object datatype which is wrong"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "Kw2D8qXKIMiE",
    "outputId": "d75d33d8-e358-4072-be1f-6788ab3f5c71"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 2356 entries, 0 to 2355\n",
      "Data columns (total 17 columns):\n",
      " #   Column                      Non-Null Count  Dtype  \n",
      "---  ------                      --------------  -----  \n",
      " 0   tweet_id                    2356 non-null   int64  \n",
      " 1   in_reply_to_status_id       78 non-null     float64\n",
      " 2   in_reply_to_user_id         78 non-null     float64\n",
      " 3   timestamp                   2356 non-null   object \n",
      " 4   source                      2356 non-null   object \n",
      " 5   text                        2356 non-null   object \n",
      " 6   retweeted_status_id         181 non-null    float64\n",
      " 7   retweeted_status_user_id    181 non-null    float64\n",
      " 8   retweeted_status_timestamp  181 non-null    object \n",
      " 9   expanded_urls               2297 non-null   object \n",
      " 10  rating_numerator            2356 non-null   int64  \n",
      " 11  rating_denominator          2356 non-null   int64  \n",
      " 12  name                        2356 non-null   object \n",
      " 13  doggo                       2356 non-null   object \n",
      " 14  floofer                     2356 non-null   object \n",
      " 15  pupper                      2356 non-null   object \n",
      " 16  puppo                       2356 non-null   object \n",
      "dtypes: float64(4), int64(3), object(10)\n",
      "memory usage: 313.0+ KB\n"
     ]
    }
   ],
   "source": [
    "# show conscise summary of the twitter-archive-enhanced-2 data\n",
    "df_twit_arch.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "E3rdtenYOyxC",
    "outputId": "0a571a89-7417-459e-9c8d-346cb67283b9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tweet_id                       0.000000\n",
       "in_reply_to_status_id         96.689304\n",
       "in_reply_to_user_id           96.689304\n",
       "timestamp                      0.000000\n",
       "source                         0.000000\n",
       "text                           0.000000\n",
       "retweeted_status_id           92.317487\n",
       "retweeted_status_user_id      92.317487\n",
       "retweeted_status_timestamp    92.317487\n",
       "expanded_urls                  2.504244\n",
       "rating_numerator               0.000000\n",
       "rating_denominator             0.000000\n",
       "name                           0.000000\n",
       "doggo                          0.000000\n",
       "floofer                        0.000000\n",
       "pupper                         0.000000\n",
       "puppo                          0.000000\n",
       "dtype: float64"
      ]
     },
     "execution_count": 253,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check percentage of missing values\n",
    "(df_twit_arch.isnull().sum()/df_twit_arch.isnull().count())*100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XifNb3KRruDx"
   },
   "source": [
    "Note columns that are almost of missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "QFdGzE5yVuXI",
    "outputId": "e4ebcf28-9f8f-4620-b240-d8ca19952f9b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tweet_id                      2356\n",
       "in_reply_to_status_id           77\n",
       "in_reply_to_user_id             31\n",
       "timestamp                     2356\n",
       "source                           4\n",
       "text                          2356\n",
       "retweeted_status_id            181\n",
       "retweeted_status_user_id        25\n",
       "retweeted_status_timestamp     181\n",
       "expanded_urls                 2218\n",
       "rating_numerator                40\n",
       "rating_denominator              18\n",
       "name                           957\n",
       "doggo                            2\n",
       "floofer                          2\n",
       "pupper                           2\n",
       "puppo                            2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 254,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# count the number of unique values in each column of the df_twit_arch\n",
    "df_twit_arch.nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "NgR1MR7QwYY7",
    "outputId": "1f03b7e4-921d-4797-b69d-3e7b5f740558"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10     2333\n",
       "11        3\n",
       "50        3\n",
       "80        2\n",
       "20        2\n",
       "2         1\n",
       "16        1\n",
       "40        1\n",
       "70        1\n",
       "15        1\n",
       "90        1\n",
       "110       1\n",
       "120       1\n",
       "130       1\n",
       "150       1\n",
       "170       1\n",
       "7         1\n",
       "0         1\n",
       "Name: rating_denominator, dtype: int64"
      ]
     },
     "execution_count": 255,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check to see if truly ratings almost always have a denominator of 10\n",
    "df_twit_arch['rating_denominator'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "nFhXIubGxrVU",
    "outputId": "cac6a417-df92-4ea3-fe63-700a25fde427"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12      558\n",
       "11      464\n",
       "10      461\n",
       "13      351\n",
       "9       158\n",
       "8       102\n",
       "7        55\n",
       "14       54\n",
       "5        37\n",
       "6        32\n",
       "3        19\n",
       "4        17\n",
       "1         9\n",
       "2         9\n",
       "420       2\n",
       "0         2\n",
       "15        2\n",
       "75        2\n",
       "80        1\n",
       "20        1\n",
       "24        1\n",
       "26        1\n",
       "44        1\n",
       "50        1\n",
       "60        1\n",
       "165       1\n",
       "84        1\n",
       "88        1\n",
       "144       1\n",
       "182       1\n",
       "143       1\n",
       "666       1\n",
       "960       1\n",
       "1776      1\n",
       "17        1\n",
       "27        1\n",
       "45        1\n",
       "99        1\n",
       "121       1\n",
       "204       1\n",
       "Name: rating_numerator, dtype: int64"
      ]
     },
     "execution_count": 256,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check to see if truly ratings almost always have a numerator greater than 10\n",
    "df_twit_arch['rating_numerator'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "ds6KEQSHymLI",
    "outputId": "659e1624-d946-4f43-ba4e-d30b75932dd9"
   },
   "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>tweet_id</th>\n",
       "      <th>in_reply_to_status_id</th>\n",
       "      <th>in_reply_to_user_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>source</th>\n",
       "      <th>text</th>\n",
       "      <th>retweeted_status_id</th>\n",
       "      <th>retweeted_status_user_id</th>\n",
       "      <th>retweeted_status_timestamp</th>\n",
       "      <th>expanded_urls</th>\n",
       "      <th>rating_numerator</th>\n",
       "      <th>rating_denominator</th>\n",
       "      <th>name</th>\n",
       "      <th>doggo</th>\n",
       "      <th>floofer</th>\n",
       "      <th>pupper</th>\n",
       "      <th>puppo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2.356000e+03</td>\n",
       "      <td>7.800000e+01</td>\n",
       "      <td>7.800000e+01</td>\n",
       "      <td>2356</td>\n",
       "      <td>2356</td>\n",
       "      <td>2356</td>\n",
       "      <td>1.810000e+02</td>\n",
       "      <td>1.810000e+02</td>\n",
       "      <td>181</td>\n",
       "      <td>2297</td>\n",
       "      <td>2356.000000</td>\n",
       "      <td>2356.000000</td>\n",
       "      <td>2356</td>\n",
       "      <td>2356</td>\n",
       "      <td>2356</td>\n",
       "      <td>2356</td>\n",
       "      <td>2356</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2356</td>\n",
       "      <td>4</td>\n",
       "      <td>2356</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>181</td>\n",
       "      <td>2218</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>957</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017-02-07 17:00:28 +0000</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>Meet Al Cabone. He's a gangsta puppa. Rather h...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-12-16 01:27:03 +0000</td>\n",
       "      <td>https://twitter.com/dog_rates/status/773308824...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>2221</td>\n",
       "      <td>1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>745</td>\n",
       "      <td>2259</td>\n",
       "      <td>2346</td>\n",
       "      <td>2099</td>\n",
       "      <td>2326</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>7.427716e+17</td>\n",
       "      <td>7.455079e+17</td>\n",
       "      <td>2.014171e+16</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.720400e+17</td>\n",
       "      <td>1.241698e+16</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>13.126486</td>\n",
       "      <td>10.455433</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6.856705e+16</td>\n",
       "      <td>7.582492e+16</td>\n",
       "      <td>1.252797e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.236928e+16</td>\n",
       "      <td>9.599254e+16</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>45.876648</td>\n",
       "      <td>6.745237</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>6.660209e+17</td>\n",
       "      <td>6.658147e+17</td>\n",
       "      <td>1.185634e+07</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.661041e+17</td>\n",
       "      <td>7.832140e+05</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>6.783989e+17</td>\n",
       "      <td>6.757419e+17</td>\n",
       "      <td>3.086374e+08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.186315e+17</td>\n",
       "      <td>4.196984e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>7.196279e+17</td>\n",
       "      <td>7.038708e+17</td>\n",
       "      <td>4.196984e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7.804657e+17</td>\n",
       "      <td>4.196984e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.993373e+17</td>\n",
       "      <td>8.257804e+17</td>\n",
       "      <td>4.196984e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.203146e+17</td>\n",
       "      <td>4.196984e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>12.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>8.924206e+17</td>\n",
       "      <td>8.862664e+17</td>\n",
       "      <td>8.405479e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.874740e+17</td>\n",
       "      <td>7.874618e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1776.000000</td>\n",
       "      <td>170.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            tweet_id  in_reply_to_status_id  ...  pupper puppo\n",
       "count   2.356000e+03           7.800000e+01  ...    2356  2356\n",
       "unique           NaN                    NaN  ...       2     2\n",
       "top              NaN                    NaN  ...    None  None\n",
       "freq             NaN                    NaN  ...    2099  2326\n",
       "mean    7.427716e+17           7.455079e+17  ...     NaN   NaN\n",
       "std     6.856705e+16           7.582492e+16  ...     NaN   NaN\n",
       "min     6.660209e+17           6.658147e+17  ...     NaN   NaN\n",
       "25%     6.783989e+17           6.757419e+17  ...     NaN   NaN\n",
       "50%     7.196279e+17           7.038708e+17  ...     NaN   NaN\n",
       "75%     7.993373e+17           8.257804e+17  ...     NaN   NaN\n",
       "max     8.924206e+17           8.862664e+17  ...     NaN   NaN\n",
       "\n",
       "[11 rows x 17 columns]"
      ]
     },
     "execution_count": 257,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show basic statistics of the df_twit_arch data \n",
    "df_twit_arch.describe(include=\"all\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "_wMTcPKTwZ3Y",
    "outputId": "26859228-1f90-49ed-d19f-d1037975101c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[\"Here's a super supportive puppo participating in the Toronto  #WomensMarch today. 13/10 https://t.co/nTz3FtorBc\",\n",
       "        13],\n",
       "       ['This is Dex. He can see into your past and future. Mesmerizing af 11/10 https://t.co/0dYI0Cpdge',\n",
       "        11],\n",
       "       ['This is Darla. She commenced a snooze mid meal. 13/10 happens to the best of us https://t.co/tD36da7qLQ',\n",
       "        13],\n",
       "       [\"This is Dotsy. She's stuck as hell. 10/10 https://t.co/A0h4lnhU4s\",\n",
       "        10],\n",
       "       [\"Meet Lilah. She agreed on one quick pic. Now she'd like to go mentally prepare for the onslaught of fireworks. 11/10 https://t.co/enCpXzZHkD\",\n",
       "        11],\n",
       "       [\"Say hello to Charlie. He's scholarly af. Quite intimidating with all his pupper knowledge 10/10 even built that fire https://t.co/9KThv6z8u5\",\n",
       "        10],\n",
       "       ['After much debate this dog is being upgraded to 10/10. I repeat 10/10',\n",
       "        10],\n",
       "       [\"This is Arlo. He's officially the king of snowy tongue slips. 13/10 would comfort during inevitable brain freeze https://t.co/oXVu9pNZZv\",\n",
       "        13],\n",
       "       ['Say hello to Jazz. She should be on the cover of Vogue. 12/10 gorgeous pupper https://t.co/mVCMemhXAP',\n",
       "        12],\n",
       "       [\"Meet Abby. She's incredibly distracting. Just wants to help steer. Hazardous af. Still 12/10 would pet while driving https://t.co/gLbLiZtwsp\",\n",
       "        12]], dtype=object)"
      ]
     },
     "execution_count": 258,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_twit_arch[['text','rating_numerator']].sample(10).values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FVaHdZSa63eY"
   },
   "source": [
    "#### assessing image_predictions dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "IZ7Ws5nf63en",
    "outputId": "c8558e8f-215c-4287-93b5-29bc5782c874"
   },
   "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>tweet_id</th>\n",
       "      <th>jpg_url</th>\n",
       "      <th>img_num</th>\n",
       "      <th>p1</th>\n",
       "      <th>p1_conf</th>\n",
       "      <th>p1_dog</th>\n",
       "      <th>p2</th>\n",
       "      <th>p2_conf</th>\n",
       "      <th>p2_dog</th>\n",
       "      <th>p3</th>\n",
       "      <th>p3_conf</th>\n",
       "      <th>p3_dog</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>666020888022790149</td>\n",
       "      <td>https://pbs.twimg.com/media/CT4udn0WwAA0aMy.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>Welsh_springer_spaniel</td>\n",
       "      <td>0.465074</td>\n",
       "      <td>True</td>\n",
       "      <td>collie</td>\n",
       "      <td>0.156665</td>\n",
       "      <td>True</td>\n",
       "      <td>Shetland_sheepdog</td>\n",
       "      <td>0.061428</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>666029285002620928</td>\n",
       "      <td>https://pbs.twimg.com/media/CT42GRgUYAA5iDo.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>redbone</td>\n",
       "      <td>0.506826</td>\n",
       "      <td>True</td>\n",
       "      <td>miniature_pinscher</td>\n",
       "      <td>0.074192</td>\n",
       "      <td>True</td>\n",
       "      <td>Rhodesian_ridgeback</td>\n",
       "      <td>0.072010</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>666033412701032449</td>\n",
       "      <td>https://pbs.twimg.com/media/CT4521TWwAEvMyu.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>German_shepherd</td>\n",
       "      <td>0.596461</td>\n",
       "      <td>True</td>\n",
       "      <td>malinois</td>\n",
       "      <td>0.138584</td>\n",
       "      <td>True</td>\n",
       "      <td>bloodhound</td>\n",
       "      <td>0.116197</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>666044226329800704</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5Dr8HUEAA-lEu.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>Rhodesian_ridgeback</td>\n",
       "      <td>0.408143</td>\n",
       "      <td>True</td>\n",
       "      <td>redbone</td>\n",
       "      <td>0.360687</td>\n",
       "      <td>True</td>\n",
       "      <td>miniature_pinscher</td>\n",
       "      <td>0.222752</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>666049248165822465</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5IQmsXIAAKY4A.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>miniature_pinscher</td>\n",
       "      <td>0.560311</td>\n",
       "      <td>True</td>\n",
       "      <td>Rottweiler</td>\n",
       "      <td>0.243682</td>\n",
       "      <td>True</td>\n",
       "      <td>Doberman</td>\n",
       "      <td>0.154629</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2070</th>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>https://pbs.twimg.com/media/DF6hr6BUMAAzZgT.jpg</td>\n",
       "      <td>2</td>\n",
       "      <td>basset</td>\n",
       "      <td>0.555712</td>\n",
       "      <td>True</td>\n",
       "      <td>English_springer</td>\n",
       "      <td>0.225770</td>\n",
       "      <td>True</td>\n",
       "      <td>German_short-haired_pointer</td>\n",
       "      <td>0.175219</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2071</th>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>https://pbs.twimg.com/media/DF_q7IAWsAEuuN8.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>paper_towel</td>\n",
       "      <td>0.170278</td>\n",
       "      <td>False</td>\n",
       "      <td>Labrador_retriever</td>\n",
       "      <td>0.168086</td>\n",
       "      <td>True</td>\n",
       "      <td>spatula</td>\n",
       "      <td>0.040836</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2072</th>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>https://pbs.twimg.com/media/DGBdLU1WsAANxJ9.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>Chihuahua</td>\n",
