Final Project 4.22.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Final Project: Movie Script Analysis ( QT & Disney )\n",
    "## Pre-cleaning and Parsing:\n",
    "### QTs:\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "qts = [os.path.join(\"data/QT\", p) for p in os.listdir(\"data/QT\")]\n",
    "\n",
    "def clean_data(text):\n",
    "    text = re.sub('\\n','',text)\n",
    "    text1 = re.sub(\"'s\",' is',text)\n",
    "    text2 = re.sub(\"'m\",' am', text1)\n",
    "    text3 = re.sub(\"n't\", ' not', text2)\n",
    "    text4 = re.sub(\"'ll\", ' will', text3)\n",
    "    text5 = re.sub(\"'ve\", ' have', text4)\n",
    "    return text5\n",
    "\n",
    "\n",
    "qt_data = []\n",
    "for p in qts:\n",
    "    with open(p) as f:\n",
    "        content = \"\"\n",
    "        for line in f:\n",
    "            line = line.strip().lower()\n",
    "            line = \" \".join(line.split(\"\\n\"))\n",
    "            line = clean_data(line)\n",
    "            line = re.sub(r\"\\s+\", \" \", line) \n",
    "            line = re.sub(r\"\\t+\", \" \", line)\n",
    "            line = line.strip().replace(u'\\u3000', u'').replace(u'\\xa0', u'')\n",
    "            # '\\a0': 不间断空白符、'\\u3000': 特殊空白符\n",
    "            # https://www.cnblogs.com/zqifa/p/python-9.html\n",
    "            line = re.sub('[^A-Za-z ]*','',line)\n",
    "            content  += \" \" + line\n",
    "        content = \" \".join([w for w in content.split(\" \") if len(w)>0])\n",
    "        qt_data.append(content)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# stemmer\n",
    "\n",
    "from nltk.stem import PorterStemmer\n",
    "def clean_article(text):\n",
    "    ps = PorterStemmer()\n",
    "    text_new = ps.stem(text) \n",
    "    return text_new\n",
    "\n",
    "qt = []\n",
    "for i in range(len(qt_data)):\n",
    "    qt.append(clean_article(qt_data[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "stopwords_qt = ['django','rider','dicky','speck','ace','le','schultz','saloon','keeper','sheriff','sharp','marshall',\n",
    "                'tatum','zz','spencer','betina','scotty','broomhilda','narrator','vo','calvin','candie','stephen','woody',\n",
    "                'german','lara','lee','flody','jano','bennett','chester','icz','billy','crash','ted','guy','betty','elspeth',\n",
    "                'kiva','athena','raven','eva','jezabel','yuppie','scum','sigfried','juancho','sarah','margaret','angela',\n",
    "                'leo','norman','mcgraw','pete','seth','richard','jacob','scott','kate','kelly','houge','santanico','frost',\n",
    "                'carlos','farmer','julie','perrier','col','landa','charlotte','driver','lt','aldo','butz','pvt','shosanna',\n",
    "                'marcel','goebbels','narrator','fredrick','zoller','bridget','hellstorm','hicox','stiglitz','wicki','wilhelm',\n",
    "                'willi','italian','donowitz','sidney','ordell','louis','jackie','melanie','nicolet','dargus','max','sheronda',\n",
    "                'wanda','amy','beaumont','bill','housewife','vernita','elle','bride','orderly','slam','sushi','chef','budd','pai',\n",
    "                'mei','mickey','mabel','earl','mallory','mcclusky','scagnetti','wurlizter','sonny','wayne','russell','steve',\n",
    "                'simon','jeff','movie','jessie','otis','young','man','woman','jody','lance','vincent','jules','brett','butch',\n",
    "                'mia','esmarelda','fabienne','maynard','zed','jimmie','wolf','pumpkin','yolanda','mr','pink','white','blue','joe',\n",
    "                'brown','orange','blonde','eddie','vic','cop','holdaway','freddy','jodie','lucy','clarence','drexl','d','cliff',\n",
    "                'alabama','dick','coccotti','marty','elvis','elliot','lenny','marvin','dimes','dario'] # characters' name in QT's movies\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# remove stopwords\n",
    "\n",
    "from nltk.corpus import stopwords\n",
    "\n",
    "stop = stopwords.words('english')\n",
    "\n",
    "def rm_swqt (text):\n",
    "    new = []\n",
    "    for i in range(len(text)):\n",
    "        new1 = ''\n",
    "        for word in text[i].split(\" \"):\n",
    "            if word not in stop:\n",
    "                if word not in stopwords_qt:\n",
    "                    new1 += word  + \" \"\n",
    "        new.append(new1)\n",
    "    return new\n",
    "                \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "qt_cleaned_data = rm_swqt(qt) # QT's cleaned data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Disneys:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "disney = [os.path.join(\"data/Disney\", p) for p in os.listdir(\"data/Disney\")]\n",
    "\n",
    "def clean_data(text):\n",
    "    text = re.sub('\\n','',text)\n",
    "    text1 = re.sub(\"'s\",' is',text)\n",
    "    text2 = re.sub(\"'m\",' am', text1)\n",
    "    text3 = re.sub(\"n't\", ' not', text2)\n",
    "    text4 = re.sub(\"'ll\", ' will', text3)\n",
    "    text5 = re.sub(\"'ve\", ' have', text4)\n",
    "    return text5\n",
    "\n",
    "\n",
    "disney_data = []\n",
    "for p in disney:\n",
    "    with open(p) as f:\n",
    "        content = \"\"\n",
    "        for line in f:\n",
    "            line = line.strip().lower()\n",
    "            line = \" \".join(line.split(\"\\n\"))\n",
    "            line = clean_data(line)\n",
    "            line = re.sub(r\"\\s+\", \" \", line) \n",
    "            line = re.sub(r\"\\t+\", \" \", line)\n",
    "            line = line.strip().replace(u'\\u3000', u'').replace(u'\\xa0', u'') # 去除‘\\t’ 也无法去除的空格\n",
    "            line = re.sub('[^A-Za-z ]*','',line)\n",
    "            content  += \" \" + line\n",
    "        content = \" \".join([w for w in content.split(\" \") if len(w)>0])\n",
    "        disney_data.append(content)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# stemmer\n",
    "\n",
    "disney = []\n",
    "for i in range(len(disney_data)):\n",
    "    disney.append(clean_article(disney_data[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "stopwords_disney = ['mulan','guard','shan','yu','li','chi','fu','general','emperor','fa','hair','dresser',\n",