       "      <td>0.716012</td>\n",
       "      <td>True</td>\n",
       "      <td>malamute</td>\n",
       "      <td>0.078253</td>\n",
       "      <td>True</td>\n",
       "      <td>kelpie</td>\n",
       "      <td>0.031379</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2073</th>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>https://pbs.twimg.com/media/DGGmoV4XsAAUL6n.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>Chihuahua</td>\n",
       "      <td>0.323581</td>\n",
       "      <td>True</td>\n",
       "      <td>Pekinese</td>\n",
       "      <td>0.090647</td>\n",
       "      <td>True</td>\n",
       "      <td>papillon</td>\n",
       "      <td>0.068957</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2074</th>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>https://pbs.twimg.com/media/DGKD1-bXoAAIAUK.jpg</td>\n",
       "      <td>1</td>\n",
       "      <td>orange</td>\n",
       "      <td>0.097049</td>\n",
       "      <td>False</td>\n",
       "      <td>bagel</td>\n",
       "      <td>0.085851</td>\n",
       "      <td>False</td>\n",
       "      <td>banana</td>\n",
       "      <td>0.076110</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2075 rows × 12 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                tweet_id  ... p3_dog\n",
       "0     666020888022790149  ...   True\n",
       "1     666029285002620928  ...   True\n",
       "2     666033412701032449  ...   True\n",
       "3     666044226329800704  ...   True\n",
       "4     666049248165822465  ...   True\n",
       "...                  ...  ...    ...\n",
       "2070  891327558926688256  ...   True\n",
       "2071  891689557279858688  ...  False\n",
       "2072  891815181378084864  ...   True\n",
       "2073  892177421306343426  ...   True\n",
       "2074  892420643555336193  ...  False\n",
       "\n",
       "[2075 rows x 12 columns]"
      ]
     },
     "execution_count": 259,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# visually asess the data\n",
    "df_img_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "bjbBh4y063er",
    "outputId": "178621d0-b230-41f6-cf35-2454589baa74"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True     1532\n",
       "False     543\n",
       "Name: p1_dog, dtype: int64"
      ]
     },
     "execution_count": 260,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check the number of correct predictions\n",
    "df_img_pred['p1_dog'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 261,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "bsFXboK8Ibp5",
    "outputId": "db99a1e8-8978-4b6e-e806-70fd183002fc"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True     1553\n",
       "False     522\n",
       "Name: p2_dog, dtype: int64"
      ]
     },
     "execution_count": 261,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check the number of correct predictions\n",
    "df_img_pred['p2_dog'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 262,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "y5huokmwIgA2",
    "outputId": "47ea9ad9-9cd9-4be9-81dc-9568e0153756"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True     1499\n",
       "False     576\n",
       "Name: p3_dog, dtype: int64"
      ]
     },
     "execution_count": 262,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check the number of correct predictions\n",
    "df_img_pred['p3_dog'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 263,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "-sTyv_EGCewM",
    "outputId": "6a327913-6a99-4171-aef7-be99fb19d60e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Labrador_retriever    104\n",
       "golden_retriever       92\n",
       "Cardigan               73\n",
       "Chihuahua              44\n",
       "Pomeranian             42\n",
       "                     ... \n",
       "bagel                   1\n",
       "waffle_iron             1\n",
       "brown_bear              1\n",
       "streetcar               1\n",
       "racket                  1\n",
       "Name: p2, Length: 405, dtype: int64"
      ]
     },
     "execution_count": 263,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_img_pred['p2'].value_counts().sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 264,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "O_ftvPQ663eu",
    "outputId": "a765a4a4-00fc-42fd-e706-e6c211b4ea60"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tweet_id      int64\n",
       "jpg_url      object\n",
       "img_num       int64\n",
       "p1           object\n",
       "p1_conf     float64\n",
       "p1_dog         bool\n",
       "p2           object\n",
       "p2_conf     float64\n",
       "p2_dog         bool\n",
       "p3           object\n",
       "p3_conf     float64\n",
       "p3_dog         bool\n",
       "dtype: object"
      ]
     },
     "execution_count": 264,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check columns data types\n",
    "df_img_pred.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 265,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "1fSelNYX63ew",
    "outputId": "43e27a4e-0f60-4463-9a26-21bb3d1ee5b5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 2075 entries, 0 to 2074\n",
      "Data columns (total 12 columns):\n",
      " #   Column    Non-Null Count  Dtype  \n",
      "---  ------    --------------  -----  \n",
      " 0   tweet_id  2075 non-null   int64  \n",
      " 1   jpg_url   2075 non-null   object \n",
      " 2   img_num   2075 non-null   int64  \n",
      " 3   p1        2075 non-null   object \n",
      " 4   p1_conf   2075 non-null   float64\n",
      " 5   p1_dog    2075 non-null   bool   \n",
      " 6   p2        2075 non-null   object \n",
      " 7   p2_conf   2075 non-null   float64\n",
      " 8   p2_dog    2075 non-null   bool   \n",
      " 9   p3        2075 non-null   object \n",
      " 10  p3_conf   2075 non-null   float64\n",
      " 11  p3_dog    2075 non-null   bool   \n",
      "dtypes: bool(3), float64(3), int64(2), object(4)\n",
      "memory usage: 152.1+ KB\n"
     ]
    }
   ],
   "source": [
    "# show conscise summary of the image_predictions data\n",
    "df_img_pred.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 266,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "B8N7XNpA63ey",
    "outputId": "6b90058a-b26e-4415-f2a6-c130568d4343"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tweet_id    0.0\n",
       "jpg_url     0.0\n",
       "img_num     0.0\n",
       "p1          0.0\n",
       "p1_conf     0.0\n",
       "p1_dog      0.0\n",
       "p2          0.0\n",
       "p2_conf     0.0\n",
       "p2_dog      0.0\n",
       "p3          0.0\n",
       "p3_conf     0.0\n",
       "p3_dog      0.0\n",
       "dtype: float64"
      ]
     },
     "execution_count": 266,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check percentage of missing values\n",
    "(df_img_pred.isnull().sum()/df_img_pred.isnull().count())*100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 267,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "5HRcRH_v63e0",
    "outputId": "f982518c-137f-4233-c859-a7cc8a5cbb11"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tweet_id    2075\n",
       "jpg_url     2009\n",
       "img_num        4\n",
       "p1           378\n",
       "p1_conf     2006\n",
       "p1_dog         2\n",
       "p2           405\n",
       "p2_conf     2004\n",
       "p2_dog         2\n",
       "p3           408\n",
       "p3_conf     2006\n",
       "p3_dog         2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 267,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# count the number of unique values in each column of the df_img_pred\n",
    "df_img_pred.nunique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 268,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "leXw4Ka363e1",
    "outputId": "725eb4df-0aeb-4a5e-f1f4-ea38b144a8e5"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://pbs.twimg.com/media/CT4udn0WwAA0aMy.jpg\"/>"
      ],
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 268,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Image for tweet_id 666020888022790149 that was correctly predicted\n",
    "Image(url = 'https://pbs.twimg.com/media/CT4udn0WwAA0aMy.jpg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 269,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "VCbzYXSV63e5",
    "outputId": "688f3245-c002-438e-b75f-c809d8c5ec68"
   },
   "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>tweet_id</th>\n",
       "      <th>jpg_url</th>\n",
       "      <th>img_num</th>\n",
       "      <th>p1</th>\n",
       "      <th>p1_conf</th>\n",
       "      <th>p1_dog</th>\n",
       "      <th>p2</th>\n",
       "      <th>p2_conf</th>\n",
       "      <th>p2_dog</th>\n",
       "      <th>p3</th>\n",
       "      <th>p3_conf</th>\n",
       "      <th>p3_dog</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2.075000e+03</td>\n",
       "      <td>2075</td>\n",
       "      <td>2075.000000</td>\n",
       "      <td>2075</td>\n",
       "      <td>2075.000000</td>\n",
       "      <td>2075</td>\n",
       "      <td>2075</td>\n",
       "      <td>2.075000e+03</td>\n",
       "      <td>2075</td>\n",
       "      <td>2075</td>\n",
       "      <td>2.075000e+03</td>\n",
       "      <td>2075</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2009</td>\n",
       "      <td>NaN</td>\n",
       "      <td>378</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>405</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>408</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>NaN</td>\n",
       "      <td>https://pbs.twimg.com/media/CvJCabcWgAIoUxW.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>golden_retriever</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>Labrador_retriever</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "      <td>Labrador_retriever</td>\n",
       "      <td>NaN</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>NaN</td>\n",
       "      <td>150</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1532</td>\n",
       "      <td>104</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1553</td>\n",
       "      <td>79</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>7.384514e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.203855</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.594548</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.345886e-01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.032417e-02</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6.785203e+16</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.561875</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.271174</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.006657e-01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.090593e-02</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>6.660209e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.044333</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.011300e-08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.740170e-10</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>6.764835e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.364412</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.388625e-02</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.622240e-02</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>7.119988e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.588230</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.181810e-01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.944380e-02</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.932034e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.843855</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.955655e-01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>9.180755e-02</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>8.924206e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.880140e-01</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.734190e-01</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            tweet_id  ... p3_dog\n",
       "count   2.075000e+03  ...   2075\n",
       "unique           NaN  ...      2\n",
       "top              NaN  ...   True\n",
       "freq             NaN  ...   1499\n",
       "mean    7.384514e+17  ...    NaN\n",
       "std     6.785203e+16  ...    NaN\n",
       "min     6.660209e+17  ...    NaN\n",
       "25%     6.764835e+17  ...    NaN\n",
       "50%     7.119988e+17  ...    NaN\n",
       "75%     7.932034e+17  ...    NaN\n",
       "max     8.924206e+17  ...    NaN\n",
       "\n",
       "[11 rows x 12 columns]"
      ]
     },
     "execution_count": 269,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show basic statistics of the df_img_pred data \n",
    "df_img_pred.describe(include=\"all\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GWVDo4tKJEtZ"
   },
   "source": [
    "#### assessing tweet_json dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "ro0HcesRJEuB",
    "outputId": "92be2e6e-4640-4c4e-b38f-50ee63e8b961"
   },
   "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>created_at</th>\n",
       "      <th>id</th>\n",
       "      <th>id_str</th>\n",
       "      <th>full_text</th>\n",
       "      <th>truncated</th>\n",
       "      <th>display_text_range</th>\n",
       "      <th>entities</th>\n",
       "      <th>extended_entities</th>\n",
       "      <th>source</th>\n",
       "      <th>in_reply_to_status_id</th>\n",
       "      <th>in_reply_to_status_id_str</th>\n",
       "      <th>in_reply_to_user_id</th>\n",
       "      <th>in_reply_to_user_id_str</th>\n",
       "      <th>in_reply_to_screen_name</th>\n",
       "      <th>user</th>\n",
       "      <th>geo</th>\n",
       "      <th>coordinates</th>\n",
       "      <th>place</th>\n",
       "      <th>contributors</th>\n",
       "      <th>is_quote_status</th>\n",
       "      <th>retweet_count</th>\n",
       "      <th>favorite_count</th>\n",
       "      <th>favorited</th>\n",
       "      <th>retweeted</th>\n",
       "      <th>possibly_sensitive</th>\n",
       "      <th>possibly_sensitive_appealable</th>\n",
       "      <th>lang</th>\n",
       "      <th>retweeted_status</th>\n",
       "      <th>quoted_status_id</th>\n",
       "      <th>quoted_status_id_str</th>\n",
       "      <th>quoted_status_permalink</th>\n",
       "      <th>quoted_status</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2017-08-01 16:23:56+00:00</td>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>892420643555336192</td>\n",
       "      <td>This is Phineas. He's a mystical boy. Only eve...</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 85]</td>\n",
       "      <td>{'hashtags': [], 'symbols': [], 'user_mentions...</td>\n",
       "      <td>{'media': [{'id': 892420639486877696, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>7315</td>\n",
       "      <td>34860</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2017-08-01 00:17:27+00:00</td>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>892177421306343424</td>\n",
       "      <td>This is Tilly. She's just checking pup on you....</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 138]</td>\n",
       "      <td>{'hashtags': [], 'symbols': [], 'user_mentions...</td>\n",
       "      <td>{'media': [{'id': 892177413194625024, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>5465</td>\n",
       "      <td>30195</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2017-07-31 00:18:03+00:00</td>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>This is Archie. He is a rare Norwegian Pouncin...</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 121]</td>\n",
       "      <td>{'hashtags': [], 'symbols': [], 'user_mentions...</td>\n",
       "      <td>{'media': [{'id': 891815175371796480, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>3605</td>\n",
       "      <td>22733</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017-07-30 15:58:51+00:00</td>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>This is Darla. She commenced a snooze mid meal...</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 79]</td>\n",
       "      <td>{'hashtags': [], 'symbols': [], 'user_mentions...</td>\n",
       "      <td>{'media': [{'id': 891689552724799489, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>7505</td>\n",
       "      <td>38132</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2017-07-29 16:00:24+00:00</td>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>This is Franklin. He would like you to stop ca...</td>\n",
       "      <td>False</td>\n",
       "      <td>[0, 138]</td>\n",
       "      <td>{'hashtags': [{'text': 'BarkWeek', 'indices': ...</td>\n",
       "      <td>{'media': [{'id': 891327551943041024, 'id_str'...</td>\n",
       "      <td>&lt;a href=\"http://twitter.com/download/iphone\" r...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>{'id': 4196983835, 'id_str': '4196983835', 'na...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>NaN</td>\n",
       "      <td>False</td>\n",
       "      <td>8057</td>\n",
       "      <td>36394</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>en</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 created_at  ...  quoted_status\n",
       "0 2017-08-01 16:23:56+00:00  ...            NaN\n",
       "1 2017-08-01 00:17:27+00:00  ...            NaN\n",
       "2 2017-07-31 00:18:03+00:00  ...            NaN\n",
       "3 2017-07-30 15:58:51+00:00  ...            NaN\n",
       "4 2017-07-29 16:00:24+00:00  ...            NaN\n",
       "\n",
       "[5 rows x 32 columns]"
      ]
     },
     "execution_count": 270,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# visually asess the data\n",
    "df_tweet_json.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "AQY01LiWJEuK",
    "outputId": "0d1348bc-1e16-4b99-8cd3-f51866ff1fa2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "created_at                       datetime64[ns, UTC]\n",
       "id                                             int64\n",
       "id_str                                         int64\n",
       "full_text                                     object\n",
       "truncated                                       bool\n",
       "display_text_range                            object\n",
       "entities                                      object\n",
       "extended_entities                             object\n",
       "source                                        object\n",
       "in_reply_to_status_id                        float64\n",
       "in_reply_to_status_id_str                    float64\n",
       "in_reply_to_user_id                          float64\n",
       "in_reply_to_user_id_str                      float64\n",
       "in_reply_to_screen_name                       object\n",
       "user                                          object\n",
       "geo                                          float64\n",
       "coordinates                                  float64\n",
       "place                                         object\n",
       "contributors                                 float64\n",
       "is_quote_status                                 bool\n",
       "retweet_count                                  int64\n",
       "favorite_count                                 int64\n",
       "favorited                                       bool\n",
       "retweeted                                       bool\n",
       "possibly_sensitive                           float64\n",
       "possibly_sensitive_appealable                float64\n",
       "lang                                          object\n",
       "retweeted_status                              object\n",
       "quoted_status_id                             float64\n",
       "quoted_status_id_str                         float64\n",
       "quoted_status_permalink                       object\n",
       "quoted_status                                 object\n",
       "dtype: object"
      ]
     },
     "execution_count": 271,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check columns data types\n",
    "df_tweet_json.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "g78SGnTRJEuM",
    "outputId": "66683f88-b7ba-4971-dbc5-45b00c3c5076"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 2330 entries, 0 to 2329\n",
      "Data columns (total 32 columns):\n",
      " #   Column                         Non-Null Count  Dtype              \n",
      "---  ------                         --------------  -----              \n",
      " 0   created_at                     2330 non-null   datetime64[ns, UTC]\n",
      " 1   id                             2330 non-null   int64              \n",
      " 2   id_str                         2330 non-null   int64              \n",
      " 3   full_text                      2330 non-null   object             \n",
      " 4   truncated                      2330 non-null   bool               \n",
      " 5   display_text_range             2330 non-null   object             \n",
      " 6   entities                       2330 non-null   object             \n",
      " 7   extended_entities              2058 non-null   object             \n",
      " 8   source                         2330 non-null   object             \n",
      " 9   in_reply_to_status_id          77 non-null     float64            \n",
      " 10  in_reply_to_status_id_str      77 non-null     float64            \n",
      " 11  in_reply_to_user_id            77 non-null     float64            \n",
      " 12  in_reply_to_user_id_str        77 non-null     float64            \n",
      " 13  in_reply_to_screen_name        77 non-null     object             \n",
      " 14  user                           2330 non-null   object             \n",
      " 15  geo                            0 non-null      float64            \n",
      " 16  coordinates                    0 non-null      float64            \n",
      " 17  place                          1 non-null      object             \n",
      " 18  contributors                   0 non-null      float64            \n",
      " 19  is_quote_status                2330 non-null   bool               \n",
      " 20  retweet_count                  2330 non-null   int64              \n",
      " 21  favorite_count                 2330 non-null   int64              \n",
      " 22  favorited                      2330 non-null   bool               \n",
      " 23  retweeted                      2330 non-null   bool               \n",
      " 24  possibly_sensitive             2196 non-null   float64            \n",
      " 25  possibly_sensitive_appealable  2196 non-null   float64            \n",
      " 26  lang                           2330 non-null   object             \n",
      " 27  retweeted_status               163 non-null    object             \n",
      " 28  quoted_status_id               26 non-null     float64            \n",
      " 29  quoted_status_id_str           26 non-null     float64            \n",
      " 30  quoted_status_permalink        26 non-null     object             \n",
      " 31  quoted_status                  24 non-null     object             \n",
      "dtypes: bool(4), datetime64[ns, UTC](1), float64(11), int64(4), object(12)\n",
      "memory usage: 518.9+ KB\n"
     ]
    }
   ],
   "source": [
    "# show conscise summary of the twitter_json data\n",
    "df_tweet_json.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 273,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "gUNOGLz8QHjn",
    "outputId": "5561e112-25ac-4abc-9f39-431b6dca82a0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<a href=\"http://twitter.com/download/iphone\" rel=\"nofollow\">Twitter for iPhone</a>     2196\n",
       "<a href=\"http://vine.co\" rel=\"nofollow\">Vine - Make a Scene</a>                          91\n",
       "<a href=\"http://twitter.com\" rel=\"nofollow\">Twitter Web Client</a>                       33\n",
       "<a href=\"https://about.twitter.com/products/tweetdeck\" rel=\"nofollow\">TweetDeck</a>      10\n",
       "Name: source, dtype: int64"
      ]
     },
     "execution_count": 273,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check the most frequent source\n",
    "df_tweet_json['source'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 274,
   "metadata": {
    "collapsed": true,
    "id": "PZx3qULAQhxK"
   },
   "outputs": [],
   "source": [
    "# # count for original tweet and retweet\n",
    "# df_tweet_json.retweeted_status.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 275,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "ouyUjhJ1JEuN",
    "outputId": "0de41a9d-0304-4870-bc81-61c8f96143b8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "created_at                         0.000000\n",
       "id                                 0.000000\n",
       "id_str                             0.000000\n",
       "full_text                          0.000000\n",
       "truncated                          0.000000\n",
       "display_text_range                 0.000000\n",
       "entities                           0.000000\n",
       "extended_entities                 11.673820\n",
       "source                             0.000000\n",
       "in_reply_to_status_id             96.695279\n",
       "in_reply_to_status_id_str         96.695279\n",
       "in_reply_to_user_id               96.695279\n",
       "in_reply_to_user_id_str           96.695279\n",
       "in_reply_to_screen_name           96.695279\n",
       "user                               0.000000\n",
       "geo                              100.000000\n",
       "coordinates                      100.000000\n",
       "place                             99.957082\n",
       "contributors                     100.000000\n",
       "is_quote_status                    0.000000\n",
       "retweet_count                      0.000000\n",
       "favorite_count                     0.000000\n",
       "favorited                          0.000000\n",
       "retweeted                          0.000000\n",
       "possibly_sensitive                 5.751073\n",
       "possibly_sensitive_appealable      5.751073\n",
       "lang                               0.000000\n",
       "retweeted_status                  93.004292\n",
       "quoted_status_id                  98.884120\n",
       "quoted_status_id_str              98.884120\n",
       "quoted_status_permalink           98.884120\n",
       "quoted_status                     98.969957\n",
       "dtype: float64"
      ]
     },
     "execution_count": 275,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check percentage of missing values\n",
    "(df_tweet_json.isnull().sum()/df_tweet_json.isnull().count())*100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "vZdWIgTyJEuO"
   },
   "source": [
    "Note columns that are almost of missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 276,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "p7ml65UtJEuT",
    "outputId": "81fce7c7-a1fb-4e93-99dc-b216fdd99214"
   },
   "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>id</th>\n",
       "      <th>id_str</th>\n",
       "      <th>in_reply_to_status_id</th>\n",
       "      <th>in_reply_to_status_id_str</th>\n",
       "      <th>in_reply_to_user_id</th>\n",
       "      <th>in_reply_to_user_id_str</th>\n",
       "      <th>geo</th>\n",
       "      <th>coordinates</th>\n",
       "      <th>contributors</th>\n",
       "      <th>retweet_count</th>\n",
       "      <th>favorite_count</th>\n",
       "      <th>possibly_sensitive</th>\n",
       "      <th>possibly_sensitive_appealable</th>\n",
       "      <th>quoted_status_id</th>\n",
       "      <th>quoted_status_id_str</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>2.330000e+03</td>\n",
       "      <td>2.330000e+03</td>\n",
       "      <td>7.700000e+01</td>\n",
       "      <td>7.700000e+01</td>\n",
       "      <td>7.700000e+01</td>\n",
       "      <td>7.700000e+01</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2330.000000</td>\n",
       "      <td>2330.000000</td>\n",
       "      <td>2196.0</td>\n",
       "      <td>2196.0</td>\n",
       "      <td>2.600000e+01</td>\n",
       "      <td>2.600000e+01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>7.419357e+17</td>\n",
       "      <td>7.419357e+17</td>\n",
       "      <td>7.440692e+17</td>\n",
       "      <td>7.440692e+17</td>\n",
       "      <td>2.040329e+16</td>\n",
       "      <td>2.040329e+16</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2573.819742</td>\n",
       "      <td>7282.009013</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.113972e+17</td>\n",
       "      <td>8.113972e+17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6.823308e+16</td>\n",
       "      <td>6.823308e+16</td>\n",
       "      <td>7.524295e+16</td>\n",
       "      <td>7.524295e+16</td>\n",
       "      <td>1.260797e+17</td>\n",
       "      <td>1.260797e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4353.406503</td>\n",
       "      <td>11310.644578</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.295843e+16</td>\n",
       "      <td>6.295843e+16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>6.660209e+17</td>\n",
       "      <td>6.660209e+17</td>\n",
       "      <td>6.658147e+17</td>\n",
       "      <td>6.658147e+17</td>\n",
       "      <td>1.185634e+07</td>\n",
       "      <td>1.185634e+07</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.721083e+17</td>\n",
       "      <td>6.721083e+17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>6.782926e+17</td>\n",
       "      <td>6.782926e+17</td>\n",
       "      <td>6.757073e+17</td>\n",
       "      <td>6.757073e+17</td>\n",
       "      <td>3.589728e+08</td>\n",
       "      <td>3.589728e+08</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>521.500000</td>\n",
       "      <td>1260.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7.761338e+17</td>\n",
       "      <td>7.761338e+17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>7.183508e+17</td>\n",
       "      <td>7.183508e+17</td>\n",
       "      <td>7.032559e+17</td>\n",
       "      <td>7.032559e+17</td>\n",
       "      <td>4.196984e+09</td>\n",
       "      <td>4.196984e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1201.000000</td>\n",
       "      <td>3155.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.281173e+17</td>\n",
       "      <td>8.281173e+17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.986712e+17</td>\n",
       "      <td>7.986712e+17</td>\n",
       "      <td>8.233264e+17</td>\n",
       "      <td>8.233264e+17</td>\n",
       "      <td>4.196984e+09</td>\n",
       "      <td>4.196984e+09</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2980.000000</td>\n",
       "      <td>8897.500000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.637581e+17</td>\n",
       "      <td>8.637581e+17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>8.924206e+17</td>\n",
       "      <td>8.924206e+17</td>\n",
       "      <td>8.862664e+17</td>\n",
       "      <td>8.862664e+17</td>\n",
       "      <td>8.405479e+17</td>\n",
       "      <td>8.405479e+17</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>73824.000000</td>\n",
       "      <td>149760.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8.860534e+17</td>\n",
       "      <td>8.860534e+17</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 id        id_str  ...  quoted_status_id  quoted_status_id_str\n",
       "count  2.330000e+03  2.330000e+03  ...      2.600000e+01          2.600000e+01\n",
       "mean   7.419357e+17  7.419357e+17  ...      8.113972e+17          8.113972e+17\n",
       "std    6.823308e+16  6.823308e+16  ...      6.295843e+16          6.295843e+16\n",
       "min    6.660209e+17  6.660209e+17  ...      6.721083e+17          6.721083e+17\n",
       "25%    6.782926e+17  6.782926e+17  ...      7.761338e+17          7.761338e+17\n",
       "50%    7.183508e+17  7.183508e+17  ...      8.281173e+17          8.281173e+17\n",
       "75%    7.986712e+17  7.986712e+17  ...      8.637581e+17          8.637581e+17\n",
       "max    8.924206e+17  8.924206e+17  ...      8.860534e+17          8.860534e+17\n",
       "\n",
       "[8 rows x 15 columns]"
      ]
     },
     "execution_count": 276,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show basic statistics of the df_tweet_json data \n",
    "df_tweet_json.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "b13cLiSqPeK6"
   },
   "source": [
    "## Assessment Summary\n",
    "\n",
    "### Quality Issues\n",
    "\n",
    "twitter-archive-enhanced dataset\n",
    "\n",
    "1.   The `timestamp` column has the `object` datatype instead of `datetime`\n",
    "2.   The `name` column has None instead of `NaN` and also too many invalid values.\n",
    "3.   The interest only is on original tweet and not retweet, therefore retweets and replies should be removed from the data.\n",
    "4.   \n",
    "`in_reply_to_status_id`, `in_reply_to_user_id`, `retweeted_status_id`,\n",
    "`retweeted_status_user_id` and `retweeted_status_timestamp` columns are constituted of over 90% missing values and therefore should be dropped\n",
    "5.   The `rating_numerator` column contains some wrongly extracted values and should be corrected\n",
    "6.   The doggo, floofer, pupper and puppo columns has None for missing values. This should as well be corrected\n",
    "\n",
    "\n",
    "image_predictions dataset\n",
    "7.   The img_num column should be dropped as it is not really serving any purpose and also the columns (p1, p1_dog, p1_conf and so on) should be just breed and confidence.\n",
    "8.   The prediction dog breeds involve both uppercase and lowercase for the first letter. There should be consistency in the name format. \n",
    "\n",
    "tweet_json<br>\n",
    "\n",
    "9.  A lot of columns are useless in the dataset as they were made up of missing values. These columns should be dropped. Also, the number of columns in `df_tweet_json`  are too many and should be reduced  to tweet ID, retweet count,favorite count and Source\n",
    "10.   The `source` column has not been properly extracted as it contains the html tag `<a>`. This should be corrected\n",
    "11.   The id column name should be made to match that of twitter achive enhanced\n",
    "\n",
    "## Tidiness Issues\n",
    "\n",
    "- The doggo, floofer, pupper and puppo columns should be made into one categorical variable\n",
    "- There are 2075 image predictions, 281 less than the number of tweets in the twitter-archive-enhanced dataset (2356), so will be classified as \"missing data\".\n",
    "- The twitter-archive-enhanced, image_predictions, and tweet_json dataframes should be made into a dataset.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7FMmYfxB7GRb"
   },
   "source": [
    "## Cleaning Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 277,
   "metadata": {
    "collapsed": true,
    "id": "Acpp6heR8ZJJ"
   },
   "outputs": [],
   "source": [
    "# make copies of individual dataframes to perform the cleaning on\n",
    "twitter_archive_df = df_twit_arch.copy()\n",
    "image_prediction_df = df_img_pred.copy()\n",
    "tweet_json_df = df_tweet_json.copy()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ctFaf2gn7mzy"
   },
   "source": [
    "#### Define_1:\n",
    "The timestamp column has the object datatype and should be converted into  datetime"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mJukQfc4730g"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 278,
   "metadata": {
    "collapsed": true,
    "id": "T5ih_0vVPjXS"
   },
   "outputs": [],
   "source": [
    "twitter_archive_df['timestamp'] = pd.to_datetime(twitter_archive_df['timestamp'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "lUbnFNONAZgQ"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 279,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "eyXQSAwNAOnh",
    "outputId": "cd88c8c3-5f9f-4348-bc27-70a7f6f62ce9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "datetime64[ns, UTC]"