    "                   'chorus','maiden','zhou','grandma','matchmaker','mushu','shang','ancestor','barry','cook','hun','ling',\n",
    "                   'yao','chien','po','cri','kee','jafar','peddler','gazeem','aladdin','guardus','iago','prince','sultan',\n",
    "                   'proprietor','abu','cave','jasmine','shopkeeper','genie','chorus','narrator','maid','majordomo','footman',\n",
    "                   'jean','belle','villagers','clothilde','tom','dick','pere','robert','apothecary','gaston','diva','vagrant',\n",
    "                   'lefou','maurice','nasty','chip','lasses','lumiere','chandelier','beast','plumette','potts','mrs',\n",
    "                   'cadenza','cogsworth','tavern','marlin','coral','nemo','bob','pearl','sheldon','ray','mr','tad','dory',\n",
    "                   'bruce','anchor','chum','nigel','bubbles','gurgle','bloat','patient','deb','peach','gill','jacques',\n",
    "                   'crush','squirt','crab','gull','darla','ice','harvesters','anna','elsa','young','queen','king','kristoff',\n",
    "                   'grand','pabbie','duke','kai','hans','both','oaken','olaf','bulda','bert','admiral','boom','binnalce',\n",
    "                   'ellen','brill','katie','nanna','banks','jane','constable','minchael','mary','poppins','albert','dawes',\n",
    "                   'sailors','eric','grimsby','seahorse','flounder','ariel','triton','sebastian','ursula','jetsam','louis',\n",
    "                   'pilot','serbian','shane','charles','seal','julie','chun','helga','zoe','todd','lulu','peter','murney',\n",
    "                   'claire','romeo','juliet','gregory','sampson','abra','benvolio','tybalt','capulet','lawyer','caroline',\n",
    "                   'mercutio','nurse','altar','judy','gideon','bonnie','hopps','stu','astronaut','jaguar','grey','friedkin',\n",
    "                   'major','kudu','pootosser','nick','yax','bogo','lionheart','badger','doctor','leopard','gazelle','bellwether',\n",
    "                   'prey','anchor','otterton','woolter']\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def rm_swdisney (text):\n",
    "    new = []\n",
    "    for i in range(len(text)):\n",
    "        new1 = ''\n",
    "        for word in text[i].split(\" \"):\n",
    "            if word not in stop:\n",
    "                if word not in stopwords_disney:\n",
    "                    new1 += word  + \" \"\n",
    "        new.append(new1)\n",
    "    return new"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "disney_cleaned_data = rm_swdisney(disney)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sentiment Analysis: \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ANEW："
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "ANEW = pd.read_csv(\"all.csv\")\n",
    "new = ANEW[['Description','Valence Mean']]\n",
    "AS = new['Description'].tolist()\n",
    "AX = new['Valence Mean'].tolist()\n",
    "ANEW = dict(zip(AS,AX))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def anew_score(text):\n",
    "    dict_anew = {}\n",
    "    score = 0\n",
    "    for word in text.split():\n",
    "        if word in ANEW.keys():\n",
    "            if word not in dict_anew.keys():\n",
    "                dict_anew[word] = 1\n",
    "            else:\n",
    "                dict_anew[word] +=1\n",
    "    for word in dict_anew.keys():\n",
    "        score += (dict_anew[word] * ANEW[word])/sum(dict_anew.values())\n",
    "    return score"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Havard IV:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:5: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \"\"\"\n",
      "/opt/anaconda3/lib/python3.7/site-packages/pandas/core/generic.py:6786: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  self._update_inplace(new_data)\n",
      "/opt/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:8: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  \n",
      "/opt/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:13: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  del sys.path[0]\n",
      "/opt/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:16: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  app.launch_new_instance()\n"
     ]
    }
   ],
   "source": [
    "po_ne = pd.read_excel('inquireraugmented.xls') \n",
    "\n",
    "# negative list:\n",
    "negative = po_ne[['Entry', 'Negativ']]\n",
    "negative.dropna(subset = ['Negativ'], inplace = True)\n",
    "negative_1 = negative[1:]\n",
    "negative_1['Entry'].replace('[^a-zA-Z_]+', '',regex = True, inplace = True)\n",
    "negative_1.drop_duplicates('Entry', 'first', inplace = True)\n",
    "negative_newlist = [str(item).lower() for item in negative_1['Entry']]\n",
    "\n",
    "# positive list:\n",
    "positive = po_ne[['Entry', 'Positiv']]\n",
    "positive.dropna(subset = ['Positiv'], inplace = True)\n",
    "positive_1 = positive[1:]\n",
    "positive_1['Entry'].replace('[^a-zA-Z_]+', '',regex = True, inplace = True)\n",
    "positive_1.drop_duplicates('Entry', 'first', inplace = True)\n",
    "positive_newlist = [str(item).lower() for item in positive_1['Entry']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define the function to test the positive_score and negative_score.\n",
    "\n",
    "def sentiment_score(text_input, sentiment_list):\n",
    "    score_Dict = {}\n",
    "    for i in range(len(text_input)):\n",
    "        number = 0\n",
    "        for word in text_input[i].split():\n",
    "            if word in sentiment_list:\n",
    "                number += 1\n",
    "                score_Dict[i] = number / len(text_input[i].split())\n",
    "    return(score_Dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "movie = ['Four Rooms','Inglourious Basterds','From Dusk Till Down','Natural Born killers','Django Unchained','Pulp Fiction','True Romance','Reservoir Dogs','Jackie Brown','Kill Bill']\n",
    "Year = ['1995','2008','1996','1995','2012','1993','1993','1990','1997','2003']\n",
    "Genres = ['comedy,drama','action,adventure,war','action,comedy,horror,thriller','action,romance,thriller,crime','adventure,drama,western','action,crime,drama,thriller','action,romance,thriller,crime','action,crime,thriller','comedy,crime','action,comedy,crime,drama,thriller']\n",
    "Rating = ['8.63','7.44','7.25','8.52','7.82','9.39','9.88','8.89','8.27','8.73']\n",
    "\n",
    "QT_df = pd.DataFrame(qt_cleaned_data)\n",
    "QT_df.columns = ['Script']\n",
    "QT_df.insert(0,'movie',movie)\n",
    "QT_df.insert(1,'Year',Year)\n",
    "QT_df.insert(2,'Genres',Genres)\n",
    "QT_df.insert(4,'Rating',Rating)\n",
    "# QT_df.to_csv('QT_data.csv')\n",
    "# QT dataframe for sentiment analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>movie</th>\n",
       "      <th>Year</th>\n",
       "      <th>Genres</th>\n",
       "      <th>Script</th>\n",
       "      <th>Rating</th>\n",
       "      <th>neg_score</th>\n",
       "      <th>pos_score</th>\n",
       "      <th>sentiment_score_iv</th>\n",
       "      <th>sentiment_score_anew</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Four Rooms</td>\n",
       "      <td>1995</td>\n",
       "      <td>comedy,drama</td>\n",
       "      <td>four rooms screenplay allison anders alexandre...</td>\n",
       "      <td>8.63</td>\n",
       "      <td>0.090876</td>\n",
       "      <td>0.093699</td>\n",
       "      <td>0.002823</td>\n",
       "      <td>5.878314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Inglourious Basterds</td>\n",
       "      <td>2008</td>\n",
       "      <td>action,adventure,war</td>\n",
       "      <td>jppinglourious basterds written quentin tarant...</td>\n",
       "      <td>7.44</td>\n",
       "      <td>0.077510</td>\n",
       "      <td>0.092213</td>\n",
       "      <td>0.014703</td>\n",
       "      <td>5.646904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>From Dusk Till Down</td>\n",
       "      <td>1996</td>\n",
       "      <td>action,comedy,horror,thriller</td>\n",
       "      <td>rkfrom dusk till dawn screenplay quentin taran...</td>\n",
       "      <td>7.25</td>\n",
       "      <td>0.113527</td>\n",
       "      <td>0.080819</td>\n",
       "      <td>-0.032708</td>\n",
       "      <td>5.639600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Natural Born killers</td>\n",
       "      <td>1995</td>\n",
       "      <td>action,romance,thriller,crime</td>\n",
       "      <td>mewsijjnatural born killers written quentin ta...</td>\n",
       "      <td>8.52</td>\n",
       "      <td>0.098840</td>\n",
       "      <td>0.086155</td>\n",
       "      <td>-0.012686</td>\n",
       "      <td>5.310696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Django Unchained</td>\n",
       "      <td>2012</td>\n",
       "      <td>adventure,drama,western</td>\n",
       "      <td>unchained written quentin tarantino ext countr...</td>\n",
       "      <td>7.82</td>\n",
       "      <td>0.089610</td>\n",
       "      <td>0.091363</td>\n",
       "      <td>0.001753</td>\n",
       "      <td>5.594078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>Pulp Fiction</td>\n",
       "      <td>1993</td>\n",
       "      <td>action,crime,drama,thriller</td>\n",
       "      <td>vbw pulp fiction quentin tarantino roger avary...</td>\n",
       "      <td>9.39</td>\n",
       "      <td>0.094869</td>\n",
       "      <td>0.089338</td>\n",
       "      <td>-0.005531</td>\n",
       "      <td>5.749571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>True Romance</td>\n",
       "      <td>1993</td>\n",
       "      <td>action,romance,thriller,crime</td>\n",
       "      <td>cdameddlucy laughs well enough king bout bout ...</td>\n",
       "      <td>9.88</td>\n",
       "      <td>0.083703</td>\n",
       "      <td>0.099120</td>\n",
       "      <td>0.015417</td>\n",
       "      <td>5.883581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Reservoir Dogs</td>\n",
       "      <td>1990</td>\n",
       "      <td>action,crime,thriller</td>\n",
       "      <td>quentin tarantino r e e r v r g october dedica...</td>\n",
       "      <td>8.89</td>\n",
       "      <td>0.102989</td>\n",
       "      <td>0.093266</td>\n",
       "      <td>-0.009723</td>\n",
       "      <td>5.462150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>Jackie Brown</td>\n",
       "      <td>1997</td>\n",
       "      <td>comedy,crime</td>\n",
       "      <td>ldajackie screenplay quentin tarantino opening...</td>\n",
       "      <td>8.27</td>\n",
       "      <td>0.079589</td>\n",
       "      <td>0.083852</td>\n",
       "      <td>0.004262</td>\n",
       "      <td>5.778873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>Kill Bill</td>\n",
       "      <td>2003</td>\n",
       "      <td>action,comedy,crime,drama,thriller</td>\n",
       "      <td>hodsover black hear labored breathing black fr...</td>\n",
       "      <td>8.73</td>\n",
       "      <td>0.090782</td>\n",
       "      <td>0.077276</td>\n",
       "      <td>-0.013506</td>\n",
       "      <td>5.655627</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  movie  Year                              Genres  \\\n",
       "0            Four Rooms  1995                        comedy,drama   \n",
       "1  Inglourious Basterds  2008                action,adventure,war   \n",
       "2   From Dusk Till Down  1996       action,comedy,horror,thriller   \n",
       "3  Natural Born killers  1995       action,romance,thriller,crime   \n",
       "4      Django Unchained  2012             adventure,drama,western   \n",
       "5          Pulp Fiction  1993         action,crime,drama,thriller   \n",
       "6          True Romance  1993       action,romance,thriller,crime   \n",
       "7        Reservoir Dogs  1990               action,crime,thriller   \n",
       "8          Jackie Brown  1997                        comedy,crime   \n",
       "9             Kill Bill  2003  action,comedy,crime,drama,thriller   \n",
       "\n",
       "                                              Script Rating  neg_score  \\\n",
       "0  four rooms screenplay allison anders alexandre...   8.63   0.090876   \n",