      ]
     },
     "execution_count": 279,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "twitter_archive_df['timestamp'].dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4NBOWbEwAw7c"
   },
   "source": [
    "#### Define_2:\n",
    "The name column has None instead of NaN and also too many invalid values. This should be corrected by regarding all None and invalid values as NaN"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "c6UrVdbHAw7s"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 280,
   "metadata": {
    "collapsed": true,
    "id": "qLruJ5jJAw7t"
   },
   "outputs": [],
   "source": [
    "# replace all invalid names from with None to later be handled as NaN\n",
    "twitter_archive_df['name'] = twitter_archive_df['name'].str.replace('^[a-z]+', 'None')\n",
    "twitter_archive_df['name'].replace({\"None\": np.nan}, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "b0rQ8-1JAw7u"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 281,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "N0BF4g0MAw7v",
    "outputId": "66046b0e-ba4a-480f-c8d8-b0cde16679a9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Charlie     12\n",
       "Oliver      11\n",
       "Lucy        11\n",
       "Cooper      11\n",
       "Penny       10\n",
       "            ..\n",
       "Fletcher     1\n",
       "Bobby        1\n",
       "Dug          1\n",
       "Ridley       1\n",
       "Andy         1\n",
       "Name: name, Length: 931, dtype: int64"
      ]
     },
     "execution_count": 281,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# count number of time each names now appear and see if the 'None' is there\n",
    "twitter_archive_df['name'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "GPWhU4I_JvZW"
   },
   "source": [
    "#### Define_3:\n",
    "The interest only is on original tweet and not retweet, therefore retweets and replies should be removed from the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "uX10aHJ-JvZ7"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 282,
   "metadata": {
    "collapsed": true,
    "id": "8scshkCUJvZ9"
   },
   "outputs": [],
   "source": [
    "# considerrows with missing values in the retweets columns as original tweet\n",
    "\n",
    "twitter_archive_df = twitter_archive_df.query('in_reply_to_status_id == \"NaN\" & retweeted_status_user_id == \"NaN\" & in_reply_to_user_id == \"NaN\" & retweeted_status_id == \"NaN\"')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jzrY2jC_JvZ_"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 283,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "yhh-tXR8JvaA",
    "outputId": "4817d120-fd3a-4082-a5af-c64c1fe97d81"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 2097 entries, 0 to 2355\n",
      "Data columns (total 17 columns):\n",
      " #   Column                      Non-Null Count  Dtype              \n",
      "---  ------                      --------------  -----              \n",
      " 0   tweet_id                    2097 non-null   int64              \n",
      " 1   in_reply_to_status_id       0 non-null      float64            \n",
      " 2   in_reply_to_user_id         0 non-null      float64            \n",
      " 3   timestamp                   2097 non-null   datetime64[ns, UTC]\n",
      " 4   source                      2097 non-null   object             \n",
      " 5   text                        2097 non-null   object             \n",
      " 6   retweeted_status_id         0 non-null      float64            \n",
      " 7   retweeted_status_user_id    0 non-null      float64            \n",
      " 8   retweeted_status_timestamp  0 non-null      object             \n",
      " 9   expanded_urls               2094 non-null   object             \n",
      " 10  rating_numerator            2097 non-null   int64              \n",
      " 11  rating_denominator          2097 non-null   int64              \n",
      " 12  name                        1390 non-null   object             \n",
      " 13  doggo                       2097 non-null   object             \n",
      " 14  floofer                     2097 non-null   object             \n",
      " 15  pupper                      2097 non-null   object             \n",
      " 16  puppo                       2097 non-null   object             \n",
      "dtypes: datetime64[ns, UTC](1), float64(4), int64(3), object(9)\n",
      "memory usage: 294.9+ KB\n"
     ]
    }
   ],
   "source": [
    "# check if nonnull values are still in the retweet columns\n",
    "twitter_archive_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "6VrdJfwWBFcT"
   },
   "source": [
    "#### Define_4:\n",
    "in_reply_to_status_id, in_reply_to_user_id, retweeted_status_id, retweeted_status_user_id and retweeted_status_timestamp columns are constituted of over 90% missing values and therefore should be dropped. Columns that are of no interest would also be dropped"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "S1jn15k7BFcz"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 284,
   "metadata": {
    "collapsed": true,
    "id": "qJfzPFNwBFc0"
   },
   "outputs": [],
   "source": [
    "# list columns to be dropped\n",
    "cols = ['in_reply_to_status_id','in_reply_to_user_id','retweeted_status_id', 'retweeted_status_user_id', 'retweeted_status_timestamp','source','expanded_urls']\n",
    "twitter_archive_df.drop(columns = cols, axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "09u1S6G-BFc2"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "HUTdm8jxBFc3",
    "outputId": "677cb2ec-067f-46b6-b1df-e054024f6c72"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 2097 entries, 0 to 2355\n",
      "Data columns (total 10 columns):\n",
      " #   Column              Non-Null Count  Dtype              \n",
      "---  ------              --------------  -----              \n",
      " 0   tweet_id            2097 non-null   int64              \n",
      " 1   timestamp           2097 non-null   datetime64[ns, UTC]\n",
      " 2   text                2097 non-null   object             \n",
      " 3   rating_numerator    2097 non-null   int64              \n",
      " 4   rating_denominator  2097 non-null   int64              \n",
      " 5   name                1390 non-null   object             \n",
      " 6   doggo               2097 non-null   object             \n",
      " 7   floofer             2097 non-null   object             \n",
      " 8   pupper              2097 non-null   object             \n",
      " 9   puppo               2097 non-null   object             \n",
      "dtypes: datetime64[ns, UTC](1), int64(3), object(6)\n",
      "memory usage: 180.2+ KB\n"
     ]
    }
   ],
   "source": [
    "# confirm they have been dropped and also check for missing values\n",
    "twitter_archive_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 285,
   "metadata": {
    "collapsed": true,
    "id": "ls5gdmVWBFc6"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8Nn_LUzPBGdJ"
   },
   "source": [
    "#### Define_5:\n",
    "The `rating_numerator` column contains some wrongly extracted values and should be corrected"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "WEGXvIzGBGdN"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 286,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "bF1WACHIBGdO",
    "outputId": "476b5515-25a7-4008-9095-95ad98b2bfa8"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/pandas/core/strings.py:2001: UserWarning: This pattern has match groups. To actually get the groups, use str.extract.\n",
      "  return func(self, *args, **kwargs)\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>text</th>\n",
       "      <th>rating_numerator</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>This is Bella. She hopes her smile made you sm...</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>695</th>\n",
       "      <td>This is Logan, the Chow who lived. He solemnly...</td>\n",
       "      <td>75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>763</th>\n",
       "      <td>This is Sophie. She's a Jubilant Bush Pupper. ...</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1712</th>\n",
       "      <td>Here we have uncovered an entire battalion of ...</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                   text  rating_numerator\n",
       "45    This is Bella. She hopes her smile made you sm...                 5\n",
       "695   This is Logan, the Chow who lived. He solemnly...                75\n",
       "763   This is Sophie. She's a Jubilant Bush Pupper. ...                27\n",
       "1712  Here we have uncovered an entire battalion of ...                26"
      ]
     },
     "execution_count": 286,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get where the numerators doesn't match\n",
    "twitter_archive_df[twitter_archive_df.text.str.contains(r\"(\\d+\\.\\d*\\/\\d+)\")][['text', 'rating_numerator']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "PQbI6y0dxqEc",
    "outputId": "243c1300-8d52-4cb6-8ada-f5de6e1422ba"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "          0\n",
      "45     13.5\n",
      "695    9.75\n",
      "763   11.27\n",
      "1712  11.26\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.7/dist-packages/pandas/core/strings.py:2001: UserWarning: This pattern has match groups. To actually get the groups, use str.extract.\n",
      "  return func(self, *args, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "correct_rating = twitter_archive_df[twitter_archive_df.text.str.contains(r\"(\\d+\\.\\d*\\/\\d+)\")]['text'].str.extract(r\"(\\d+\\.\\d*(?=\\/\\d+))\")\n",
    "print(correct_rating)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 288,
   "metadata": {
    "collapsed": true,
    "id": "pHPAk00ZyLgU"
   },
   "outputs": [],
   "source": [
    "# assign the correct ratings to their appropriate rows\n",
    "twitter_archive_df.loc[correct_rating.index, 'rating_numerator'] = correct_rating.values\n",
    "# datatype to float\n",
    "twitter_archive_df['rating_numerator'] = twitter_archive_df['rating_numerator'].astype('float')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CWi12cuABGdQ"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "wOzG_sMqBGdS",
    "outputId": "e0b3ca7a-9936-400e-b673-8504c3f3a8ea"
   },
   "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>tweet_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>text</th>\n",
       "      <th>rating_numerator</th>\n",
       "      <th>rating_denominator</th>\n",
       "      <th>name</th>\n",
       "      <th>doggo</th>\n",
       "      <th>floofer</th>\n",
       "      <th>pupper</th>\n",
       "      <th>puppo</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>883482846933004288</td>\n",
       "      <td>2017-07-08 00:28:19+00:00</td>\n",
       "      <td>This is Bella. She hopes her smile made you sm...</td>\n",
       "      <td>13.50</td>\n",
       "      <td>10</td>\n",
       "      <td>Bella</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>695</th>\n",
       "      <td>786709082849828864</td>\n",
       "      <td>2016-10-13 23:23:56+00:00</td>\n",
       "      <td>This is Logan, the Chow who lived. He solemnly...</td>\n",
       "      <td>9.75</td>\n",
       "      <td>10</td>\n",
       "      <td>Logan</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>763</th>\n",
       "      <td>778027034220126208</td>\n",
       "      <td>2016-09-20 00:24:34+00:00</td>\n",
       "      <td>This is Sophie. She's a Jubilant Bush Pupper. ...</td>\n",
       "      <td>11.27</td>\n",
       "      <td>10</td>\n",
       "      <td>Sophie</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>pupper</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1712</th>\n",
       "      <td>680494726643068929</td>\n",
       "      <td>2015-12-25 21:06:00+00:00</td>\n",
       "      <td>Here we have uncovered an entire battalion of ...</td>\n",
       "      <td>11.26</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                tweet_id                 timestamp  ...  pupper  puppo\n",
       "45    883482846933004288 2017-07-08 00:28:19+00:00  ...    None   None\n",
       "695   786709082849828864 2016-10-13 23:23:56+00:00  ...    None   None\n",
       "763   778027034220126208 2016-09-20 00:24:34+00:00  ...  pupper   None\n",
       "1712  680494726643068929 2015-12-25 21:06:00+00:00  ...    None   None\n",
       "\n",
       "[4 rows x 10 columns]"
      ]
     },
     "execution_count": 289,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# check the corrected rows\n",
    "twitter_archive_df.loc[correct_rating.index]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 289,
   "metadata": {
    "collapsed": true,
    "id": "gZKbBTeiBGdV"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "hq3hXQK_BHQx"
   },
   "source": [
    "#### Define_6:\n",
    "The source column has not been properly extracted as it contains the html tag `<a>`. This should be corrected and properly extracted"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4N5BENZIBHQ1"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 290,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "yS3URZBIBHQ3",
    "outputId": "6bc04128-4ad4-448e-bf67-490ad723bdc6"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<a href=\"http://twitter.com/download/iphone\" rel=\"nofollow\">Twitter for iPhone</a>     2196\n",
       "<a href=\"http://vine.co\" rel=\"nofollow\">Vine - Make a Scene</a>                          91\n",
       "<a href=\"http://twitter.com\" rel=\"nofollow\">Twitter Web Client</a>                       33\n",
       "<a href=\"https://about.twitter.com/products/tweetdeck\" rel=\"nofollow\">TweetDeck</a>      10\n",
       "Name: source, dtype: int64"
      ]
     },
     "execution_count": 290,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# have a look again\n",
    "tweet_json_df['source'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "rzTTPOEdqdbA"
   },
   "source": [
    "Reference: https://pandas.pydata.org/docs/reference/api/pandas.Series.str.extract.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 291,
   "metadata": {
    "collapsed": true,
    "id": "5JNGzBebooMJ"
   },
   "outputs": [],
   "source": [
    "# extract the correct sources by creating group from any line starting with < followed by 'a' and many or more characters and ends with </a>  \n",
    "tweet_json_df['source'] = tweet_json_df['source'].str.extract('^<a.+>(.+)</a>$')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eecihzrVBHQ6"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 292,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "__YDTMjfBHQ8",
    "outputId": "e86c5fc1-207d-4af1-9191-606c087c19e4"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Twitter for iPhone     2196\n",
       "Vine - Make a Scene      91\n",
       "Twitter Web Client       33\n",
       "TweetDeck                10\n",
       "Name: source, dtype: int64"
      ]
     },
     "execution_count": 292,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# have a look again\n",
    "tweet_json_df['source'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9nNEz66JBH8-"
   },
   "source": [
    "#### Define_7:\n",
    "A lot of columns are useless in the tweet_json_df dataset as they were made up of missing values. These columns should be dropped and also the dataset be reduced to tweet ID, retweet count,favorite count and Source"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "A9kItSyLBH9C"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 293,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "2l4mLAy0sXKk",
    "outputId": "0ac9478a-9dc3-4c2f-c8e2-4013b0a210c8"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['created_at', 'id', 'id_str', 'full_text', 'truncated',\n",
       "       'display_text_range', 'entities', 'extended_entities', 'source',\n",
       "       'in_reply_to_status_id', 'in_reply_to_status_id_str',\n",
       "       'in_reply_to_user_id', 'in_reply_to_user_id_str',\n",
       "       'in_reply_to_screen_name', 'user', 'geo', 'coordinates', 'place',\n",
       "       'contributors', 'is_quote_status', 'retweet_count', 'favorite_count',\n",
       "       'favorited', 'retweeted', 'possibly_sensitive',\n",
       "       'possibly_sensitive_appealable', 'lang', 'retweeted_status',\n",
       "       'quoted_status_id', 'quoted_status_id_str', 'quoted_status_permalink',\n",
       "       'quoted_status'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 293,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tweet_json_df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 294,
   "metadata": {
    "collapsed": true,
    "id": "8h42xoZaBH9F"
   },
   "outputs": [],
   "source": [
    "# select 'id','retweet_count', 'favorite_count','source'\n",
    "tweet_json_df = tweet_json_df[['id','retweet_count', 'favorite_count','source']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "uiPI5qSTBH9I"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 295,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "XOEzeYM9BH9K",
    "outputId": "5511a793-807d-4339-a8eb-bdb66ad41617"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 2330 entries, 0 to 2329\n",
      "Data columns (total 4 columns):\n",
      " #   Column          Non-Null Count  Dtype \n",
      "---  ------          --------------  ----- \n",
      " 0   id              2330 non-null   int64 \n",
      " 1   retweet_count   2330 non-null   int64 \n",
      " 2   favorite_count  2330 non-null   int64 \n",
      " 3   source          2330 non-null   object\n",
      "dtypes: int64(3), object(1)\n",
      "memory usage: 72.9+ KB\n"
     ]
    }
   ],
   "source": [
    "# see the columns present\n",
    "tweet_json_df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "l3tezM_UBIqh"
   },
   "source": [
    "#### Define_8:\n",
    "The doggo, floofer, pupper and puppo columns should be made into one categorical variable and also np.nan inplace of None"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XiCBQY8vBIqk"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "dtq3zyG94l_7"
   },
   "source": [