       "1  jppinglourious basterds written quentin tarant...   7.44   0.077510   \n",
       "2  rkfrom dusk till dawn screenplay quentin taran...   7.25   0.113527   \n",
       "3  mewsijjnatural born killers written quentin ta...   8.52   0.098840   \n",
       "4  unchained written quentin tarantino ext countr...   7.82   0.089610   \n",
       "5  vbw pulp fiction quentin tarantino roger avary...   9.39   0.094869   \n",
       "6  cdameddlucy laughs well enough king bout bout ...   9.88   0.083703   \n",
       "7  quentin tarantino r e e r v r g october dedica...   8.89   0.102989   \n",
       "8  ldajackie screenplay quentin tarantino opening...   8.27   0.079589   \n",
       "9  hodsover black hear labored breathing black fr...   8.73   0.090782   \n",
       "\n",
       "   pos_score  sentiment_score_iv  sentiment_score_anew  \n",
       "0   0.093699            0.002823              5.878314  \n",
       "1   0.092213            0.014703              5.646904  \n",
       "2   0.080819           -0.032708              5.639600  \n",
       "3   0.086155           -0.012686              5.310696  \n",
       "4   0.091363            0.001753              5.594078  \n",
       "5   0.089338           -0.005531              5.749571  \n",
       "6   0.099120            0.015417              5.883581  \n",
       "7   0.093266           -0.009723              5.462150  \n",
       "8   0.083852            0.004262              5.778873  \n",
       "9   0.077276           -0.013506              5.655627  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Qt_script_list = list(QT_df['Script'])\n",
    "negativeDict = sentiment_score(Qt_script_list, negative_newlist)\n",
    "positiveDict = sentiment_score(Qt_script_list, positive_newlist)\n",
    "neg = list(negativeDict.values())\n",
    "pos = list(positiveDict.values())\n",
    "QT_df.insert(5,'neg_score',neg)\n",
    "QT_df.insert(6,'pos_score',pos)\n",
    "QT_df['sentiment_score_iv'] = QT_df['pos_score'] - QT_df['neg_score']\n",
    "QT_df['sentiment_score_anew'] = QT_df['Script'].apply(lambda x: anew_score(x))\n",
    "QT_df # QT dataframe with \"Harvard IV score\" and \"ANEW score\" "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "movie = ['Aladdin','Finding Nemo','Mary Poppins','Mulan','The Little Mermaid','The Pacifier','Zootopia',' Beauty and Beast','ROMEO & JULIET','Frozen']\n",
    "Year = ['1992','2003','1964','1998','1989','2004','2016','2016','1995','2013']\n",
    "Genres = ['family,animation,comedy,musical,fantasy','family,animation,comedy,adventure','family,animation,comedy,musical,fantasy,drama','family,animation,comedy,musical,adventure','family,animation,comedy,musical,romance,fantasy','comedy,action,family,thriller','adventure,animation,comedy','family,fantasy,musical','drama,romance,comedy','adventure,animation,comedy']\n",
    "Rating = ['9.24','9.20',' 9.38','9.82','9.73','6.0','10.0','9.67','10.0','10.0']\n",
    "\n",
    "disney_df = pd.DataFrame(disney_cleaned_data)\n",
    "disney_df.columns = ['Script']\n",
    "disney_df.insert(0,'movie',movie)\n",
    "disney_df.insert(1,'Year',Year)\n",
    "disney_df.insert(2,'Genres',Genres)\n",
    "disney_df.insert(4,'Rating',Rating)\n",
    "# Disney dataframe for sentiment analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>movie</th>\n",
       "      <th>Year</th>\n",
       "      <th>Genres</th>\n",
       "      <th>Script</th>\n",
       "      <th>Rating</th>\n",
       "      <th>neg_score</th>\n",
       "      <th>pos_score</th>\n",
       "      <th>sentiment_score_iv</th>\n",
       "      <th>sentiment_score_anew</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Aladdin</td>\n",
       "      <td>1992</td>\n",
       "      <td>family,animation,comedy,musical,fantasy</td>\n",
       "      <td>gsavjgecaladdin complete script compiled ben s...</td>\n",
       "      <td>9.24</td>\n",
       "      <td>0.089018</td>\n",
       "      <td>0.105943</td>\n",
       "      <td>0.016925</td>\n",
       "      <td>5.900127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Finding Nemo</td>\n",
       "      <td>2003</td>\n",
       "      <td>family,animation,comedy,adventure</td>\n",
       "      <td>cntan finding transcript v copyright walt disn...</td>\n",
       "      <td>9.20</td>\n",
       "      <td>0.090095</td>\n",
       "      <td>0.109044</td>\n",
       "      <td>0.018949</td>\n",
       "      <td>6.036038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Mary Poppins</td>\n",
       "      <td>1964</td>\n",
       "      <td>family,animation,comedy,musical,fantasy,drama</td>\n",
       "      <td>abhap right ladies gents comical poems suitabl...</td>\n",
       "      <td>9.38</td>\n",
       "      <td>0.073023</td>\n",
       "      <td>0.147867</td>\n",
       "      <td>0.074844</td>\n",
       "      <td>6.217262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Mulan</td>\n",
       "      <td>1998</td>\n",
       "      <td>family,animation,comedy,musical,adventure</td>\n",
       "      <td>chzgrmmshhycdisney compiled adams theater show...</td>\n",
       "      <td>9.82</td>\n",
       "      <td>0.082271</td>\n",
       "      <td>0.087397</td>\n",
       "      <td>0.005126</td>\n",
       "      <td>5.921347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>The Little Mermaid</td>\n",
       "      <td>1989</td>\n",
       "      <td>family,animation,comedy,musical,romance,fantasy</td>\n",
       "      <td>shthe little mermaid ocean birds flying porpoi...</td>\n",
       "      <td>9.73</td>\n",
       "      <td>0.099754</td>\n",
       "      <td>0.121891</td>\n",
       "      <td>0.022137</td>\n",
       "      <td>6.156296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>The Pacifier</td>\n",
       "      <td>2004</td>\n",
       "      <td>comedy,action,family,thriller</td>\n",
       "      <td>cthe pacifier fade ext pacific ocean day fishi...</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.078187</td>\n",
       "      <td>0.081914</td>\n",
       "      <td>0.003727</td>\n",
       "      <td>5.925138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>Zootopia</td>\n",
       "      <td>2016</td>\n",
       "      <td>adventure,animation,comedy</td>\n",
       "      <td>spyblgbazootopia written jared bush phil johns...</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.081025</td>\n",
       "      <td>0.092248</td>\n",
       "      <td>0.011222</td>\n",
       "      <td>5.872771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Beauty and Beast</td>\n",
       "      <td>2016</td>\n",
       "      <td>family,fantasy,musical</td>\n",
       "      <td>mgncbeauty written stephen chbosky evan spilio...</td>\n",
       "      <td>9.67</td>\n",
       "      <td>0.073202</td>\n",
       "      <td>0.093116</td>\n",
       "      <td>0.019915</td>\n",
       "      <td>6.004309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>ROMEO &amp; JULIET</td>\n",
       "      <td>1995</td>\n",
       "      <td>drama,romance,comedy</td>\n",
       "      <td>sacwilliam shakespeare adapted craig pearce ba...</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.099545</td>\n",
       "      <td>0.090849</td>\n",
       "      <td>-0.008696</td>\n",
       "      <td>5.981083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>Frozen</td>\n",
       "      <td>2013</td>\n",
       "      <td>adventure,animation,comedy</td>\n",
       "      <td>quhfrozen written jennifer lee final shooting ...</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.087554</td>\n",
       "      <td>0.101502</td>\n",
       "      <td>0.013948</td>\n",
       "      <td>5.924542</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                movie  Year                                           Genres  \\\n",
       "0             Aladdin  1992          family,animation,comedy,musical,fantasy   \n",
       "1        Finding Nemo  2003                family,animation,comedy,adventure   \n",
       "2        Mary Poppins  1964    family,animation,comedy,musical,fantasy,drama   \n",
       "3               Mulan  1998        family,animation,comedy,musical,adventure   \n",
       "4  The Little Mermaid  1989  family,animation,comedy,musical,romance,fantasy   \n",
       "5        The Pacifier  2004                    comedy,action,family,thriller   \n",
       "6            Zootopia  2016                       adventure,animation,comedy   \n",
       "7    Beauty and Beast  2016                           family,fantasy,musical   \n",
       "8      ROMEO & JULIET  1995                             drama,romance,comedy   \n",
       "9              Frozen  2013                       adventure,animation,comedy   \n",
       "\n",
       "                                              Script Rating  neg_score  \\\n",
       "0  gsavjgecaladdin complete script compiled ben s...   9.24   0.089018   \n",
       "1  cntan finding transcript v copyright walt disn...   9.20   0.090095   \n",
       "2  abhap right ladies gents comical poems suitabl...   9.38   0.073023   \n",
       "3  chzgrmmshhycdisney compiled adams theater show...   9.82   0.082271   \n",
       "4  shthe little mermaid ocean birds flying porpoi...   9.73   0.099754   \n",
       "5  cthe pacifier fade ext pacific ocean day fishi...    6.0   0.078187   \n",
       "6  spyblgbazootopia written jared bush phil johns...   10.0   0.081025   \n",
       "7  mgncbeauty written stephen chbosky evan spilio...   9.67   0.073202   \n",
       "8  sacwilliam shakespeare adapted craig pearce ba...   10.0   0.099545   \n",
       "9  quhfrozen written jennifer lee final shooting ...   10.0   0.087554   \n",
       "\n",
       "   pos_score  sentiment_score_iv  sentiment_score_anew  \n",
       "0   0.105943            0.016925              5.900127  \n",
       "1   0.109044            0.018949              6.036038  \n",
       "2   0.147867            0.074844              6.217262  \n",
       "3   0.087397            0.005126              5.921347  \n",
       "4   0.121891            0.022137              6.156296  \n",
       "5   0.081914            0.003727              5.925138  \n",
       "6   0.092248            0.011222              5.872771  \n",
       "7   0.093116            0.019915              6.004309  \n",
       "8   0.090849           -0.008696              5.981083  \n",
       "9   0.101502            0.013948              5.924542  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "disney_script_list = list(disney_df['Script'])\n",
    "negativeDict1 = sentiment_score(disney_script_list, negative_newlist)\n",
    "positiveDict1 = sentiment_score(disney_script_list, positive_newlist)\n",
    "neg = list(negativeDict1.values())\n",
    "pos = list(positiveDict1.values())\n",
    "disney_df.insert(5,'neg_score',neg)\n",
    "disney_df.insert(6,'pos_score',pos)\n",
    "disney_df['sentiment_score_iv'] = disney_df['pos_score'] - disney_df['neg_score']\n",
    "disney_df['sentiment_score_anew'] = disney_df['Script'].apply(lambda x: anew_score(x))\n",
    "disney_df # Disney dataframe with \"Harvard IV score\" and \"ANEW score\" "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Type</th>\n",
       "      <th>movie</th>\n",
       "      <th>Year</th>\n",
       "      <th>Genres</th>\n",
       "      <th>Script</th>\n",
       "      <th>Rating</th>\n",
       "      <th>neg_score</th>\n",
       "      <th>pos_score</th>\n",
       "      <th>sentiment_score_iv</th>\n",
       "      <th>sentiment_score_anew</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>QT</td>\n",
       "      <td>Four Rooms</td>\n",
       "      <td>1995</td>\n",
       "      <td>comedy,drama</td>\n",
       "      <td>four rooms screenplay allison anders alexandre...</td>\n",
       "      <td>8.63</td>\n",
       "      <td>0.090876</td>\n",
       "      <td>0.093699</td>\n",
       "      <td>0.002823</td>\n",
       "      <td>5.878314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>QT</td>\n",
       "      <td>Inglourious Basterds</td>\n",
       "      <td>2008</td>\n",
       "      <td>action,adventure,war</td>\n",
       "      <td>jppinglourious basterds written quentin tarant...</td>\n",
       "      <td>7.44</td>\n",
       "      <td>0.077510</td>\n",