    "Ref: https://pandas.pydata.org/docs/reference/api/pandas.Series.str.cat.html"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 296,
   "metadata": {
    "collapsed": true,
    "id": "TdtCOPc6BIqm"
   },
   "outputs": [],
   "source": [
    "# create new temprary columns with the None removed\n",
    "dog_stage = ['doggo','pupper', 'floofer', 'puppo' ]\n",
    "for i in dog_stage:\n",
    "        twitter_archive_df[i] = twitter_archive_df[i].replace('None', '')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 297,
   "metadata": {
    "collapsed": true,
    "id": "tbODzYR54Adu"
   },
   "outputs": [],
   "source": [
    "# concatenates the Series of twitter_archive_df.doggo and elements of others element-wise\n",
    "twitter_archive_df['dog_stage'] = twitter_archive_df.doggo.str.cat(twitter_archive_df.floofer).str.cat(twitter_archive_df.pupper).str.cat(twitter_archive_df.puppo)\n",
    "# the temporary columns are dropped\n",
    "twitter_archive_df = twitter_archive_df.drop(['doggo','floofer','pupper','puppo'], axis = 1)\n",
    "# use np.nan to fill the empty\n",
    "twitter_archive_df['dog_stage'] = twitter_archive_df['dog_stage'].replace('', np.nan)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "IXUmOkJnBIqo"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "nftMkI_QBIqr",
    "outputId": "0ee88796-54ec-4970-e2b8-9429f571c0d5"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0        NaN\n",
       "1        NaN\n",
       "2        NaN\n",
       "3        NaN\n",
       "4        NaN\n",
       "5        NaN\n",
       "6        NaN\n",
       "7        NaN\n",
       "8        NaN\n",
       "9      doggo\n",
       "10       NaN\n",
       "11       NaN\n",
       "12     puppo\n",
       "13       NaN\n",
       "14     puppo\n",
       "15       NaN\n",
       "16       NaN\n",
       "17       NaN\n",
       "18       NaN\n",
       "20       NaN\n",
       "21       NaN\n",
       "22       NaN\n",
       "23       NaN\n",
       "24       NaN\n",
       "25       NaN\n",
       "26       NaN\n",
       "27       NaN\n",
       "28       NaN\n",
       "29    pupper\n",
       "31       NaN\n",
       "Name: dog_stage, dtype: object"
      ]
     },
     "execution_count": 298,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show the first 30 rows\n",
    "twitter_archive_df['dog_stage'].head(30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 298,
   "metadata": {
    "collapsed": true,
    "id": "mn6w8XvABIqv"
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "54K-vZ5_6oWV"
   },
   "source": [
    "#### Define_9: \n",
    "*The* img_num column in image_prediction dataset should be dropped as it is really of no use and also the columns (p1, p1_dog, p1_conf and so on) should be just breed and confidence. A new breed name should be generated basedon the highest confidence of prediction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cOxUTc6J61q6"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 299,
   "metadata": {
    "collapsed": true,
    "id": "Tl3tEUj661rA"
   },
   "outputs": [],
   "source": [
    "image_prediction_df['dog_breed'] = 'None'\n",
    "\n",
    "breed = []\n",
    "confidence = []\n",
    "# iterating over image_prediction_df row by row and taking the highest confident prediction other wise np.nan\n",
    "for index, row in image_prediction_df.iterrows():\n",
    "    if row['p1_dog'] and row['p1_conf'] == max([row['p1_conf'], row['p2_conf'], row['p3_conf']]):\n",
    "        breed.append(row['p1'])\n",
    "        confidence.append(row['p1_conf'])\n",
    "    elif row['p2_dog'] and row['p2_conf'] == max([row['p1_conf'], row['p2_conf'], row['p3_conf']]):\n",
    "        breed.append(row['p2'])\n",
    "        confidence.append(row['p2_conf'])\n",
    "    elif row['p3_dog'] and row['p3_conf'] == max([row['p1_conf'], row['p2_conf'], row['p3_conf']]):\n",
    "        breed.append(row['p3'])\n",
    "        confidence.append(row['p3_conf'])\n",
    "    else:\n",
    "        breed.append(np.nan)\n",
    "        confidence.append(np.nan)\n",
    "        \n",
    "image_prediction_df['breed'] = breed\n",
    "image_prediction_df['confidence'] = confidence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 300,
   "metadata": {
    "collapsed": true,
    "id": "CAkHeKpFA18r"
   },
   "outputs": [],
   "source": [
    "# select the newly created columns which automatically drops other columns\n",
    "image_prediction_df = image_prediction_df[['tweet_id', 'jpg_url', 'breed', 'confidence']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "AkB07tKi61rF"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 301,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "wT3y1y6v61rG",
    "outputId": "5b47fe7a-9e79-4449-c41c-64ea58266174"
   },
   "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>tweet_id</th>\n",
       "      <th>jpg_url</th>\n",
       "      <th>breed</th>\n",
       "      <th>confidence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>666020888022790149</td>\n",
       "      <td>https://pbs.twimg.com/media/CT4udn0WwAA0aMy.jpg</td>\n",
       "      <td>Welsh_springer_spaniel</td>\n",
       "      <td>0.465074</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>666029285002620928</td>\n",
       "      <td>https://pbs.twimg.com/media/CT42GRgUYAA5iDo.jpg</td>\n",
       "      <td>redbone</td>\n",
       "      <td>0.506826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>666033412701032449</td>\n",
       "      <td>https://pbs.twimg.com/media/CT4521TWwAEvMyu.jpg</td>\n",
       "      <td>German_shepherd</td>\n",
       "      <td>0.596461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>666044226329800704</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5Dr8HUEAA-lEu.jpg</td>\n",
       "      <td>Rhodesian_ridgeback</td>\n",
       "      <td>0.408143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>666049248165822465</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5IQmsXIAAKY4A.jpg</td>\n",
       "      <td>miniature_pinscher</td>\n",
       "      <td>0.560311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>666050758794694657</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5Jof1WUAEuVxN.jpg</td>\n",
       "      <td>Bernese_mountain_dog</td>\n",
       "      <td>0.651137</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>666051853826850816</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5KoJ1WoAAJash.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>666055525042405380</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5N9tpXIAAifs1.jpg</td>\n",
       "      <td>chow</td>\n",
       "      <td>0.692517</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>666057090499244032</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5PY90WoAAQGLo.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>666058600524156928</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5Qw94XAAA_2dP.jpg</td>\n",
       "      <td>miniature_poodle</td>\n",
       "      <td>0.201493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>666063827256086533</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5Vg_wXIAAXfnj.jpg</td>\n",
       "      <td>golden_retriever</td>\n",
       "      <td>0.775930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>666071193221509120</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5cN_3WEAAlOoZ.jpg</td>\n",
       "      <td>Gordon_setter</td>\n",
       "      <td>0.503672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>666073100786774016</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5d9DZXAAALcwe.jpg</td>\n",
       "      <td>Walker_hound</td>\n",
       "      <td>0.260857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>666082916733198337</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5m4VGWEAAtKc8.jpg</td>\n",
       "      <td>pug</td>\n",
       "      <td>0.489814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>666094000022159362</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5w9gUW4AAsBNN.jpg</td>\n",
       "      <td>bloodhound</td>\n",
       "      <td>0.195217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>666099513787052032</td>\n",
       "      <td>https://pbs.twimg.com/media/CT51-JJUEAA6hV8.jpg</td>\n",
       "      <td>Lhasa</td>\n",
       "      <td>0.582330</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>666102155909144576</td>\n",
       "      <td>https://pbs.twimg.com/media/CT54YGiWUAEZnoK.jpg</td>\n",
       "      <td>English_setter</td>\n",
       "      <td>0.298617</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>666104133288665088</td>\n",
       "      <td>https://pbs.twimg.com/media/CT56LSZWoAAlJj2.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>666268910803644416</td>\n",
       "      <td>https://pbs.twimg.com/media/CT8QCd1WEAADXws.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>666273097616637952</td>\n",
       "      <td>https://pbs.twimg.com/media/CT8T1mtUwAA3aqm.jpg</td>\n",
       "      <td>Italian_greyhound</td>\n",
       "      <td>0.176053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>666287406224695296</td>\n",
       "      <td>https://pbs.twimg.com/media/CT8g3BpUEAAuFjg.jpg</td>\n",
       "      <td>Maltese_dog</td>\n",
       "      <td>0.857531</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>666293911632134144</td>\n",
       "      <td>https://pbs.twimg.com/media/CT8mx7KW4AEQu8N.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>666337882303524864</td>\n",
       "      <td>https://pbs.twimg.com/media/CT9OwFIWEAMuRje.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>666345417576210432</td>\n",
       "      <td>https://pbs.twimg.com/media/CT9Vn7PWoAA_ZCM.jpg</td>\n",
       "      <td>golden_retriever</td>\n",
       "      <td>0.858744</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>666353288456101888</td>\n",
       "      <td>https://pbs.twimg.com/media/CT9cx0tUEAAhNN_.jpg</td>\n",
       "      <td>malamute</td>\n",
       "      <td>0.336874</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>666362758909284353</td>\n",
       "      <td>https://pbs.twimg.com/media/CT9lXGsUcAAyUFt.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>666373753744588802</td>\n",
       "      <td>https://pbs.twimg.com/media/CT9vZEYWUAAlZ05.jpg</td>\n",
       "      <td>soft-coated_wheaten_terrier</td>\n",
       "      <td>0.326467</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>666396247373291520</td>\n",
       "      <td>https://pbs.twimg.com/media/CT-D2ZHWIAA3gK1.jpg</td>\n",
       "      <td>Chihuahua</td>\n",
       "      <td>0.978108</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>666407126856765440</td>\n",
       "      <td>https://pbs.twimg.com/media/CT-NvwmW4AAugGZ.jpg</td>\n",
       "      <td>black-and-tan_coonhound</td>\n",
       "      <td>0.529139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>666411507551481857</td>\n",
       "      <td>https://pbs.twimg.com/media/CT-RugiWIAELEaq.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              tweet_id  ... confidence\n",
       "0   666020888022790149  ...   0.465074\n",
       "1   666029285002620928  ...   0.506826\n",
       "2   666033412701032449  ...   0.596461\n",
       "3   666044226329800704  ...   0.408143\n",
       "4   666049248165822465  ...   0.560311\n",
       "5   666050758794694657  ...   0.651137\n",
       "6   666051853826850816  ...        NaN\n",
       "7   666055525042405380  ...   0.692517\n",
       "8   666057090499244032  ...        NaN\n",
       "9   666058600524156928  ...   0.201493\n",
       "10  666063827256086533  ...   0.775930\n",
       "11  666071193221509120  ...   0.503672\n",
       "12  666073100786774016  ...   0.260857\n",
       "13  666082916733198337  ...   0.489814\n",
       "14  666094000022159362  ...   0.195217\n",
       "15  666099513787052032  ...   0.582330\n",
       "16  666102155909144576  ...   0.298617\n",
       "17  666104133288665088  ...        NaN\n",
       "18  666268910803644416  ...        NaN\n",
       "19  666273097616637952  ...   0.176053\n",
       "20  666287406224695296  ...   0.857531\n",
       "21  666293911632134144  ...        NaN\n",
       "22  666337882303524864  ...        NaN\n",
       "23  666345417576210432  ...   0.858744\n",
       "24  666353288456101888  ...   0.336874\n",
       "25  666362758909284353  ...        NaN\n",
       "26  666373753744588802  ...   0.326467\n",
       "27  666396247373291520  ...   0.978108\n",
       "28  666407126856765440  ...   0.529139\n",
       "29  666411507551481857  ...        NaN\n",
       "\n",
       "[30 rows x 4 columns]"
      ]
     },
     "execution_count": 301,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show the first 30 row\n",
    "image_prediction_df.head(30)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "55cH3cdDBJZ_"
   },
   "source": [
    "#### Define_10:\n",
    "\n",
    "The `id` column in the `tweet_json_df` doesn't match that of `twitter_achived_enhanced` and therefore should be renamed for consistency"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "j9HNB40JBJaC"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 302,
   "metadata": {
    "collapsed": true,
    "id": "a3L55ICzBJaE"
   },
   "outputs": [],
   "source": [
    "# rename the column name\n",
    "tweet_json_df.rename(columns = {'id':'tweet_id'}, inplace = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "nVHhK8XSBJaI"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "jmjON4K0BJaK",
    "outputId": "b4701a2a-f3ea-4548-90ee-884d66481056"
   },
   "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>tweet_id</th>\n",
       "      <th>retweet_count</th>\n",
       "      <th>favorite_count</th>\n",
       "      <th>source</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>7315</td>\n",
       "      <td>34860</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>5465</td>\n",
       "      <td>30195</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>3605</td>\n",
       "      <td>22733</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>7505</td>\n",
       "      <td>38132</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>8057</td>\n",
       "      <td>36394</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2325</th>\n",
       "      <td>666049248165822465</td>\n",
       "      <td>38</td>\n",
       "      <td>92</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2326</th>\n",
       "      <td>666044226329800704</td>\n",
       "      <td>121</td>\n",
       "      <td>258</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2327</th>\n",
       "      <td>666033412701032449</td>\n",
       "      <td>39</td>\n",
       "      <td>105</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2328</th>\n",
       "      <td>666029285002620928</td>\n",
       "      <td>41</td>\n",
       "      <td>117</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2329</th>\n",
       "      <td>666020888022790149</td>\n",
       "      <td>442</td>\n",
       "      <td>2338</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2330 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                tweet_id  retweet_count  favorite_count              source\n",
       "0     892420643555336193           7315           34860  Twitter for iPhone\n",
       "1     892177421306343426           5465           30195  Twitter for iPhone\n",
       "2     891815181378084864           3605           22733  Twitter for iPhone\n",
       "3     891689557279858688           7505           38132  Twitter for iPhone\n",
       "4     891327558926688256           8057           36394  Twitter for iPhone\n",
       "...                  ...            ...             ...                 ...\n",
       "2325  666049248165822465             38              92  Twitter for iPhone\n",
       "2326  666044226329800704            121             258  Twitter for iPhone\n",
       "2327  666033412701032449             39             105  Twitter for iPhone\n",
       "2328  666029285002620928             41             117  Twitter for iPhone\n",
       "2329  666020888022790149            442            2338  Twitter for iPhone\n",
       "\n",
       "[2330 rows x 4 columns]"
      ]
     },
     "execution_count": 303,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show\n",
    "tweet_json_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "S4bONzckEMPa"
   },
   "source": [
    "#### Define_11:\n",
    "\n",
    "Preview the cleaned image_predictions dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kYue7iP1EMPz"
   },
   "source": [
    "##### Code/Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "Iz3s-OkHEMP2",
    "outputId": "866c2a92-0e91-4e5b-ceb4-cf1d99b7f9d7"
   },
   "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>tweet_id</th>\n",
       "      <th>jpg_url</th>\n",
       "      <th>breed</th>\n",
       "      <th>confidence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>666020888022790149</td>\n",
       "      <td>https://pbs.twimg.com/media/CT4udn0WwAA0aMy.jpg</td>\n",
       "      <td>Welsh_springer_spaniel</td>\n",
       "      <td>0.465074</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>666029285002620928</td>\n",
       "      <td>https://pbs.twimg.com/media/CT42GRgUYAA5iDo.jpg</td>\n",
       "      <td>redbone</td>\n",
       "      <td>0.506826</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>666033412701032449</td>\n",
       "      <td>https://pbs.twimg.com/media/CT4521TWwAEvMyu.jpg</td>\n",
       "      <td>German_shepherd</td>\n",
       "      <td>0.596461</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>666044226329800704</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5Dr8HUEAA-lEu.jpg</td>\n",
       "      <td>Rhodesian_ridgeback</td>\n",
       "      <td>0.408143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>666049248165822465</td>\n",
       "      <td>https://pbs.twimg.com/media/CT5IQmsXIAAKY4A.jpg</td>\n",
       "      <td>miniature_pinscher</td>\n",
       "      <td>0.560311</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2070</th>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>https://pbs.twimg.com/media/DF6hr6BUMAAzZgT.jpg</td>\n",