       "      <td>0.092213</td>\n",
       "      <td>0.014703</td>\n",
       "      <td>5.646904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>QT</td>\n",
       "      <td>From Dusk Till Down</td>\n",
       "      <td>1996</td>\n",
       "      <td>action,comedy,horror,thriller</td>\n",
       "      <td>rkfrom dusk till dawn screenplay quentin taran...</td>\n",
       "      <td>7.25</td>\n",
       "      <td>0.113527</td>\n",
       "      <td>0.080819</td>\n",
       "      <td>-0.032708</td>\n",
       "      <td>5.639600</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>QT</td>\n",
       "      <td>Natural Born killers</td>\n",
       "      <td>1995</td>\n",
       "      <td>action,romance,thriller,crime</td>\n",
       "      <td>mewsijjnatural born killers written quentin ta...</td>\n",
       "      <td>8.52</td>\n",
       "      <td>0.098840</td>\n",
       "      <td>0.086155</td>\n",
       "      <td>-0.012686</td>\n",
       "      <td>5.310696</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>QT</td>\n",
       "      <td>Django Unchained</td>\n",
       "      <td>2012</td>\n",
       "      <td>adventure,drama,western</td>\n",
       "      <td>unchained written quentin tarantino ext countr...</td>\n",
       "      <td>7.82</td>\n",
       "      <td>0.089610</td>\n",
       "      <td>0.091363</td>\n",
       "      <td>0.001753</td>\n",
       "      <td>5.594078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>QT</td>\n",
       "      <td>Pulp Fiction</td>\n",
       "      <td>1993</td>\n",
       "      <td>action,crime,drama,thriller</td>\n",
       "      <td>vbw pulp fiction quentin tarantino roger avary...</td>\n",
       "      <td>9.39</td>\n",
       "      <td>0.094869</td>\n",
       "      <td>0.089338</td>\n",
       "      <td>-0.005531</td>\n",
       "      <td>5.749571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>QT</td>\n",
       "      <td>True Romance</td>\n",
       "      <td>1993</td>\n",
       "      <td>action,romance,thriller,crime</td>\n",
       "      <td>cdameddlucy laughs well enough king bout bout ...</td>\n",
       "      <td>9.88</td>\n",
       "      <td>0.083703</td>\n",
       "      <td>0.099120</td>\n",
       "      <td>0.015417</td>\n",
       "      <td>5.883581</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>QT</td>\n",
       "      <td>Reservoir Dogs</td>\n",
       "      <td>1990</td>\n",
       "      <td>action,crime,thriller</td>\n",
       "      <td>quentin tarantino r e e r v r g october dedica...</td>\n",
       "      <td>8.89</td>\n",
       "      <td>0.102989</td>\n",
       "      <td>0.093266</td>\n",
       "      <td>-0.009723</td>\n",
       "      <td>5.462150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>QT</td>\n",
       "      <td>Jackie Brown</td>\n",
       "      <td>1997</td>\n",
       "      <td>comedy,crime</td>\n",
       "      <td>ldajackie screenplay quentin tarantino opening...</td>\n",
       "      <td>8.27</td>\n",
       "      <td>0.079589</td>\n",
       "      <td>0.083852</td>\n",
       "      <td>0.004262</td>\n",
       "      <td>5.778873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>QT</td>\n",
       "      <td>Kill Bill</td>\n",
       "      <td>2003</td>\n",
       "      <td>action,comedy,crime,drama,thriller</td>\n",
       "      <td>hodsover black hear labored breathing black fr...</td>\n",
       "      <td>8.73</td>\n",
       "      <td>0.090782</td>\n",
       "      <td>0.077276</td>\n",
       "      <td>-0.013506</td>\n",
       "      <td>5.655627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>Disney</td>\n",
       "      <td>Aladdin</td>\n",
       "      <td>1992</td>\n",
       "      <td>family,animation,comedy,musical,fantasy</td>\n",
       "      <td>gsavjgecaladdin complete script compiled ben s...</td>\n",
       "      <td>9.24</td>\n",
       "      <td>0.089018</td>\n",
       "      <td>0.105943</td>\n",
       "      <td>0.016925</td>\n",
       "      <td>5.900127</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>Disney</td>\n",
       "      <td>Finding Nemo</td>\n",
       "      <td>2003</td>\n",
       "      <td>family,animation,comedy,adventure</td>\n",
       "      <td>cntan finding transcript v copyright walt disn...</td>\n",
       "      <td>9.20</td>\n",
       "      <td>0.090095</td>\n",
       "      <td>0.109044</td>\n",
       "      <td>0.018949</td>\n",
       "      <td>6.036038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>Disney</td>\n",
       "      <td>Mary Poppins</td>\n",
       "      <td>1964</td>\n",
       "      <td>family,animation,comedy,musical,fantasy,drama</td>\n",
       "      <td>abhap right ladies gents comical poems suitabl...</td>\n",
       "      <td>9.38</td>\n",
       "      <td>0.073023</td>\n",
       "      <td>0.147867</td>\n",
       "      <td>0.074844</td>\n",
       "      <td>6.217262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>Disney</td>\n",
       "      <td>Mulan</td>\n",
       "      <td>1998</td>\n",
       "      <td>family,animation,comedy,musical,adventure</td>\n",
       "      <td>chzgrmmshhycdisney compiled adams theater show...</td>\n",
       "      <td>9.82</td>\n",
       "      <td>0.082271</td>\n",
       "      <td>0.087397</td>\n",
       "      <td>0.005126</td>\n",
       "      <td>5.921347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>Disney</td>\n",
       "      <td>The Little Mermaid</td>\n",
       "      <td>1989</td>\n",
       "      <td>family,animation,comedy,musical,romance,fantasy</td>\n",
       "      <td>shthe little mermaid ocean birds flying porpoi...</td>\n",
       "      <td>9.73</td>\n",
       "      <td>0.099754</td>\n",
       "      <td>0.121891</td>\n",
       "      <td>0.022137</td>\n",
       "      <td>6.156296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>Disney</td>\n",
       "      <td>The Pacifier</td>\n",
       "      <td>2004</td>\n",
       "      <td>comedy,action,family,thriller</td>\n",
       "      <td>cthe pacifier fade ext pacific ocean day fishi...</td>\n",
       "      <td>6.0</td>\n",
       "      <td>0.078187</td>\n",
       "      <td>0.081914</td>\n",
       "      <td>0.003727</td>\n",
       "      <td>5.925138</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>Disney</td>\n",
       "      <td>Zootopia</td>\n",
       "      <td>2016</td>\n",
       "      <td>adventure,animation,comedy</td>\n",
       "      <td>spyblgbazootopia written jared bush phil johns...</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.081025</td>\n",
       "      <td>0.092248</td>\n",
       "      <td>0.011222</td>\n",
       "      <td>5.872771</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>Disney</td>\n",
       "      <td>Beauty and Beast</td>\n",
       "      <td>2016</td>\n",
       "      <td>family,fantasy,musical</td>\n",
       "      <td>mgncbeauty written stephen chbosky evan spilio...</td>\n",
       "      <td>9.67</td>\n",
       "      <td>0.073202</td>\n",
       "      <td>0.093116</td>\n",
       "      <td>0.019915</td>\n",
       "      <td>6.004309</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>Disney</td>\n",
       "      <td>ROMEO &amp; JULIET</td>\n",
       "      <td>1995</td>\n",
       "      <td>drama,romance,comedy</td>\n",
       "      <td>sacwilliam shakespeare adapted craig pearce ba...</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.099545</td>\n",
       "      <td>0.090849</td>\n",
       "      <td>-0.008696</td>\n",
       "      <td>5.981083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>Disney</td>\n",
       "      <td>Frozen</td>\n",
       "      <td>2013</td>\n",
       "      <td>adventure,animation,comedy</td>\n",
       "      <td>quhfrozen written jennifer lee final shooting ...</td>\n",
       "      <td>10.0</td>\n",
       "      <td>0.087554</td>\n",
       "      <td>0.101502</td>\n",
       "      <td>0.013948</td>\n",
       "      <td>5.924542</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     Type                 movie  Year  \\\n",
       "0      QT            Four Rooms  1995   \n",
       "1      QT  Inglourious Basterds  2008   \n",
       "2      QT   From Dusk Till Down  1996   \n",
       "3      QT  Natural Born killers  1995   \n",
       "4      QT      Django Unchained  2012   \n",
       "5      QT          Pulp Fiction  1993   \n",
       "6      QT          True Romance  1993   \n",
       "7      QT        Reservoir Dogs  1990   \n",
       "8      QT          Jackie Brown  1997   \n",
       "9      QT             Kill Bill  2003   \n",
       "0  Disney               Aladdin  1992   \n",
       "1  Disney          Finding Nemo  2003   \n",
       "2  Disney          Mary Poppins  1964   \n",
       "3  Disney                 Mulan  1998   \n",
       "4  Disney    The Little Mermaid  1989   \n",
       "5  Disney          The Pacifier  2004   \n",
       "6  Disney              Zootopia  2016   \n",
       "7  Disney      Beauty and Beast  2016   \n",
       "8  Disney        ROMEO & JULIET  1995   \n",
       "9  Disney                Frozen  2013   \n",
       "\n",
       "                                            Genres  \\\n",
       "0                                     comedy,drama   \n",
       "1                             action,adventure,war   \n",
       "2                    action,comedy,horror,thriller   \n",
       "3                    action,romance,thriller,crime   \n",
       "4                          adventure,drama,western   \n",
       "5                      action,crime,drama,thriller   \n",
       "6                    action,romance,thriller,crime   \n",
       "7                            action,crime,thriller   \n",
       "8                                     comedy,crime   \n",
       "9               action,comedy,crime,drama,thriller   \n",
       "0          family,animation,comedy,musical,fantasy   \n",
       "1                family,animation,comedy,adventure   \n",
       "2    family,animation,comedy,musical,fantasy,drama   \n",
       "3        family,animation,comedy,musical,adventure   \n",
       "4  family,animation,comedy,musical,romance,fantasy   \n",
       "5                    comedy,action,family,thriller   \n",
       "6                       adventure,animation,comedy   \n",
       "7                           family,fantasy,musical   \n",
       "8                             drama,romance,comedy   \n",
       "9                       adventure,animation,comedy   \n",
       "\n",
       "                                              Script Rating  neg_score  \\\n",
       "0  four rooms screenplay allison anders alexandre...   8.63   0.090876   \n",
       "1  jppinglourious basterds written quentin tarant...   7.44   0.077510   \n",
       "2  rkfrom dusk till dawn screenplay quentin taran...   7.25   0.113527   \n",
       "3  mewsijjnatural born killers written quentin ta...   8.52   0.098840   \n",
       "4  unchained written quentin tarantino ext countr...   7.82   0.089610   \n",
       "5  vbw pulp fiction quentin tarantino roger avary...   9.39   0.094869   \n",
       "6  cdameddlucy laughs well enough king bout bout ...   9.88   0.083703   \n",
       "7  quentin tarantino r e e r v r g october dedica...   8.89   0.102989   \n",
       "8  ldajackie screenplay quentin tarantino opening...   8.27   0.079589   \n",
       "9  hodsover black hear labored breathing black fr...   8.73   0.090782   \n",
       "0  gsavjgecaladdin complete script compiled ben s...   9.24   0.089018   \n",
       "1  cntan finding transcript v copyright walt disn...   9.20   0.090095   \n",
       "2  abhap right ladies gents comical poems suitabl...   9.38   0.073023   \n",
       "3  chzgrmmshhycdisney compiled adams theater show...   9.82   0.082271   \n",
       "4  shthe little mermaid ocean birds flying porpoi...   9.73   0.099754   \n",
       "5  cthe pacifier fade ext pacific ocean day fishi...    6.0   0.078187   \n",
       "6  spyblgbazootopia written jared bush phil johns...   10.0   0.081025   \n",
       "7  mgncbeauty written stephen chbosky evan spilio...   9.67   0.073202   \n",
       "8  sacwilliam shakespeare adapted craig pearce ba...   10.0   0.099545   \n",