       "      <td>basset</td>\n",
       "      <td>0.555712</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2071</th>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>https://pbs.twimg.com/media/DF_q7IAWsAEuuN8.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2072</th>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>https://pbs.twimg.com/media/DGBdLU1WsAANxJ9.jpg</td>\n",
       "      <td>Chihuahua</td>\n",
       "      <td>0.716012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2073</th>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>https://pbs.twimg.com/media/DGGmoV4XsAAUL6n.jpg</td>\n",
       "      <td>Chihuahua</td>\n",
       "      <td>0.323581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2074</th>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>https://pbs.twimg.com/media/DGKD1-bXoAAIAUK.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2075 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                tweet_id  ... confidence\n",
       "0     666020888022790149  ...   0.465074\n",
       "1     666029285002620928  ...   0.506826\n",
       "2     666033412701032449  ...   0.596461\n",
       "3     666044226329800704  ...   0.408143\n",
       "4     666049248165822465  ...   0.560311\n",
       "...                  ...  ...        ...\n",
       "2070  891327558926688256  ...   0.555712\n",
       "2071  891689557279858688  ...        NaN\n",
       "2072  891815181378084864  ...   0.716012\n",
       "2073  892177421306343426  ...   0.323581\n",
       "2074  892420643555336193  ...        NaN\n",
       "\n",
       "[2075 rows x 4 columns]"
      ]
     },
     "execution_count": 304,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show\n",
    "image_prediction_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ye2cpAdME40x"
   },
   "source": [
    "#### Define_12:\n",
    "\n",
    "Preview the cleaned twitter-archive-enhanced dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "J07POiEBE41e"
   },
   "source": [
    "##### Code/Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 305,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "Tjg4Vsr7E41f",
    "outputId": "3746e3f1-3964-4491-d0b3-de2e0f829b55"
   },
   "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>tweet_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>text</th>\n",
       "      <th>rating_numerator</th>\n",
       "      <th>rating_denominator</th>\n",
       "      <th>name</th>\n",
       "      <th>dog_stage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>2017-08-01 16:23:56+00:00</td>\n",
       "      <td>This is Phineas. He's a mystical boy. Only eve...</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Phineas</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>2017-08-01 00:17:27+00:00</td>\n",
       "      <td>This is Tilly. She's just checking pup on you....</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Tilly</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>2017-07-31 00:18:03+00:00</td>\n",
       "      <td>This is Archie. He is a rare Norwegian Pouncin...</td>\n",
       "      <td>12.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Archie</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>2017-07-30 15:58:51+00:00</td>\n",
       "      <td>This is Darla. She commenced a snooze mid meal...</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Darla</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>2017-07-29 16:00:24+00:00</td>\n",
       "      <td>This is Franklin. He would like you to stop ca...</td>\n",
       "      <td>12.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Franklin</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2351</th>\n",
       "      <td>666049248165822465</td>\n",
       "      <td>2015-11-16 00:24:50+00:00</td>\n",
       "      <td>Here we have a 1949 1st generation vulpix. Enj...</td>\n",
       "      <td>5.0</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2352</th>\n",
       "      <td>666044226329800704</td>\n",
       "      <td>2015-11-16 00:04:52+00:00</td>\n",
       "      <td>This is a purebred Piers Morgan. Loves to Netf...</td>\n",
       "      <td>6.0</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2353</th>\n",
       "      <td>666033412701032449</td>\n",
       "      <td>2015-11-15 23:21:54+00:00</td>\n",
       "      <td>Here is a very happy pup. Big fan of well-main...</td>\n",
       "      <td>9.0</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2354</th>\n",
       "      <td>666029285002620928</td>\n",
       "      <td>2015-11-15 23:05:30+00:00</td>\n",
       "      <td>This is a western brown Mitsubishi terrier. Up...</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2355</th>\n",
       "      <td>666020888022790149</td>\n",
       "      <td>2015-11-15 22:32:08+00:00</td>\n",
       "      <td>Here we have a Japanese Irish Setter. Lost eye...</td>\n",
       "      <td>8.0</td>\n",
       "      <td>10</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2097 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                tweet_id                 timestamp  ...      name  dog_stage\n",
       "0     892420643555336193 2017-08-01 16:23:56+00:00  ...   Phineas        NaN\n",
       "1     892177421306343426 2017-08-01 00:17:27+00:00  ...     Tilly        NaN\n",
       "2     891815181378084864 2017-07-31 00:18:03+00:00  ...    Archie        NaN\n",
       "3     891689557279858688 2017-07-30 15:58:51+00:00  ...     Darla        NaN\n",
       "4     891327558926688256 2017-07-29 16:00:24+00:00  ...  Franklin        NaN\n",
       "...                  ...                       ...  ...       ...        ...\n",
       "2351  666049248165822465 2015-11-16 00:24:50+00:00  ...       NaN        NaN\n",
       "2352  666044226329800704 2015-11-16 00:04:52+00:00  ...       NaN        NaN\n",
       "2353  666033412701032449 2015-11-15 23:21:54+00:00  ...       NaN        NaN\n",
       "2354  666029285002620928 2015-11-15 23:05:30+00:00  ...       NaN        NaN\n",
       "2355  666020888022790149 2015-11-15 22:32:08+00:00  ...       NaN        NaN\n",
       "\n",
       "[2097 rows x 7 columns]"
      ]
     },
     "execution_count": 305,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show\n",
    "twitter_archive_df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "QkyPdzZ_FQJS"
   },
   "source": [
    "#### Define_13:\n",
    "Merge the three datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "XBSAhDryFQJv"
   },
   "source": [
    "##### Code:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 306,
   "metadata": {
    "collapsed": true,
    "id": "voqPrkRtFQJx"
   },
   "outputs": [],
   "source": [
    "# merge based on the tweet_id of twitter_archive_enahnced that matches with that of other dataframes\n",
    "df_new = pd.merge(twitter_archive_df, tweet_json_df, on='tweet_id')\n",
    "df_new = pd.merge(df_new, image_prediction_df, on = 'tweet_id')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "gPOUeo2lFQJy"
   },
   "source": [
    "##### Test:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 307,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "30ESOuqFFQJz",
    "outputId": "a55472b8-f36f-4d47-877f-f5103d554007"
   },
   "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>tweet_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>text</th>\n",
       "      <th>rating_numerator</th>\n",
       "      <th>rating_denominator</th>\n",
       "      <th>name</th>\n",
       "      <th>dog_stage</th>\n",
       "      <th>retweet_count</th>\n",
       "      <th>favorite_count</th>\n",
       "      <th>source</th>\n",
       "      <th>jpg_url</th>\n",
       "      <th>breed</th>\n",
       "      <th>confidence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>2017-08-01 16:23:56+00:00</td>\n",
       "      <td>This is Phineas. He's a mystical boy. Only eve...</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Phineas</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7315</td>\n",
       "      <td>34860</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "      <td>https://pbs.twimg.com/media/DGKD1-bXoAAIAUK.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>2017-08-01 00:17:27+00:00</td>\n",
       "      <td>This is Tilly. She's just checking pup on you....</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Tilly</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5465</td>\n",
       "      <td>30195</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "      <td>https://pbs.twimg.com/media/DGGmoV4XsAAUL6n.jpg</td>\n",
       "      <td>Chihuahua</td>\n",
       "      <td>0.323581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>891815181378084864</td>\n",
       "      <td>2017-07-31 00:18:03+00:00</td>\n",
       "      <td>This is Archie. He is a rare Norwegian Pouncin...</td>\n",
       "      <td>12.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Archie</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3605</td>\n",
       "      <td>22733</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "      <td>https://pbs.twimg.com/media/DGBdLU1WsAANxJ9.jpg</td>\n",
       "      <td>Chihuahua</td>\n",
       "      <td>0.716012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>891689557279858688</td>\n",
       "      <td>2017-07-30 15:58:51+00:00</td>\n",
       "      <td>This is Darla. She commenced a snooze mid meal...</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Darla</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7505</td>\n",
       "      <td>38132</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "      <td>https://pbs.twimg.com/media/DF_q7IAWsAEuuN8.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>891327558926688256</td>\n",
       "      <td>2017-07-29 16:00:24+00:00</td>\n",
       "      <td>This is Franklin. He would like you to stop ca...</td>\n",
       "      <td>12.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Franklin</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8057</td>\n",
       "      <td>36394</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "      <td>https://pbs.twimg.com/media/DF6hr6BUMAAzZgT.jpg</td>\n",
       "      <td>basset</td>\n",
       "      <td>0.555712</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             tweet_id                 timestamp  ...      breed  confidence\n",
       "0  892420643555336193 2017-08-01 16:23:56+00:00  ...        NaN         NaN\n",
       "1  892177421306343426 2017-08-01 00:17:27+00:00  ...  Chihuahua    0.323581\n",
       "2  891815181378084864 2017-07-31 00:18:03+00:00  ...  Chihuahua    0.716012\n",
       "3  891689557279858688 2017-07-30 15:58:51+00:00  ...        NaN         NaN\n",
       "4  891327558926688256 2017-07-29 16:00:24+00:00  ...     basset    0.555712\n",
       "\n",
       "[5 rows x 13 columns]"
      ]
     },
     "execution_count": 307,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_new.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "55mgXKW2I6wL"
   },
   "source": [
    "## Storing Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 308,
   "metadata": {
    "collapsed": true,
    "id": "jqnpu27JI52B"
   },
   "outputs": [],
   "source": [
    "# Store the clean DataFrame in a CSV file with the name twitter_archive_master.csv\n",
    "df_new.to_csv('twitter_archive_master.csv',index=False, encoding = 'utf-8')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "HEanAmX3KgEu"
   },
   "source": [
    "## Analyzing, and Visualizing Data\n",
    "\n",
    "   1. Visualizing the most popular dog breed\n",
    "   2. Analyse and visualize the relationship between retweet counts, and favorite counts.\n",
    "   3. Analyse and Visualize the most popular source for tweets\n",
    "   4. Analyse and Visualize the most most common dog stage."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 309,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "Q__ScPQVYhJu",
    "outputId": "4c59b0d1-acc6-4649-e198-139962b7607b"
   },
   "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>tweet_id</th>\n",
       "      <th>timestamp</th>\n",
       "      <th>text</th>\n",
       "      <th>rating_numerator</th>\n",
       "      <th>rating_denominator</th>\n",
       "      <th>name</th>\n",
       "      <th>dog_stage</th>\n",
       "      <th>retweet_count</th>\n",
       "      <th>favorite_count</th>\n",
       "      <th>source</th>\n",
       "      <th>jpg_url</th>\n",
       "      <th>breed</th>\n",
       "      <th>confidence</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>892420643555336193</td>\n",
       "      <td>2017-08-01 16:23:56+00:00</td>\n",
       "      <td>This is Phineas. He's a mystical boy. Only eve...</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Phineas</td>\n",
       "      <td>NaN</td>\n",
       "      <td>7315</td>\n",
       "      <td>34860</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "      <td>https://pbs.twimg.com/media/DGKD1-bXoAAIAUK.jpg</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>892177421306343426</td>\n",
       "      <td>2017-08-01 00:17:27+00:00</td>\n",
       "      <td>This is Tilly. She's just checking pup on you....</td>\n",
       "      <td>13.0</td>\n",
       "      <td>10</td>\n",
       "      <td>Tilly</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5465</td>\n",
       "      <td>30195</td>\n",
       "      <td>Twitter for iPhone</td>\n",
       "      <td>https://pbs.twimg.com/media/DGGmoV4XsAAUL6n.jpg</td>\n",
       "      <td>Chihuahua</td>\n",
       "      <td>0.323581</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             tweet_id                 timestamp  ...      breed  confidence\n",
       "0  892420643555336193 2017-08-01 16:23:56+00:00  ...        NaN         NaN\n",
       "1  892177421306343426 2017-08-01 00:17:27+00:00  ...  Chihuahua    0.323581\n",
       "\n",
       "[2 rows x 13 columns]"
      ]
     },
     "execution_count": 309,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# show the first 2 rows to easily have a view of the columns and their elements\n",
    "df_new.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Z6gY44q6c9Cr"
   },
   "source": [
    "#### Analysis/Visualization:\n",
    "Analysing and Visualizing the most popular dog breed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 310,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "hgPPrCcgMUEV",
    "outputId": "f7ec1e51-3c07-404c-8f04-c34791d807a8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "golden_retriever      137\n",
      "Labrador_retriever     92\n",
      "Pembroke               88\n",
      "Chihuahua              78\n",
      "pug                    54\n",
      "Name: breed, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA/AAAAIFCAYAAAB1dxYQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOzdeZxkVX3//9dbBtwVlREVRFBxQdGE4EL4CiguuESI288lKooSo0ajJCBu4IK74hYXBAXigoq7oqIoLkRQIAoIUZEdQQbRiKgo+vn9cW5DUdMz0zN01e3b/Xo+HvXornNvVX3qds10v88959xUFZIkSZIkaWG7Xt8FSJIkSZKkNTPAS5IkSZI0AAZ4SZIkSZIGwAAvSZIkSdIAGOAlSZIkSRoAA7wkSZIkSQNggJckzVmS7ZIckeSCJH9K8tskP0jymiS37ammnZJUkgdP+HXuluQb3XuuJLutpp79k1xvrH3z7nHPmnCdx3avM3O7KMlXktxvkq87x9oOTXLOFF5np7Fj8IfuM3tUkmcl2WDSNayirvGfzR+S/G+SVyS5fh81rUpX3/591yFJujYDvCRpTpLsBRwHLAdeDjwYeCLwVWBP4IP9VTcVbwPuCDwB2A741ir22wnYj35/x55Cq3E74EXApsC3kty9x5r68ALaMXgosBfwC+A/ge8nWd5TTaM/m0cCR9A+L2/pqR5J0oAs67sASdLCl+SBwJuBd1TVi8Y2H5Xk9cDj1/Ac6wNXVVVNqMxJuzvw7ar6St+FzMHlVXV89/3xSY4Hzgb+hRZqBy/J9avqyjXsdsbIcQD4eJJDgG/QOpz+YWIFrtrlYzV9I8mdaR1D/7qqByUJsH5V/WnSBUqSFi7PwEuS5mIf4NLu60qq6oqqOnTm/shw8ecmeVOSXwBXAhsmWZ7k/Ul+muT3Sc5P8tEkm4w+ZzcMvZJsneSb3b4XJXn1+PD0zo2SvDvJpd3tw0k2XNMbS7J+ktcmOaebFnBOd3/9bvtOSQrYHHjqzPDnVTzX/rSzqQB/XsW+63Xv4aIkv0nyhSSbzvJceyb5UZI/du/nkCS3XNP7mU1VnQOsAO48l/fc7TP6M3xbkku6n8EXk2w+VutKw61HHr/76mpL8qokJ3dTEy7tpincf2yfmSHxj0nygSQrgF+u47H4HvA+4FFJ7jTyGjfrPj+/SHJlkp8keVEXnEdr2SbJd7rh7+cneWn3Hq5Lx9RvgfVHG7qfyYeTPDPJ/wJ/op2xJ8m9k3w+ya+7Oo5L8oDxJ02yY5Jjklye5IokX01yz7F91ut+9hd1P99jk9xjlue6S5LPdJ+DPyY5L8knk3gySJKmyAAvSVqt7g/0HYGvrcPZv5cBd6ENsf9H4I/ALbuv+wK7AP8BbAkcl+QGszzHZ4GvA7sBHwVeAbxylv3eARTwZOBVwGO7tjU5DHgJcDjwKOBQWkfFYd32k2nDnVcAR3HN8OfZHAwc0n3//1ax7760IP1M4IXd9g+P7pDkDbSh3l8HHk07RrsAX06y3hze07UkuTntuP+ma1rTex6vd0vgGcDzgL8Djh4N+9fRJsCBwK7A7sAlwLeTbD3Lvu8CAjy123ddHdV93R6g6xD6Eu09vpV2Zv4rtGkTB8w8KMlGwDG0Y/l02hnzh61tLUmWdbebJXkU8BTg47Ps+kDgxbTP8y7AKUm2Af67q+HZtM/5r4CvJ/m7kdd4ZFfr74B/ov27uCnwnSS3H3mN/YGXAh+h/Rs7Gvj8LLV8ifaz+pfuPb+E1inn35KSNE1V5c2bN2/evK3yBmxMC8avn2XbstHbSPvm3WNOBrKG518PuH23/z+OtO/ftb1kbP8PAJcDG3b3d+r2O2xsv3fTOgpW+frAPbvH7j/W/vKu/V4jbRcAh87heM3UvWysfeaYHDvW/u9d++1G9vsL8Mqx/bbv9tttDa9/LPDdkZ/LnWidIEULaHN6zyP1ng5cb5Y69hhpm+35Zh6/+0jbocA5a/gsLAN+QpuuMdM+8zP+zBw/szP7P3gV2+/abd+nu/+o8Vq79oNpIXWj7v7ruvubjuxzQ9pogJpDXcd2rzN++zxwg7F9zwF+D9xmrP0Y4Axgg7Hjdgbw2ZG2M4Fjxh57M9pImrd3929BC/jvG9tvn9GfKbBRd//Rczn+3rx58+Ztcjd7TSVJ6yTJbYA/j95mGU772apaaWhxkn/phof/DrgKOK/bdNdZXuoTY/ePAG5CC6KjvjR2/1Tg+rQOiFXZofv64bH2mfs7ruax6+qosfundl83674+hHZW8yMjZ2qXASfQOi52YM2255qfy5nA3wPPqarPsvbv+ciq+uvMnao6jtaZsapRCGslyYPTpkj8ivZZ+DNt1MZsn4XPzMdr0s7iQwul0I7JX2kjPEZ9GNiAa97r/YHjq+qCmR2q6g+s/NlbnR8B9+lu/492Rvu+wCfHh+t3r3Xx1UUnN6T9fD4J/HXksxHaaI0duv22pHXcjH+Gfg98j2s+A1sDN2b2f2OjfgWcBbwhybO755ck9cAAL0lak1/RzmRvNtZ+KdcEkQ+s4rEXjTck+VfgPbTA8RhaeJmZ8zzbEPrxuc4z9zcZa79s7P7MAmezPeeMmTnl43VePLZ9Pq2pzlt3X89krIOENgT6VnN4jZmQuC1tdMPGVfX+btvavufZ5pr/kpWP/1rrhoMfRTsLvAftc3AfWv2z/dxW+jyto5kh5DPPd0vgslp5isj4MbktbYj/uLWZj/+7qjqxux1XVe+jDcV/FG2Y/Kjx93tL2tn2V7DyZ+P5wC266QAzn6FDZtnvUVzzGZq59OOq/o0B3dCC1rF0IvB64KdJzkryL2vxviVJ88CFRyRJq1VVVyX5NvCQJBvMhJyquor2Bz3dPN5ZHz5L2xNpQ3v3mmlIssVqStiYdvZv9D7AhXN8C6szE6ZvA/x8pP02Y9un6Vfd14cCv17N9tX5XVWduIpta/ueZxvBsDHww5H7V9LOVI+aS0fDY2ln3R9TVX+eaUxyC66Zrz9qvq5g8Mju63e7r5cBtxz9fHfGj8lFXBOOR61ulMdc/Lj7ei/gyyPt4+/3N7SRAv9JW79gJVX11240A7T1C74+y24z73Gmg2DjkRpm7o8/71nA07pRAvemdRi8J8k5VfXl8f0lSZPhGXhJ0ly8iTYP9o3z8Fw3op0JHPWM1ez/hLH7T6SdsT11ln3X1rdHnnPUU7qvx67Dc86cUb/huhQEfI0W0jYbOVM7ejt7HZ93xtq+58dlZNX/JNvTriv/vZF9zmXlKQ2PZM1uRJvvf3VQTfIgVh7tMW+SbAc8hza9Y+ZYfov2N9H4pRCfQgu7M+/1eGC7jFw1oBvWPpf3ujr36r6uWN1OVXUF8B1agD55ts9Ht+tPaHPo77GKz9Ap3X6nAFcw+7+xVdVQVfVD2uJ6sPLPXZI0QZ6BlyStUVUdk+QltDmw96Kd/TubNsz5LrQ/+K9gbmdIvwLsk+SlwPeBBwGPW83+z+4C5A9oq18/i7a41v+t6/uZUVWnJfkYsH83R/i/afOdXwF8rKrWpZPg9O7rXkm+DPxlNWfDZ6vp50neCLw7yV1p4fKPtGHfDwEOrqpvrkNdM8+/tu/5psBnk7wfWE4bQv0zrn0G+Ajg5UleRgu5DwCeNIdyvgL8G3Bokg/RPkuvYH5GVwDcvVtnYRltuPhDaSvYn05bwX3Gl2ln49+XZDntbPQjaJ+111fVpd1+b6PNWf9qklfROmte3H2d6+iAm+aay+RtQAvvM+/503N4/ItpnTBfTbum/UW0zrVtgPWq6iVVVUmeB3wuyQa0Oe6X0s6s/z1wXlW9rap+k+RA4GVJLqetQH8f2nSGq3X/5t9BWyn/TNow/t1poye+Mcf3LUmaBwZ4SdKcVNWbkhxHu/TZ62hh7o+0s30fp61k/Zc5PNWrgQ2BF9E6AL5FC+ZnrWL/XWmXD3sF8H/Aa4HXrPs7Wcnu3Ws/k7YS+y9oIw1etY7P90XaHP/n0i53F65ZNG1OquqlSc6gXbbtebRweD5tBfKfrWNdo3Zn7u/59bTL3h1KW/Dsm8DzR4e8d/tsSBtW/RLavPan0hbeW6Wq+mqSF9BC6WOB04CndTXNh3d2X6+kTT34Ee14/tfoUPlu2PkjaZ/rfWjD/8/p6nr7yH6XJtm5e97Du+d8Hy1AP22ONd2La87o/5n2c/0s8Jqqmm3awLVU1clJ7gPs19Vxc9qZ+5O7Wmb2OyrJDrRLOR5MGxFyMa2DZfSSdfvTPp/Pov38TqBdRm90SP3FtIUmX0wbffFH2giYR1XVSXN835KkeZBZFgeWJKl3SfanhZT1u/n2mqIkm9NGWTy7qg7ut5qFK8l6tPB8aVXt3Hc9kqTFzTPwkiRJc5TkNbRh5OfSztQ/i3ZW/RF91iVJWhoM8JIkSXNXtKkRt+u+PwXYzZXYJUnT4BB6SZIkSZIGwMvISZIkSZI0AItuCP1GG21Um2++ed9lSJIkSZK01k466aRLq2r5bNsWXYDffPPNOfHEOV9uV5IkSZKkBSPJuava5hB6SZIkSZIGwAAvSZIkSdIAGOAlSZIkSRoAA7wkSZIkSQMw1QCf5INJLkly2izb9kpSSTbq7ifJO5OcmeSUJNtMs1ZJkiRJkhaSaZ+BPxTYZbwxye2BhwLnjTQ/HNiyu+0JvHcK9UmSJEmStCBNNcBX1beBy2bZdCCwN1AjbbsCh1dzPLBhkttOoUxJkiRJkhac3ufAJ9kVuLCqfjS2aRPg/JH7F3Rtsz3HnklOTHLiihUrJlSpJEmSJEn96TXAJ7kR8FLgldflearqoKratqq2Xb58+fwUJ0mSJEnSArKs59e/E7AF8KMkAJsCJye5L3AhcPuRfTft2iRJkiRJWnJ6PQNfVadW1a2ravOq2pw2TH6bqroY+DzwtG41+vsD/1dVF/VZryRJkiRJfZn2ZeQ+BnwPuGuSC5LssZrdjwLOAs4EPgA8dwolSpIkSZK0IE11CH1VPWkN2zcf+b6A5026JkmSJEmShqD3VeglSZIkSdKa9b2IXW/+7j8O77uEQTnpzU/ruwRJkiRJWtI8Ay9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAAzwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQABnhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAAzwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQABnhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAAzwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQABnhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAAzwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQABnhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGoCpBvgkH0xySZLTRtrenOR/k5yS5DNJNhzZtm+SM5P8JMnDplmrJEmSJEkLybTPwB8K7DLW9jXgnlV1L+CnwL4ASbYCngjco3vMe5KsN71SJUmSJElaOKYa4Kvq28BlY21HV9VV3d3jgU2773cFjqiqK6vqbOBM4L5TK1aSJEmSpAVkoc2Bfybw5e77TYDzR7Zd0LWtJMmeSU5McuKKFSsmXKIkSZIkSdO3YAJ8kpcBVwEfWdvHVtVBVbVtVW27fPny+S9OkiRJkqSeLeu7AIAkuwOPAnauquqaLwRuP7Lbpl2bJEmSJElLTu9n4JPsAuwNPLqqfj+y6fPAE5NcP8kWwJbA9/uoUZIkSZKkvk31DHySjwE7ARsluQDYj7bq/PWBryUBOL6qnlNVP07yCeB02tD651XVX6ZZryRJkiRJC8VUA3xVPWmW5kNWs/8BwAGTq0iSJEmSpGHofQi9JEmSJElaMwO8JEmSJEkDYICXJEmSJGkADPCSJEmSJA2AAV6SJEmSpAEwwEuSJEmSNAAGeEmSJEmSBsAAL0mSJEnSABjgJUmSJEkaAAO8JEmSJEkDYICXJEmSJGkADPCSJEmSJA2AAV6SJEmSpAEwwEuSJEmSNAAGeEmSJEmSBsAAL0mSJEnSABjgJUmSJEkaAAO8JEmSJEkDYICXJEmSJGkADPCSJEmSJA2AAV6SJEmSpAEwwEuSJEmSNAAGeEmSJEmSBsAAL0mSJEnSABjgJUmSJEkaAAO8JEmSJEkDYICXJEmSJGkADPCSJEmSJA2AAV6SJEmSpAEwwEuSJEmSNAAGeEmSJEmSBsAAL0mSJEnSABjgJUmSJEkaAAO8JEmSJEkDYICXJEmSJGkADPCSJEmSJA2AAV6SJEmSpAGYaoBP8sEklyQ5baTtlkm+luRn3ddbdO1J8s4kZyY5Jck206xVkiRJkqSFZNpn4A8FdhlrewlwTFVtCRzT3Qd4OLBld9sTeO+UapQkSZIkacGZaoCvqm8Dl4017woc1n1/GLDbSPvh1RwPbJjkttOpVJIkSZKkhWUhzIHfuKou6r6/GNi4+34T4PyR/S7o2laSZM8kJyY5ccWKFZOrVJIkSZKkniyEAH+1qiqg1uFxB1XVtlW17fLlyydQmSRJkiRJ/VoIAf6XM0Pju6+XdO0XArcf2W/Trk2SJEmSpCVnIQT4zwNP775/OvC5kfandavR3x/4v5Gh9pIkSZIkLSnLpvliST4G7ARslOQCYD/gDcAnkuwBnAs8odv9KOARwJnA74FnTLNWSZIkSZIWkqkG+Kp60io27TzLvgU8b7IVSZIkSZI0DAthCL0kSZIkSVoDA7wkSZIkSQNggJckSZIkaQAM8JIkSZIkDYABXpIkSZKkATDAS5IkSZI0AAZ4SZIkSZIGwAAvSZIkSdIAGOAlSZIkSRoAA7wkSZIkSQNggJckSZIkaQAM8JIkSZIkDcCyvgvQ0nPeq7fuu4TB2OyVp/ZdgiRJkqQFwjPwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQABnhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAJb1XYCk6dj+Xdv3XcKgHPevx/VdgiRJknQtnoGXJEmSJGkADPCSJEmSJA2AAV6SJEmSpAEwwEuSJEmSNAAGeEmSJEmSBsAAL0mSJEnSABjgJUmSJEkaAAO8JEmSJEkDYICXJEmSJGkADPCSJEmSJA2AAV6SJEmSpAEwwEuSJEmSNAAGeEmSJEmSBsAAL0mSJEnSAMw5wCfZIclNVrHtJkl2uC6FJHlRkh8nOS3Jx5LcIMkWSU5IcmaSjyfZ4Lq8hiRJkiRJQ7U2Z+C/CWy1im137bavkySbAC8Atq2qewLrAU8E3ggcWFV3Bn4N7LGuryFJkiRJ0pCtTYDParZdH/jLdaxlGXDDJMuAGwEXAQ8Cjuy2Hwbsdh1fQ5IkSZKkQVq2uo1JNgfuONK07SzD6G8IPBM4b12LqKoLk7yle44/AEcDJwG/qaqrut0uADZZRZ17AnsCbLbZZutahiRJkiRJC9ZqAzzwdGA/oLrbu7j2mfjq7l8FPG9di0hyC2BXYAvgN8AngV3m+viqOgg4CGDbbbetda1DkiRJkqSFak0B/lDgWFpI/wYtpJ8+ts+VwE+r6rLrUMeDgbOragVAkk8D2wMbJlnWnYXfFLjwOryGJEmSJEmDtdoAX1XnAucCJHkgcHJVXT6BOs4D7p/kRrQh9DsDJ9IWxnsccARtNMDnJvDakiRJkiQteHNexK6qvjWh8E5VnUBbrO5k4NSuroOAfYAXJzkTuBVwyCReX5IkSZKkhW5NQ+iv1l2DfV/gScBmtJXnR1VVzfn5xlXVfrT59qPOAu67rs8pSQvBt3bYse8SBmXHb3+r7xIkSZIWpLUJ3G+mzYH/MvBp2tx3SZIkSZI0BWsT4B8H7FdVB0yqGEmSJEmSNLs5z4EHbgJ8b1KFSJIkSZKkVVubAP8FYIdJFSJJkiRJklZtbYbQvws4PMlfgaOAla77XlVnzVdhkiRJkiTpGmsT4GeGz+/PyqvFz1jvOlUjSZIkSZJmtTYB/plATaoQSZIkSZK0anMO8FV16ATrkCRJkiRJq7E2i9hJkiRJkqSezPkMfJIPrmGXqqo9rmM9kiRJkiRpFmszB/5BrDwH/pbATYHfdDdJkiRJkjQBazMHfvPZ2pPsALwPeMo81SRJkiRJksZc5znwVfVt4EDadeIlSZIkSdIEzNcidmcBfztPzyVJkiRJksZc5wCfZBmwO3DBda5GkiRJkiTNam1Wof/GLM0bAHcBbgU8Z76KkiRJkiRJ17Y2q9Bfj5VXob8c+DRwRFUdO19FSZIkSZKka1ubVeh3mmAdkiRJkiRpNeZrETtJkiRJkjRBaxXgk2yd5MgkK5Jc1X39RJKtJ1WgJEmSJElau0Xs7gN8C/gD8HngYuA2wD8Aj0yyQ1WdNJEqJUmSJEla4tZmEbvXA6cBO1fV5TONSW4KfL3b/tD5LU+SJEmSJMHaDaG/P/D60fAO0N1/I7DdfBYmSZIkSZKusTYBfvwScmu7XZIkSZIkraO1CfAnAC/thsxfLcmNgX2A4+ezMEmSJEmSdI21mQP/UuBY4NwkXwQuoi1i9wjgxsCO816dJEmSJEkC1iLAV9X3k9wfeCXwMOCWwGXAN4HXVNWpkylRkiRJkiStNsAnuR7wSODsqjqtqk4BHje2z9bA5oABXpIkSZKkCVnTHPh/Aj4GXLGafS4HPpbkSfNWlSRJkiRJupa5BPgPVdXZq9qhqs4BDgGePo91SZIkSZKkEWsK8NsAR8/heb4ObHvdy5EkSZIkSbNZU4C/KfDrOTzPr7t9JUmSJEnSBKwpwF8K3GEOz7NZt68kSZIkSZqANQX47zK3ue27d/tKkiRJkqQJWFOAfzuwc5IDk2wwvjHJ+kneDjwIOHASBUqSJEmSpDVcB76qvpdkL+CtwFOSHA2c222+A/AQ4FbAXlV1/EQrlSRJkiRpCVttgAeoqrcnORnYB/hH4Ibdpj8AxwJvqKrvXNdCkmwIHAzcEyjgmcBPgI8DmwPnAE+oqrksqidJkiRJ0qKypiH0AFTVt6vqkbSV5m/T3W5WVY+cj/DeeQfwlaq6G3Bv4AzgJcAxVbUlcEx3X5IkSZKkJWdOAX5GVf21qi7pbn+ZryKS3BzYATike50/VdVvgF2Bw7rdDgN2m6/XlCRJkiRpSNYqwE/QFsAK4ENJ/ifJwUluDGxcVRd1+1wMbDzbg5PsmeTEJCeuWLFiSiVLkiRJkjQ9CyXALwO2Ad5bVX8LXMHYcPmqKtrc+JVU1UFVtW1Vbbt8+fKJFytJkiRJ0rQtlAB/AXBBVZ3Q3T+SFuh/meS2AN3XS3qqT5IkSZKkXi2IAF9VFwPnJ7lr17QzcDrweeDpXdvTgc/1UJ4kSZIkSb1b42XkpuhfgY8k2QA4C3gGrYPhE0n2oF1//gk91idJkiRJUm8WTICvqh8C286yaedp1yJJkiRJ0kKzIIbQS5IkSZKk1TPAS5IkSZI0AAZ4SZIkSZIGwAAvSZIkSdIAGOAlSZIkSRoAA7wkSZIkSQNggJckSZIkaQAM8JIkSZIkDcCyvguQJGlS3r3XF/ouYVCe/9Z/6LsESZK0Gp6BlyRJkiRpAAzwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQABnhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAAzwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQABnhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAAzwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQABnhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAAzwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQAy/ouYFSS9YATgQur6lFJtgCOAG4FnAQ8tar+1GeNkiRpzQ74p8f1XcKgvOzDR/ZdgiRpABbaGfgXAmeM3H8jcGBV3Rn4NbBHL1VJkiRJktSzBRPgk2wKPBI4uLsf4EHATJf0YcBu/VQnSZIkSVK/FkyAB94O7A38tbt/K+A3VXVVd/8CYJPZHphkzyQnJjlxxYoVk69UkiRJkqQpWxABPsmjgEuq6qR1eXxVHVRV21bVtsuXL5/n6iRJkiRJ6t9CWcRue+DRSR4B3AC4GfAOYMMky7qz8JsCF/ZYoyRJ0oJ3xgHf6LuEQbn7yx7UdwmSNGcL4gx8Ve1bVZtW1ebAE4FvVNVTgG8CM8vYPh34XE8lSpIkSZLUq4VyBn5V9gGOSPJa4H+AQ3quR5IkSZrV/vvv33cJg+LxktbeggvwVXUscGz3/VnAffusR5IkSZKkhWBBDKGXJEmSJEmrZ4CXJEmSJGkADPCSJEmSJA3AgpsDL0mSJElr4xOfdNmsuXrC47/fdwm6DjwDL0mSJEnSABjgJUmSJEkaAAO8JEmSJEkDYICXJEmSJGkADPCSJEmSJA2AAV6SJEmSpAEwwEuSJEmSNAAGeEmSJEmSBsAAL0mSJEnSABjgJUmSJEkaAAO8JEmSJEkDYICXJEmSJGkADPCSJEmSJA2AAV6SJEmSpAEwwEuSJEmSNAAGeEmSJEmSBsAAL0mSJEnSACzruwBJkiRJ0vDc+8iv9l3CoPzocQ+7zs/hGXhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAAzwkiRJkiQNgAFekiRJkqQBMMBLkiRJkjQABnhJkiRJkgbAAC9JkiRJ0gAY4CVJkiRJGgADvCRJkiRJA2CAlyRJkiRpAAzwkiRJkiQNwIII8Elun+SbSU5P8uMkL+zab5nka0l+1n29Rd+1SpIkSZLUhwUR4IGrgL2qaivg/sDzkmwFvAQ4pqq2BI7p7kuSJEmStOQsiABfVRdV1cnd95cDZwCbALsCh3W7HQbs1k+FkiRJkiT1a0EE+FFJNgf+FjgB2LiqLuo2XQxsvIrH7JnkxCQnrlixYip1SpIkSZI0TQsqwCe5CfAp4N+q6rej26qqgJrtcVV1UFVtW1XbLl++fB5fVuYAACAASURBVAqVSpIkSZI0XQsmwCdZnxbeP1JVn+6af5nktt322wKX9FWfJEmSJEl9WhABPkmAQ4AzquptI5s+Dzy9+/7pwOemXZskSZIkSQvBsr4L6GwPPBU4NckPu7aXAm8APpFkD+Bc4Ak91SdJkiRJUq8WRICvqu8CWcXmnadZiyRJkiRJC9GCGEIvSZIkSZJWzwAvSZIkSdIAGOAlSZIkSRoAA7wkSZIkSQNggJckSZIkaQAM8JIkSZIkDYABXpIkSZKkATDAS5IkSZI0AAZ4SZIkSZIGwAAvSZIkSdIAGOAlSZIkSRoAA7wkSZIkSQNggJckSZIkaQAM8JIkSZIkDYABXpIkSZKkATDAS5IkSZI0AAZ4SZIkSZIGwAAvSZIkSdIAGOAlSZIkSRoAA7wkSZIkSQNggJckSZIkaQAM8JIkSZIkDYABXpIkSZKkATDAS5IkSZI0AAZ4SZIkSZIGwAAvSZIkSdIAGOAlSZIkSRoAA7wkSZIkSQNggJckSZIkaQAM8JIkSZIkDYABXpIkSZKkATDAS5IkSZI0AAZ4SZIkSZIGwAAvSZIkSdIAGOAlSZIkSRoAA7wkSZIkSQMwiACfZJckP0lyZpKX9F2PJEmSJEnTtuADfJL1gP8EHg5sBTwpyVb9ViVJkiRJ0nQt+AAP3Bc4s6rOqqo/AUcAu/ZckyRJkiRJU5Wq6ruG1UryOGCXqnpWd/+pwP2q6vkj++wJ7NndvSvwk6kXOn82Ai7tu4glyOPeH499fzz2/fHY98dj3x+PfX889v3x2Pdj6Mf9DlW1fLYNy6ZdySRU1UHAQX3XMR+SnFhV2/Zdx1Ljce+Px74/Hvv+eOz747Hvj8e+Px77/njs+7GYj/sQhtBfCNx+5P6mXZskSZIkSUvGEAL8D4Atk2yRZAPgicDne65JkiRJkqSpWvBD6KvqqiTPB74KrAd8sKp+3HNZk7QopgIMkMe9Px77/njs++Ox74/Hvj8e+/547Pvjse/Hoj3uC34RO0mSJEmSNIwh9JIkSZIkLXkGeEmSJEmSBsAAL0mSJEnSABjgJUmSJEkaAAN8T5IsS/LcJLfruxZJkiRJ0sLnKvQ9SnIFsFVVndt3LUtNkmXAnsBnq+oXfdcjTUv32b8vsBmwwei2qjq8l6KkCUpyNHBsd/t+VV3Va0GLXJLN5rpvVZ03yVqkPiR5OPA84I7Aw6rq/CTPAs6uqmP6rU6LwYK/DvwidzywDWCAn7KquirJm4Ev9V3LUpVkY+CpwJ2AV1TVpUm2B35RVWf3W93ilORuwBeALYAAf6H9HvgzcCVggJ+QJK9cxaYC/gicCXylqv4wvaqWjO8DDwdeCfw5yfcw0E/SObTP9VysN8E6lrwkNwL+Brg1Y6Nuq+rTvRS1yCV5CvA+4GBgZ2D9btN6wN6AAX7ClkIHigG+Xx8A3prkDsBJwBWjG6vq5F6qWjrsQOlJkr+j/RI7G7gH8GbgUuAhwF2AJ/dX3aL2dtr/NX8DXNx9vTnwXuDlPda1FDyeNurhxsDMqJ/b0f7fXwHcHrgkyY5VdVY/JS5OVfVygCQ3BP4e2IkW6PendZ7crK/aFqn7jHx/F+BNtEDzva5tO+CfgX2mXNeSkuTBwMeAW82yubDzZFL2Bp5dVUd0oXHG8cCre6ppyVgqHSgOoe9Rkr+uZnNVlf+5TlCSJwKvA96JHShTleSbwLerar8klwP3rqqzkmwHHFFVd+i5xEUpya+AHavqtCT/B9y3qn6SZEfgXVV1r55LXLSS7A78E7B7VV3QtW0KfBD4MG000CeAy6tqt77qXMy6UT87AQ8CHghsCpxQVQ/ss67FLMm3aP+3HDnW/jjghVX1gH4qW/yS/Bj4AfBSpwpOT5LfA3evqnPH/r65E3BaVd2w5xIXtSQ/Al7fdaCMHv97A0dX1cY9lzgvPAPfry36LmCJ+2j39W2zbLN3erL+DthjlvaLgEXxn+sCFeD33fcrgE2AnwAXAHfuq6glYj9g15nwDlBVFyTZm7YWx+FJXgZ8rrcKF6kk76EF9zsAJwDfAp4NHF9VV/ZY2lJwX+CUWdpPof0e0ORsDjza8D51v6CNPBkf3bkD8PPpl7PkbMk1o31G/Y5FNNrKAN8jF6/rnR0o/fkDcItZ2u8GXDLlWpaS04B7A2fR5gXvk+QvtDBzZp+FLQEbAzeYpf36tPmpAL8EbjS1ipaO59A6rN4AfBk4qRx+OC3nAM8F/m2s/bk4fW3SjgPuiqFx2g4C3jkyfP72SR5Am0qyf29VLR1LogPFAN+zpbDQwkJlB0qvPgfsl+Tx3f1KsjnwRuBTfRW1BBxAm4MNbc77l4Bv0tYfeEJfRS0RXwfen2RP2pQdaGcg3wt8rbu/NW1dCM2vLWln4HeidVbdNMl3aZ/9Y50uNVEvAj6TZBfaHGCA+9HODj+mr6KWiPcBb+kuV3wqbbHSq/m5n4yqelOSm9P+X78B7f+ZK4G3VNV/9lrc0rAkOlCcA9+jsYUWngPco5un8c/AY6rqYb0WuATYgdKPJDcDjgLuRQuUF9POUB4HPKKqrljNwzWPktwS+LVnJCcrya1pq/w/lLb6P7RVoY8Gnl5VlyR5ILB+VR3dU5lLQnc1hr1paxKs53ozk9Wt9fBc2ggrgDOA91XV+f1Vtfi5zlK/uisAbEX7f/70qvpdzyUtGUkOoHUezox6m+lAeUV/Vc0vA3yPlspCCwuVHSj9S/Ig2pUArgecXFVfT7J+Vf15DQ+VBinJXWnDWgH+t6p+2mc9S0GS6wHb0hau2wnYnvaH3Um0M/D79ledNBndFY5WyVGIWswWeweKAb5HrlTZLztQ+pPkNbP1hCbZADiyqh7dQ1mLXpLPr267x12LUZLf0tYaOJlrrv/+XUf6TEeSrWmXjbsjsEdVXZRkN+DcqvqffquTrrs1/W4d5e/ZyUryQdoVLi4fa78x7YoYz+ynsvnlHPh+LYmFFhawJbFS5QK1R5IVVfXOmYYk6wOfpl3aSZPxq7H769MWtbs97dhrQpK8c3Xbq+oF06plCXo8BvZeJHko8Hna4oE7AzMnJu4E7A54ycQJSrKMdiWAzYANRrdV1eG9FLU4jf9uVX+eDrwEuHys/YbA0wADvK6zJbHQwgJmB0p/Hg58M8mvquoj3Zn3z9DC+4P6LW3xqqpnzNae5K3Ab6dczlKz9dj99WlzgtcDPAs5QVX1VYAkN6BdLrGAn1fVH3stbGl4DfDiqnpPN9JtxrHAXv2UtDR0az18gXbFndDW3lhGW8zuStqaHJoHq/rdqunp1vNJd7tFkqtGNq8HPJJ2pZdFwQDfI1eq7J0dKD2pqh8l2RX4YpI/As+gXZP8QVVlT/b0vR/4LvCqvgtZrKrqgeNtXaA8BPjO9CtaOrqzkK8Hnk87CxngyiTvAl7mmhsTdU/agqXjLgNuOeValpq309Z5+BvaQrF/A9ycduWLl/dYlzQJl9I6Zws4fZbtBew31YomyADfs6p6Wbda4qJdaGGhsgOlX1X1nSRPpp15/zEtvF/Wc1lL1V3XvIvmW1X9McnrgK/QFtTUZLwJeBJtsdLvdm0PoIX66wH/3lNdS8FltM7Zc8batwEumHo1S8t9gB2r6opuRfplVXVykr2Bd9GuAqN5kOSbtIC4RlXlKMPJeCCtc/YbwGNp//fM+BNtzY1f9FHYJBjge5Tk34CPVtUlwIl917MU2YEyPatZ5OVS4Arg0CSAi7xMyizzsAPcljal4YPTr0jARsBN+i5ikXsy8MyqGj0T/PMkK2hXITHAT85HgTcneQIt4CxLsiPwFuBDvVa2+AX4fff9ClpHyk9oHSd37quoReq0ke/XA55CG/VwQtd2X9rv2g9Pua4lo6q+BZBkC+C8xX5pXAN8v14MvCnJN4D/Aj5TVb9fw2M0T+xAmbpVDY3/6lSrWNrG52H/lfaH3YswwE9UkhePN9H+oHsKsw8x1vy5ObOva/JzYMMp17LUvBw4lLbWTGhDW0ML9gf0V9aScBptkdKzgO8D+yT5C/Bs4Mw+C1tsqupfZ75PciBwGG0l9Bppfzvts6/JuhVwq5kTQuOq6uTpljMZXkauR2mfrp1oZwceS1vU6HO0Hrqjq+qv/VW3+CU5D7gNbbiNHSiSJibJ2WNNM50n36BdznJ8xVzNkyTHAydV1fPG2t8L/E1VbddPZUtHd3ncv6WNdPufqvpZzyUtekkeBty4qj6d5I7Al2jTpS4FnlBVx/ZZ32KV5FfAdlX107H2uwDHV5VrP0xQN12kuHZnydVht6rWm3pRE2CAXyC6VbgfRQvzjwB+U1W367eqxc0OlP51f1RsRfvP9YyqOqvnkqSJS3ITAKfrTEeSHWijHC4Eju+a7w/cDnh4VX13VY/VddNd7/0LVfWXvmvR1St1/3qxDy/uU5LLgGdX1afG2h8LfMAAP1lJ7jDWtD6t8/BlwL5V9eXpVzX/DPALSJLb0sLks4C7VtX1ei5pybADZbqS3Iy2+vZjaWciofWWfgrYw7ORk9Gtev5C2vWYb007G3a1qnJRownqpu28mDYXFdqlLN8GvN0/qCcrye2A59Eu3QdwBvCexbSo0UKU5Ara9ZgPAw4ZPyspLTZJ3gLsAbyRa3cY7g18qKq8fGIPkjwU2K+qtu+7lvlggO9ZkpsCj6PNg9yRNi/po8CHq2p8yKUmyA6U6UnyIeDvgT2B/+6at6etxH1cVe3RV22LWZIPAv8IfJIWHq/1C6CqvIzchCR5E+3z/mbge13zdrQF1D5QVXv3VZs0Kd3fOE+mXSr0PrTP/iHAJ6rqij5rW4y6xWL/qap+u5qFYwEXi52UJDNXtnghbZ0TgIuAdwBvdTRKP5JsCfywqm7cdy3zwQDfoyRH0s72/hb4OC20/6DfqpYWO1D60c0R262qvjPWvgNtLYJb9VPZ4tYN7XtCVX2971qWmu7Y71lVR461Pw54v5/5yUpyI9p1sGcbefLpXopaYpLcA3gm7fftjWh/9xxSVcev9oGas65z/AVVdXn3/SpV1TOmVNaS1Y02pKp+23ctS0U3TeRaTbSOlP2BO1bVNlMvagJchb5fV9KGEB9tj9z0zdKBsq8dKFNzQ2Zflf4y4AZTrmUp+T1wft9FLGGnrKLN0T4TlOTBwMdoqxOPK9plnzRhVfXjboXuK2jDif8/YPckJ9PmDM/270NrYTSUG9D7NbrGT5LTPSk0NZcyNrqQFuLPp/2fsyh4Bl5LVpKPcM2CdXagTFGSr9E6Tp46s/J/khsDhwM3q6qH9FnfYpXkBcA9gOc453q6Zi4hVFUvHGs/EFivql7QT2WLX5IfAz8AXuqc9+lLsj5t6s4zaetvnAAcTOs4vwXwOuB+VXX33oqU5olr/PQryY5jTTNXfDmzqq7qoaSJMMD3LMlzaQvrbAHcs6rOSvIS4Kyq+kS/1UmTkWRr4Cu0YZQzZ122pp0hflhV/biv2habWeZB7gD8H+16zH8e3eCcyMnpLln2ZNpcyJkhw/ejrYT+EeDqPywM8/OrW0jtXlU127XgNUFJ3gU8iXZG7L+Ag6vq9LF9bgP8wnVn5peLlvbDNX40DQ6h71G3IvHetJUq3zCy6ULg+YABfsLsQOlHVZ3aLSjyFK5ZFfq/gI9U1R/6q2xRGp+q8JleqtDdgJO772cuc3Nxdxs982iv+vw7jnb9awP89G1F+3vm01X1p1XscynwwOmVtGS8h2sWLf1v/L9lWh7Nymv8HJtkT9rvXwP8hHVXlrons3dcHdVLUfPMM/A9SvK/wF5V9aUklwP37gLkPYBvu6jRZM3SgXKP7vg/lTYfb4deC1ykuuGU5wM7e6Zd0iQkGV2oaHPgtbRL9p3KyiNPTkZaZFy0tB9Jfg9sO8tIk3sCJyyWVdAXqiQPoZ0QuvUsm6uqFsWaJ56B79cdgNNmaf8zbZEvTdZzaEH9S0leO9J+Mm2esCagqv6c5M94NkDS5JxI+z8mI20HzbKfi9hNWJJ70S6rtRXteJ8OvLmqZvv7R/PHRUv7cRzwmiTja/y8imuG1Gty/hP4IvAa4Jcs0r81DfD9OgvYBjh3rP0RtF9wmiw7UPrzLmDfJM9YTIuKDEGSZ9DmpG4GbDC6raru2EtR0vzbou8CBEkeDXwa+A7w5a75/wH/k+QxVfWF3opb/N4EvDiJi5ZO14uArwIXJllpjZ/eqlo6bgu8rqrGs9WiYoDv11uAd3fXpw2wXTd8e2/aaq2aLDtQ+vMAYEfaL7jTaJcVupqLqU1Gkv8A9gXeT1vM7j3Anbvv39JjadK8Gv3jLckBwPlV9b7RfZI8B9gEeMWUy1tKXgscUFX7jTYmeXW3zQA/OQ+h/a7dJYmLlk5JVZ3mGj+9+iJtEcGz+i5kkpwD37MkzwZeDty+a/oFsF9VHdJfVUtDdybytbQOk/cD/0wLM3sDz6yqj/dY3qLWrdK6Sl6/djKS/JR2Ka0jx9bdeAWwWVU9u+cSpXmX5Dzg8VV1wlj7fYAjq+oOsz9S11WSP9IWiD1zrH1L4NSqukE/lS1+/p7VUpTk5rQru/yMNsp2vOPq8D7qmm8G+AUiyUbA9arqkr5rWUrsQNFS0i2uc7eqOi/JJcBDq+qHSe4MfL+qbtlzidK860LkVlV11lj7HYHTDZGT03We/Md4h3iSJwJvtPNEi5HrPvQnyROAw4Dr06YtjAbdqqqb9VLYPHMI/QJRVZf2XcNSVFUfAD5gB0o/ktyJay6hdfr4H9iadxcDGwHn0aaObAf8kDbyxN5cLVbn0YYSj///sgNwwfTLWVI+ALy/6yQcvSb2vwNv7q0qaUJc96F3bwHeDexfVVesaeehMsBPWbegxY5V9eskp7KaP5qr6l7Tq2xpswNlupLcCjiEdr3Uv17TnC/Spi+MX7tc8+MbtGN+Mu34H9j1Vm8DfKLPwqQJej/ts74B7d8AwM7A62mXEdXkvBb4HbAXbVVo6Ea6Ae/sq6ilwkVLe+G6D/3aEHjfYg7vYIDvw6eAK7vvj+yzkKXIDpQF42DaWd8HADPzUu8HvJd2xuYxPdW12O0JXA+gqt6X5Ne0s2GfooUcadGpqrd2o6zeyTUh5k/AO6rqTf1Vtvh1q58fSOtAuWnXdnm/VS0NLlram7vQFq0b91+0NZY0WZ8CHgz8vO9CJsk58D1Jcj3a6pTnVdXv+q5nqUiyH20e0u+771epql41pbKWnG4u9s5V9b2x9u2Ar1fVjfupTNJi1V2Leavu7hn+7tVi5qKl/XDdh351n+8X0i7ldworL2L3tj7qmm8G+J4kCe1M/Fbjq7Nq8uxA6VeSc4F/qKpTxtrvDXyhqjbrp7LFLck2q9teVSdPqxZJi1+SWwD7Aw8Ebk03AmhGVd26h7KWBBct7UcXIPeirfGw0roPVXVAX7UtBUnOXs3mWixTRxxC35OqqiQ/AZYDBvjpK9riXVvh8e/Dq4G3J3lqVV0IkGQT4K3dNk3GibTPfkbaRntx15tuOZIWucOBe9BWhf4lLpY5TS5a2g/XfehRVW3Rdw3T4Bn4HiV5OO0SZs8DflT+MKaqmwO/5/gwbk3GLGsObAHcALiwu78J8EfgbNcfmIwk40P31gf+FngZsG9VfXnlR0nSuumGbu/o6J7pS3IwcMH/3969x1s+13scf71n3HMrIk6uIddBJCV04VBKkkvkKNJJpxMl5ERFhVwriTpyi0JFRSR0QSK5328zTISD3HJn5n3++P62WdasGWPstb57r/V+Ph77sdb6/n5rz3vvYc/+/H7f7+dre19JO1N6EVxG07Q0U+i7L30foltSwFfU/MM2B2VK2QtMaW4HQL/sVThS5QJKb71cz4FW6T/QW5L+Hfia7XVqZ4mI/iHpGmAn21fUzjJomqWCY2y/0LzemjKV+zbgh7afn97749XJNrn1SFoO2ILOuy/sWCXUMEsBX5GkTzD9Lugn9i7N4MkFlIhC0rLANWkeGBHDSdL6lAvluwM32J5UOdJAkDQrsD/wfdsTa+cZJNPaJhfINrk9IGkTSif6q4E1gL8BbwJmBy62vWnFeMMmBXwMrFxAGRkkDV1EeZHtpyrF6WuS2psWCViE0mRqadvTbXIXEfFKNL1NTqOsv56K7fTd6BJJTwAr276rdpZBIumXwLLAp5l6m9w7bGeb3C6SdCXwC9sHDu2+QOlBcBJwabrQx6smaRKwiO0H2sYXAB7IP2zRr5q12EdQOhNPddc3/+13h6TJTH3RSsDdwNa2L+t9qojoV5IuAl4L/IAOTexsn14j1yCQdDpwtu3jamcZJNkmt67mwtW4ZsvEh4H1bN8gaRXK/w99sctRutDXpWmMzw4818sggygXUKo6mbJ84XOkM3Evvbvt9WTgQcpdgRcq5ImI/rYmsJbtG2oHGUC/Bw6QNA64Eniy9aDtM6qk6n8P0va9bjwFZPp89w0tjwW4j7Lrwg2Umve1tUINtxTwFUjarXlqYOfmatGQscC6wC09DzZ4cgGlntWBt9q+uXaQQWL7wtoZImKg3ASkn0wdRzaPu3Q4ZrJtaLdkm9y6/gq8k/Kz52zgMEmrAh8G+mbXqRTwdXyueRSwE9Da1OU54C5g5x5nGhi5gDIiXAu8HkgB32OS5gJWAxZi6t4DuSMTEcNpH+BwSfsA1wMv6Xxu++EqqQaA7TEvf1Z0weeBJYG7JLVvk7uQpBcvqGTL3K7YDZi7eb4vMA/wEcruC7tN4z2jTtbAVyTpj8Dmth+pnWWQSLqzeboEcA+dL6B81fZfia6QtBJlDfwRlKlN7b/U/b1Grn4naQPgFGCBDoedZSMRMZyavhtDWn/hFPmZE30oW+ZGL6SAHyEkLQw8aHvyy54cwyIXUOppmomcAqxIfqnrGUk3UrZU+bLte2vniYj+1mwjN01Z1tM9kgR8BvgssBSlI/0ESXsBE2z/rGrAASdpG+BM253Wy0dMVwr4iiTNAhxA+QE7J7Bc88P1IGCi7aOqBhwguYDSW5KuAh6lrAnr1Jn4yhq5+p2kJyndWcfXzhIREd0j6fPAnsBBwLeAlZrfMf8D+JTt9aoGHHCSHgdWsz2hdpbRrtkuboYKWtt90ZMja+Dr2hf4ILAd8NOW8cuBLwEp4Luo0wUUIBdQemN5yj9ct9UOMmAuAd4MpICPiJ6RtCiwODBb67jti+okGgg7Uwr1syV9s2X8KmClSpliimk1Uo5X7r9rB+i1FPB1bQPsaPvCtnViN1CKyeiufckFlFoup0zpSwHfZZLe0vLyB8ChzS/TnRpKXdXLbBHR35qfNT8F1qPcIRMvvVOW5VLdswTl98l2z1NuWkT0Bdsn1s7Qayng61oUmNhhfBbyd9MLuYBSz9GUbVYOI4Vkt13BlF+ch/xvh/OyrVBEDLfvUBrFrkjpv7ExsDBlO60vVMw1CCYAb2Hq3zPfT9liKyJGqRSJdd1IuSp9V9v4VkDWAHdfLqDUc0rzmEKy+5aqHSAiBtb6wCa2b5FkSq+ZSyQ9C3wDOL9uvL52KHBks3WogLc369/3BHasmiyiSyTNBuxNuUm3ODBr6/F+aZKcIqWu/YCTJS1GKVi2lLQ8sC2wSdVkgyEXUOpJUdkjtjtdpIqI6IU5gYea5w8DC1GWTt0EZA/sLrJ9fEuvn7mAk4B7gV1sn1Y1XET3fAPYGjgQ+DawB7Ak8FHgK/ViDa8U8BXZPkvSVsCXgcnA1yjNRT5o+4Kq4QZDLqBUkqKydyRtDpxl+/nm+TTZPqNHsSJiMNxCaVp6F3ANsLOkuylbm/2jYq6BYPsY4BhJCwJjbD9QO1O8aCJtywdjWGwF7Gz7XEmHAr+2PV7SzcCGwA/rxhse2UauEkmzAvsD308xU4+kjSgXUNYAxlAuoHzd9nlVgw0ASe+j/BK3NLCR7bsl7QTcafv3ddP1j6a/wxtsP9DW66Gd+2VqWUSMDJI+Bsxq+4Smoea5wILAs8D2tn9eNWDEMJP0K+BHwDnZmrj3JD0FLG/775LuAz5g+0pJSwHX9ss2cmNqBxhUtp8H/otsI1GFpFklHQzcYnt923Pbnsv2O1O8d1/zS93PgNsp0+mH1iiNpazPi2Fi+8W7Ls3zaX2keI+IYWX7J7ZPaJ5fBSwDvAtYLMV7d0l6raTvSrpO0v2SHmj9qJ2vjz0JnAbcI+kAScvWDjRg/k7pcQVwB7BR8/ztwNNVEnVB7sBXJOl04Gzbx9XOMogkPQGsbPuu2lkGjaRrgQNtnyrpX8CqtidIWhU4z/bClSNGRMRMkvReYAHbP2sZ24uyfesswAXAR20/Widh/5N0FmW/9xOB/+Ol2/dhuy+mEo9EkuYFPgbsAKwJ/JlyV/7ntvumiByJJB0IPGF7f0lbUJom3wP8G3CI7b2rBhwmKeArkvRfwFeBUylN055sPZ71qN2VCyj1NFOcVrA9sa2AfxNwg+3sUdslkt5Iad64EG2zsGwfXiVURPQVSecDvx36mSJpLeAy4FjgZkpjqZNt71EvZX9r/m1dP9uy1iVpJWAnYGfK0pHTgO/YvrlqsAEh6W3AOsBttn9TO89wSRO7uo5sHnfpcCxbaXXf74EDJI0jF1B67V5gOabexm89YHzv4wyGZunCccALwIO89I6MgRTwETEcVgH2anm9JfAX258CaBrZfZNSyEd3jCdLZauStCjwIeADlH93TwcWA66T9D+2D62Zr19JWphStA/dqHgGWFzSZ2wfXTXcMMkd+BhYaehVj6Q9KVPLdqI0NfoAZZuPQ4F9bX+/Xrr+JWk85er/V2xPqp0nIvqTpGeAZW3f3bz+C6Wp1zeb10tSZlvNXS1kn5O0PrAPsDvle52f+T3QNKn+ELAjpev51cAxwCm2n2jO2RT4se35qwXtU5K2oyxXEPAIbTcqbC/a8Y2jTO7AjwKSrgfeP/QPYQwP27kyXYntg5s1YucDcwB/pEwtOzTFe1ctDPwov8hFRJfdvbqPdQAAE+NJREFUB7wJuFvS7MDqvHQP5nkoP/Oje+4A5qTsroP00p7JuUnRNfdRisefAnvZvq7DORdRissYfvsDB1N2lHqhdphuSQE/OizJlC7d0WO5gDJ8JM0FHAJsRvlv+izgsObwTUNXp6NrzgHeBkyoHSQi+tpvgYObxnWbUpaoXdxyfBylwIzuOQWYj7JMc6omdtE1X6A0q3tmWic0zRuX6l2kgTIvcEI/F++QAj5iRixJLqAMl/2ATwA/oWznsS0wxvaWNUP1M0mbt7w8HzioaapzPfB867np+xARw+SrwBmUbvNPAB+3/VzL8R0pP4+ie9YE1rJ9Q+0gg8T2SQCS5qBsm2hg/PQK+hhWPwE2Ab5XO0g3ZQ38KNDapbt2lkGU7//wadZg72371Ob1WsAlwByZ1t0dL9ProVX6PkTEsJI0H2VLp0lt469rxp/r/M54tSRdAexi+y+1swwSSbMABwL/DcxGmU7/LKWg3Nv289N5e7xKkmYDfgU8R+cbFV+vkWu45Q58RPTSYrRMo7R9uaQXgEWBLFHogvR6iIhabD82jfGHe51lAO0DHC5pHzoXMvk76I6DgW0o28b9uRlbl1LUj6E0FYzu+TSwMfAQU2ZADDHQFwV87sCPArkDXFe+/8NH0iTgDbYfbBn7FzDO9p31kvU3Se8DjqZ8nx9vOzYfcC3wKduZ0hoR0QfaZl+1/rIvMuOqayTdD+xo+5y28U0oTWQXqZNsMEh6ADjQ9rdrZ+mm3IGPiF4ScLKk1u7DcwDHSHpqaMD2pj1P1t8+BxzSXrxDuUMm6SDg82RNakREv3h37QADaj5gfIfx8UC2jeu+scCZtUN0Wwr40eHTlA6iEaPdiR3GTu55isGzCqUz7rT8Adi7R1kiIqLLbF9YO8OAupbS+f+zbeO7Atf0Ps7AOR74GH0yVX5aUsBXJultwHuBhShrY15ke5fm8acVovU1SbNS1iZtb/vWlzk9F1CGie0damcYUK8HptfMzsACPcoSERE9IGlhSiG5IuXn/I3A0bbzO0337AmcI2kD4LJmbG1Kr5/3VUs1OOYCdpK0EXAdU/d+2KVKqmGWAr4iSbtTml3cAdzL1I0WoktsPy9pKWbg+5wLKNEH7qHsu3z7NI6PA/7RuzgREdFNktYBzqXcgLi0Gd4O2E3SRrYvneab49W4C1iOcuFk+Wbs58BRpO7qhRWAq5vny7cd65vaKk3sKpJ0N3CQ7SNrZxlEkg4BsL1H7SwR3STpu8CGwBq2n247NhdwBXC+7V1r5IuIiOEl6VJK9/mdbU9uxsYAPwBWtv2Omvn6VdOsdxHbD7SNLwA8kOaBMRxSwFck6TFg9XQ3r0PSUZR1MncCVwJPth7vl2k2EZIWolyRngwcCdzSHFqBsletgLdkWmVERH+Q9DSwWvsyQUnLA1fbnrNOsv7WdP9fuHW3nWZ8CeAm26+pkyz6SaZy1HUKZa/Co2oHGVArAFc1z5duO5YrW9E3bD8g6R2UreQOoBTsUP47/x3w2RTvERF95TFgKaC9z89SwKO9j9PfJB3RPDVwYOvOOpTO6GuRJnYxTFLA13U3sF+zTqlTo4XDq6QaELazxUoMDNsTgfdLei2wDKWIv932I3WTRUREF5wKHCtpT+Avzdg6wEGUG0gxvFZpHkW5QfRcy7HnKDeMDu11qOhPmUJfkaQ7p3PYttvvCkcXSJqDUtAYGG/7mcqRIiIiImaapNkojZI/w5Qbds9TZmJ9yfZz03pvzDxJxwO72n68dpboXyngY2A1W8kdQFkDPBvlqumzwPeAvW0/P523R0RERIwoTWPSQ4DNgFmBP1J6nzxKuUnx1HTeHhGjQKbQjxDNXp0PDnUKjZ44CNgG2JmyJzzAusCBwBhg90q5IiIiImbGfsAngJ8ATwPbAmNsb1kzVEQMn9yBr6i5A7w/ZXrTnMBytidIOgiYaDvN7bpI0v3AjrbPaRvfBPiR7UXqJIuIiIh45SSNp8wiPLV5vRZwCTCH7UlVw0XEsBhTO8CA+xrwQWA7ytTtIZdTrp5Gd80HjO8wPh6Yv8dZIiIiIl6txYCLh17Yvhx4AVi0WqKIGFYp4OvaBtjZ9q8p+zMPuQFYrk6kgXIt0Gmv913JVh8REREx+ozlpR3QoRTwWTYb0SfyP3NdiwITO4zPQv5uemFP4BxJGwCXNWNrU/5e3lctVURERMTMEXCypNaZnXMAx7TuTW57054ni4hhkSKxrhuB9YC72sa3Aq7seZoBY/siScsBnwWWb4Z/Dhxl+956ySIiIiJmyokdxk7ueYqI6JoU8HXtR7lKuhhlytOWkpandAzdpGqyAdEU6nvXzhERERHxatneoXaGiOiudKGvTNJGwJeBNSg9Ca4Cvm77vKrB+pSk9Wb0XNsXdTNLRERERETEK5ECPgaKpMmAKWvEaJ7T4TW2x/YwWkRERERExHSlC30MmtcDCzWPHwBuBbYHlmk+tgduAdLcJSIiIiIiRpTcge+xljvALyt3gLtL0pXAXrbPbxvfEDjY9up1kkVEREREREwtTex6byumFPALA18Hfglc2oy9HdgM+Frvow2cFYF7Ooz/gyld6SMiIiIiIkaE3IGvSNKZwFm2j2kb/xSwme10ou8iSVcAdwA72H66GZsTOB5YxvaaNfNFRERERES0SgFfkaQngNVs39E2vgxwre3X1Ek2GCS9FfgNMCtwXTO8CjAJ2MT232pli4iIiIiIaJcmdnU9BGzRYXwL4MEeZxk4TYG+NLAXZfu+q5rnS6V4j4iIiIiIkSZ34CuStD1luvYFTFkDvzawAfBJ2yfWyhYREREREREjSwr4yiS9DdgFWKEZuhk4wvZf66UaHJJmAdYCFgdmaz1m+8dVQkVERERERHSQAj4GlqTlgbOApQBR1r7PAjwPPGt73orxIiIiIiIiXiLbyPWYpNfN6Lm2H+5mluA7wJXAasD9zeN8wNHAPhVzRURERERETCV34HtM0mSm7AM/zdMA2x7bg0gDS9I/gfVt3yDpMWAt27dKWh/4nu1xlSNGRERERES8KHfge+/dtQPEiwQ81Tx/EPg34FbgHmCZWqEiIiIiIiI6SQHfY7YvrJ0hXnQDsCowAbgc+JKkScCngDtqBouIiIiIiGiXKfSVSZod+BiwImVq/Y3AKbafrRpsAEjaCHiN7TMkLQ2cDbwZeAjYyvafauaLiIiIiIholQK+IkkrAucC8wLXN8OrAI8BG9u+uVa2QSBpLkq3+UktY68DHnH+x4iIiIiIiBEmBXxFks6nrMH+D9uPN2PzAicDs9veqGa+fiZpLPAMsKrtm2rniYiIiIiIeDlZA1/XOsBbh4p3ANuPS9obuKxerP5ne5KkicBstbNERERERETMiDG1Awy4Z4D5O4zP1xyL7voG8C1JC9YOEhERERER8XIyhb4iSScCb6V0PR+64/524IfA5bZ3qJVtEEi6HlgKmJWyddyTrcezD3xERERERIwkmUJf167AicDFwFAjtbHAr4Ev1Ao1QH5RO0BERERERMSMyh34EUDSMsAKzcubbWcP8oiIiIiIiHiJFPAVSTpuGodMWQN/B3Ca7Xt7l2rwSHoPsGLz8ibbf6iZJyIiIiIiopMU8BVJOgtYF5gM3NAMrwwIuBJYCZgbWNf2NVVC9jFJSwGnA+OAoYskiwLXAx+xPaFWtoiIiIiIiHbpQl/XJcBvgTfaXs/2esAbgXOA84AlgLOBw+pF7GvHAv8Clra9uO3FgaWBR4EfVU0WERERERHRJnfgK5J0H/Ae2ze3ja8I/N72IpJWBy6wvUCVkH1M0tPA2ravbRtfDbjU9px1kkVEREREREwtd+DrmhtYpMP4G5pjAI+T3QK65e9ApyJ9DuDuHmeJiIiIiIiYrhTwdf0SOFbSlpKWbD62pEztPqM5Zy3gtmoJ+9sXgSMkrS1prKQxktYGvtMci4iIiIiIGDEyhb4iSXMBhwM7MOUu+wvAccDutp9spnOTJnbDQ9K/KF3+h8wBjKU0EoRyUWsS8IzteXscLyIiIiIiYppSwI8Akl4DvKl5Od72kzXz9DNJH5/Rc22f2M0sERERERERr0QK+IiIiIiIiIhRIM3RIgBJbwBmax2z/fdKcSIiIiIiIqaSAj4GlqT5gCOArWgr3htje5soIiIiIiJi2tKFPgbZocCqwGbAM8C2wB7APcDWFXNFRERERERMJWvgY2BJugfYxvbFkh4H3mL7DknbADva3rByxIiIiIiIiBflDnwMsvmBic3zx4AFmueXAu+okigiIiIiImIaUsDHIBsPLN08vxn4qCQBmwOPVEsVERERERHRQQr4GGQnAOOa598CPg08BxzSvI6IiIiIiBgxsgY+oiFpcWBN4HbgK7a3qhwpIiIiIiLiRSngI9pIWhW4yna2kYuIiIiIiBEjU+gjIiIiIiIiRoEU8BERERERERGjQAr4iIiIiIiIiFFgltoBInpN0pkvc8q8PQkSERERERHxCqSAj0H0zxk4fmcvgkRERERERMyodKGPiIiIiIiIGAWyBj4iIiIiIiJiFEgBHxERERERETEKpICPiIiIviDpXZIs6RO1s0RERHRDCviIiIguaykshz4mSXpE0g2STpS0sSSNwJyW9IykCZKOl7RC7YwRERGDLF3oIyIieucU4BxAwDzAm4HNgO2BCyRtafvRivmGDOUEmBMYB+wEfETSKrYnVksWERExwFLAR0RE9M5Vtk9uHZC0G3AwsBulcH5fjWBtOuW8HfgusDnw7em9WdI8tv/VxXwREREDKVPoIyIiKrI9yfYXgT8DG0t6Z+txSUtKOknS/0l6VtJ4SQdImqv9c0kaJ+k8SU9K+mczPX/BZir8Ca8y6r3N43Nt2SxpX0lbS7pS0tPA91rO2aDJ9GgzHf86STt3+gMkrSnpl5Iear7WWyXtLWmqGw6SPiTp6uZz3i3pG8CsHc6bo8l3q6SnmhzXSzrkVX4/IiIiei534CMiIkaGY4F3AptQinkkLQFcDswHHAXcDrwL+B9gHUnvtf1Cc+6ywMWUi/NHAP8A3g+cOxNZ5pK0YPN8TmBlYH/gIeD0DudvBuwCHA38AHi8yfSfzevLmvc/CWwIHC3pTbb3GPoEkjYBzgDuAA4DHgbeDnwdWA3YsuXcDzc57mqOvwDsQPnetfs+sCPwY+Bwyu8+ywLvmfFvR0RExMiQAj4iImJkuK55XK5l7ADg9cAmtofWpB/V3D3eHfg4pfCHUiDPC7zT9iXN2JGSTgPWeIVZ9ms+Wt0ErGv7/g7nrwSMs33z0ICkRSgXEk61vW3LuUdJ+i6wm6SjbU+QNEfzdfwVeM/QRQngh5KuBQ6X9C7bf5I0ljKV/2FgLdsPNX/eD5nyPWz1YeC3tj/+yr4FERERI0+m0EdERIwMjzeP8wJIGgNsClzdUrwPORCYTClOaYra9wOXtxTvQw6biSz/S7lTviHwQeBLwILAOc2sgHZntxbvjS2A2YFjm2n8L34AZ1F+B9mgOXdDYGHgeGD+tnOHvvZ/bx7XABYDjh8q3gFsP0a529/uMWAlSSu/gq8/IiJiRMod+IiIiJFh3uZxqJB/PTA3cGP7ibYflnQfsHTLua8Bbu3weTuNvZzbbV/Q8vo3ki6kTIU/CPho2/m3dfgcQ1vOXdDh2JCF2849bgbOHfqab+lwzk0dxj4PnARcL2kC8EfKBYSzbE+ezp8XEREx4qSAj4iIGBnGNY8zU3B3ne2/SnqMzmvHn+owNrSv/fbAfdP4tBPazt0DuGYa5947jfHpsv1rSUtSZiisT7nr/0ngYkkb2H5uOm+PiIgYUVLAR0REjAyfbB7Pbh4fBP5FWV/+EpJeCyzClGL3QUqDuDd3+LydxmbWLJRp8TPi9ubxoba7+dM798kZOHeo6F++w7EVO73B9sPAycDJkgR8C9gT+BDw85f58yIiIkaMrIGPiIioSNJYSYdSOtCfM7SGvZnefRawuqSN2962F+Xf8F82504CfgusJWmdtnO/OEw5N6RM079yBt/yM+BZYD9Jc3b4fPNJGroY8DvgAWAvSa/rcO6ckuZpXl4J3APs0NIpH0nzAju3vW+spPlbx2wbuLp5OdWfFRERMZLlDnxERETvvEXSds3zeSh3xzcDlgDOA7ZtO//LlAZvv5J0FGWLtfWArYGLgBNbzt0H2Ag4V9KRlCJ3E8r6eADPZM7ZKbMA/hN4vvlzXpbteyR9BvgRcLOkk4CJTZ5VKF/3isBdtp+UtD3wK+BWScc1X+v8lDvtm1Ma9v3J9iRJX6BcILhc0jGUbeR2BP4JLN4SYx7gPklnUor2B4ClgM8Aj1AukERERIwaKeAjIiJ6Z5vmYzLwBKXIvhA4xfZU+7XbnijpbZS9zrejFLT3ULrQf7NluzVs3yppPeBQYFfgGeA3wGcp086fnomcNFn/SbnAcKDtv83oJ7F9vKTbKFvefbrJ/xBlnf9XgPtbzv2dpLdSZhdsRyn0HwHGU/Zvv67l3F9I2gL4KrAvpTA/gXJR47yWCE8B3wHeS1n7PjdlPf6ZzdcyU+vqIyIialGZSRYRERH9SNIawBXA/9j+Vu08ERERMfOyBj4iIqJPtK81bxq27dm8PL/3iSIiImI4ZQp9RERE/7hG0h+A6ykN5z4IrAucZntGm89FRETECJUp9BEREX1C0sGUon0xykX6O4GfAAfZfr5mtoiIiHj1UsBHREREREREjAJZAx8RERERERExCqSAj4iIiIiIiBgFUsBHREREREREjAIp4CMiIiIiIiJGgRTwEREREREREaPA/wM5pLD3uM0lZAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1224x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# print the counts of the 5 most popular Dog breeds\n",
    "print(df_new['breed'].value_counts()[:5])\n",
    "\n",
    "# set the plot size\n",
    "plt.figure(figsize = (17,6))\n",
    "#value_counts already in descending order of value\n",
    "ax = sns.barplot(x = df_new['breed'].value_counts().index[0:10], y =df_new['breed'].value_counts().values[:10], data = df_new);\n",
    "ax.set_xticklabels(ax.get_xticklabels(),rotation = 90, fontsize = 14);\n",
    "#label x axis\n",
    "plt.xlabel(\"Dog Breeds\",fontsize = 18);\n",
    "plt.ylabel(\"Count\",fontsize = 16);\n",
    "# title the graph\n",
    "plt.title(\"Graph of the Popular Dog Breeds\",fontsize = 16);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "iH6Bj_PgcBA_"
   },
   "source": [
    "As it could be seen from the analysis and the graph, the most popular Dog breed of all is the **golden_retriever** which counts 137. What does a typical golden_retriver look like? This would be answered below"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 311,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "4bG-8s2ydPk3",
    "outputId": "fe2530d1-c7ec-4f2f-86bc-9e97d7e8f8d0"
   },
   "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>breed</th>\n",
       "      <th>favorite_count</th>\n",
       "      <th>jpg_url</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>92</th>\n",
       "      <td>golden_retriever</td>\n",
       "      <td>76197</td>\n",
       "      <td>https://pbs.twimg.com/media/DBQwlFCXkAACSkI.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>829</th>\n",
       "      <td>golden_retriever</td>\n",
       "      <td>51043</td>\n",
       "      <td>https://pbs.twimg.com/media/CiyHLocU4AI2pJu.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>442</th>\n",
       "      <td>golden_retriever</td>\n",
       "      <td>48559</td>\n",
       "      <td>https://pbs.twimg.com/ext_tw_video_thumb/79546...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>182</th>\n",
       "      <td>golden_retriever</td>\n",
       "      <td>42569</td>\n",
       "      <td>https://pbs.twimg.com/media/C79sB4xXwAEvwKY.jpg</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>golden_retriever</td>\n",
       "      <td>41487</td>\n",
       "      <td>https://pbs.twimg.com/media/DELC9dZXUAADqUk.jpg</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                breed  ...                                            jpg_url\n",
       "92   golden_retriever  ...    https://pbs.twimg.com/media/DBQwlFCXkAACSkI.jpg\n",
       "829  golden_retriever  ...    https://pbs.twimg.com/media/CiyHLocU4AI2pJu.jpg\n",
       "442  golden_retriever  ...  https://pbs.twimg.com/ext_tw_video_thumb/79546...\n",
       "182  golden_retriever  ...    https://pbs.twimg.com/media/C79sB4xXwAEvwKY.jpg\n",
       "39   golden_retriever  ...    https://pbs.twimg.com/media/DELC9dZXUAADqUk.jpg\n",
       "\n",
       "[5 rows x 3 columns]"
      ]
     },
     "execution_count": 311,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#let's extract the image of the golden_retriever with the highest favorite\n",
    "extract_gold_ret_img = df_new[df_new['breed'] == 'golden_retriever'].sort_values(by = 'favorite_count',ascending = False)\n",
    "# extract just the image url (jpg_url), breed and favorite_count columns\n",
    "extract_gold_ret_img = extract_gold_ret_img[['breed','favorite_count','jpg_url']][0:5]\n",
    "# view\n",
    "extract_gold_ret_img"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 312,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "JUx8yfxUgPNN",
    "outputId": "fbd6f3ad-0a88-41ec-86a1-cc59fdf93034"
   },
   "outputs": [
    {
     "data": {
      "image/jpeg": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 312,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#get the most popular by favorite count and view\n",
    "extract_gold_ret_img = extract_gold_ret_img['jpg_url'].values[0]\n",
    "# trypical golden_retriever\n",
    "Image(extract_gold_ret_img)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "JbiXnNfdiYWy"
   },
   "source": [
    "Such a very cute breed!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7Cz0vp-liiuI"
   },
   "source": [
    "#### Analysis/Visualization:\n",
    "Analyse and visualize the relationship between retweet counts, and favorite counts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 313,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "7qjphWi1iiuN",
    "outputId": "1cb5a790-a25d-4d50-e5f1-2c3563637a95"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Retweet Count')"
      ]
     },
     "execution_count": 313,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1224x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# set the plot size\n",
    "plt.figure(figsize = (17,6))\n",
    "# regression plot of retweets vs favorite count\n",
    "sns.regplot(x=\"favorite_count\", y=\"retweet_count\", data=df_new)\n",
    "# title the graph\n",
    "plt.title('Retweet Count vs Favorite Count',fontsize = 16)\n",
    "#label x axis\n",
    "plt.xlabel('Favorite Count',fontsize = 16)\n",
    "plt.ylabel('Retweet Count')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ZVxOToduiiuS"
   },
   "source": [
    "As it would be thought logical, a positive linear correlation is seen between number of retweets and the number of favorites. We could however say,a high number of retweets should be expected for a post with a high number of favorites. to further be sure about this correlation, I will find the pearson correlation between this two columns below."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 314,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "R3zcwezjiiuU",
    "outputId": "93f9b242-31b0-43e9-e467-5c9797e28afd"
   },
   "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>favorite_count</th>\n",
       "      <th>retweet_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>favorite_count</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.928452</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>retweet_count</th>\n",
       "      <td>0.928452</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                favorite_count  retweet_count\n",
       "favorite_count        1.000000       0.928452\n",
       "retweet_count         0.928452       1.000000"
      ]
     },
     "execution_count": 314,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get pearson correlation between retwet and favorite\n",
    "df_new[[\"favorite_count\", \"retweet_count\"]].corr()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7F9dsZ2kmeJs"
   },
   "source": [
    "YES! The assumption made in the comment for the earlier graph prove correct. A 92% correlation is found which is a very strong positive correlation.."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "bsTUvHernfp7"
   },
   "source": [
    "#### Analysis/Visualization:\n",
    "Analyse and Visualize the most popular source for tweets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 315,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "Qczd_2YcnpKA",
    "outputId": "00c22d2b-b784-4b83-f7de-403b58ff2bd7"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Twitter for iPhone    1925\n",
      "Twitter Web Client      28\n",
      "TweetDeck               10\n",
      "Name: source, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# print the counts of the most popular source for tweets\n",
    "print(df_new['source'].value_counts())\n",
    "\n",
    "# set the plot size\n",
    "plt.figure(figsize = (12,6))\n",
    "#value_counts already in descending order of value\n",
    "ax = sns.barplot(x = df_new['source'].value_counts().index, y =df_new['source'].value_counts().values, data = df_new);\n",
    "ax.set_xticklabels(ax.get_xticklabels(),fontsize = 14);\n",
    "#label x axis\n",
    "plt.xlabel(\"Tweet Sources\",fontsize = 18);\n",
    "plt.ylabel(\"Count\",fontsize = 16);\n",
    "# title the graph\n",
    "plt.title(\"Graph of the sources of tweets\",fontsize = 16);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "F5WMWZbepmIm"
   },
   "source": [
    "As seen above, majorly all tweets has been made from an Iphone."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "UU81wYz2q06_"
   },
   "source": [
    "#### Analysis/Visualization:\n",
    "Analyse and Visualize the most most common of all dog stages."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 316,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "id": "Wu8E7ZY7q07g",
    "outputId": "4b194e42-3cdb-4574-e431-0f63c4b8a4f8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "pupper          201\n",
      "doggo            62\n",
      "puppo            22\n",
      "doggopupper       8\n",
      "floofer           7\n",
      "doggofloofer      1\n",
      "doggopuppo        1\n",
      "Name: dog_stage, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# print the counts of the 5 most popular Dog breeds\n",
    "print(df_new['dog_stage'].value_counts())\n",
    "\n",
    "# set the plot size\n",
    "plt.figure(figsize = (12,6))\n",
    "#value_counts already in descending order of value\n",
    "ax = sns.barplot(x = df_new['dog_stage'].value_counts().index, y =df_new['dog_stage'].value_counts().values, data = df_new);\n",
    "ax.set_xticklabels(ax.get_xticklabels(),rotation = 90, fontsize = 14);\n",
    "#label x axis\n",
    "plt.xlabel(\"Dog Stage\",fontsize = 18);\n",
    "plt.ylabel(\"Count\",fontsize = 16);\n",
    "# title the graph\n",
    "plt.title(\"Graph of the Dog Stages\",fontsize = 16);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-ZceKKwcq07k"
   },
   "source": [
    "As it could be seen from the analysis and the graph, the most common Dog stage is the **pupper**  which counts 201. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2RHsjgt42kjT"
   },
   "source": [
    "## References\n",
    "- https://realpython.com/regex-python/<br>\n",
    "- Alot of [Stack Overflows](https://stackoverflow.com/)<br>\n",
    "- https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.iterrows.html<br>\n",
    "- https://seaborn.pydata.org/tutorial/regression.html\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [
    "oke6ugRxw7M3",
    "u4tm-69izgyW",
    "FVaHdZSa63eY",
    "GWVDo4tKJEtZ",
    "ctFaf2gn7mzy",
    "4NBOWbEwAw7c",
    "GPWhU4I_JvZW",
    "6VrdJfwWBFcT",
    "8Nn_LUzPBGdJ",
    "WEGXvIzGBGdN",
    "CWi12cuABGdQ",
    "hq3hXQK_BHQx",
    "4N5BENZIBHQ1",
    "9nNEz66JBH8-",
    "A9kItSyLBH9C",
    "uiPI5qSTBH9I",
    "l3tezM_UBIqh",
    "54K-vZ5_6oWV",
    "AkB07tKi61rF",
    "55cH3cdDBJZ_",
    "S4bONzckEMPa",
    "ye2cpAdME40x"
   ],
   "name": "wrangle_act.ipynb",
   "provenance": [],
   "toc_visible": true
  },
  "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}