       "9  quhfrozen written jennifer lee final shooting ...   10.0   0.087554   \n",
       "\n",
       "   pos_score  sentiment_score_iv  sentiment_score_anew  \n",
       "0   0.093699            0.002823              5.878314  \n",
       "1   0.092213            0.014703              5.646904  \n",
       "2   0.080819           -0.032708              5.639600  \n",
       "3   0.086155           -0.012686              5.310696  \n",
       "4   0.091363            0.001753              5.594078  \n",
       "5   0.089338           -0.005531              5.749571  \n",
       "6   0.099120            0.015417              5.883581  \n",
       "7   0.093266           -0.009723              5.462150  \n",
       "8   0.083852            0.004262              5.778873  \n",
       "9   0.077276           -0.013506              5.655627  \n",
       "0   0.105943            0.016925              5.900127  \n",
       "1   0.109044            0.018949              6.036038  \n",
       "2   0.147867            0.074844              6.217262  \n",
       "3   0.087397            0.005126              5.921347  \n",
       "4   0.121891            0.022137              6.156296  \n",
       "5   0.081914            0.003727              5.925138  \n",
       "6   0.092248            0.011222              5.872771  \n",
       "7   0.093116            0.019915              6.004309  \n",
       "8   0.090849           -0.008696              5.981083  \n",
       "9   0.101502            0.013948              5.924542  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "disney_df.insert(0,'Type','Disney')\n",
    "QT_df.insert(0,'Type','QT')\n",
    "all_df = QT_df.append(disney_df)\n",
    "all_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_df.to_csv('NEW_dataframe.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hypothesis test1: Tarantino’s movies will have a higher negative sentiment rating than that of Disney.\n",
    "\n",
    "Null hypothesis: 𝜇1 − 𝜇2  =  0\n",
    "\n",
    "Alternative hypothesis: 𝜇1 − 𝜇2 > 0 @ 𝛼 = 0.05\n",
    "\n",
    "Where 𝜇1 stands for the mean negative score for QT; 𝜇2 stands for the mean negative for Disney."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from scipy import stats\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a20214290>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# distribution plot of the negative score\n",
    "sns.distplot(QT_df['neg_score'])\n",
    "sns.distplot(disney_df['neg_score'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LeveneResult(statistic=0.0036038653544321012, pvalue=0.9527914891381134)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.levene(QT_df['neg_score'],disney_df['neg_score']) # p-value > 0.05, we regard it has the queal variance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ttest_indResult(statistic=1.4900812759200774, pvalue=0.15351758148378744)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats.ttest_ind(QT_df['neg_score'],disney_df['neg_score']) # pvalue is 0.15, larger than 0.05, we fail to reject the null hypothesis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>statistic</th>\n",
       "      <th>pvalue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>Levene</td>\n",
       "      <td>0.00362</td>\n",
       "      <td>0.9528</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>Ttest</td>\n",
       "      <td>1.4901</td>\n",
       "      <td>0.1535</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       statistic  pvalue\n",
       "Levene   0.00362  0.9528\n",
       "Ttest     1.4901  0.1535"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd \n",
    "from pandas import Series,DataFrame\n",
    "data_p = {'statistic':['0.00362','1.4901'],\n",
    "        'pvalue':['0.9528','0.1535']\n",
    "       }\n",
    "p = DataFrame(data_p, index = ['Levene','Ttest'])\n",
    "p"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hypothesis test2: Overtime, the total sentiment score rating has increased for both movie categories.\n",
    "### Hypothesis test3: The sentiment score calculated by \"Harvard IV\" and \"ANEW\" has the same trend."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"NEW_dataframe.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Rating')"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6wAAAHnCAYAAAC8O9DCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOzdd3hUZf7+8fcnjS4gIGCJiAWwUKQrCaEK0gVFEAWVEsi44rrq17Xsrt/Vn6tfK1WqqIggSJG2AlICiDRpVhAVQRSVTjCQ5Pn9MQOGkMCgSc5kcr+ua67JnHPmnHtCUO485zzHnHOIiIiIiIiIhJoIrwOIiIiIiIiIZEeFVUREREREREKSCquIiIiIiIiEJBVWERERERERCUkqrCIiIiIiIhKSVFhFREREREQkJKmwiohI2DIzZ2ZLvM4R6sysSuB79Xo+Ha9P4Hh98uN4IiJScKmwiogUAmZWz8zGm9l2MztqZgfNbKOZ/cfMKmXZNiFQJs7lUeUsx/9npm0nnGG7ppm2+zZXPnyIMrMbM33W/l7nERERCUVRXgcQEZG8Y2YGPAs8DKQBC4B3gRjghsDyQWbWwzk3O/C2b4F/ZdlVGeB+4ADwcjaH2h9kpDTgVjO73zmX3Xv6BbbJrf8/1QBScmlfue1ESXWBr0d5mCW/TQdWAbu9DiIiIqFNhVVEJLw9gb+Ufgu0d859mnmlmXUF3gLeM7M459zHzrlvgX9m2a4K/sK63zl3yrpzNBvoDNwBDMtyjLJAV+B9oMufOMZJzrkvcmM/uc3MygC3AluBzcAtZna9c269t8nyh3PuAP5ffoiIiJyRTgkWEQlTgZL5BHAc6Ji1rAI456YBDwDRwGv5EGs+sBP/SGpWdwJFgdE5vdnMIsws0czWmNlhMzsS+HqgmZ32/7Ss17Ca2WuBZR1z2H+jwPp3sywvbmaPmtmGwDEPm9lHZtYjuI99ml5AMeD1wAOy/55kPkX7n2ZW28zmmNl+M0sxs6VmdkM277nQzJ40sxVm9qOZHTOzH8zsbTOrEUxAM3sncNz4HNZ3C6wfkmlZVTMbZWbbAqee7zWzzWY20szKZdou22tYzaymmU0ys2/NLNXMfjaz9Wb2splFB5NbRETCiwqriEj4uhv/mTTTnXObz7DdGOAHoJaZNcrjTOnAuMCx6mVZ1w//SPDCM7z/TWAEUBF/7lFABWB4YN3ZvB547p3D+rsCzyevsw2Mhi4HnsmUf0LguG+b2b+DOG5W/YAM4A1gHvAT0NPMSpzhPfWAlfhL/Rj8o9VNgEVmVi3LtvHA/+A/VXsa8BL+U3C7AWvMrFYQGYcHngfksP7EKc2jAMysMrAG/8/dp8Cr+P9MvsH/y4jKZzqYmdUEPgY6BbK+CEwBfgYGAUWCyCwiImFGpwSLiISvJoHnMxVAnHNpgVHInviLzqo8zjUWeBx/aVsL/pFN4NrAcpfdmwKjmT2BT4B459zhwPLHgaX4C98c59zbOR3YOfeRmX0FtDezcs65XzPtvwhwO7AH/0jwCS8DdYBHnHPPZdq+KDAD+LuZTXXObQjmwwc+a03gA+fczsCyicBfA8cfm8Nb2wF3O+dez7SvAcBI/KdrD8q07YdARefcoSzHrgWswH9dc9sz5XTOLTOzT4GugWuOf8m0n8uAlsDKTL8M6QacDwx2zr2S5bgl8Bf0M+mNv4x3ds7NzPL+soTutcgiIpKHNMIqIhK+ToxofR/Etie2uTiPspzknNsBfAD0yDSi2A//6OX4M7z1nsDz/5woq4H9HQEeCbzsG0SECfgnnbo9y/IOQFlgonMuDSBwGmsvYG3msho47m+B4xr+Ih2sEyOTr2daduJzZ3tacMCKzGU1YBz+SaoaZMm2J2tZDSzfiL/MNgvyFNsR+Ec2s45I98f/ubM7jfxoNsc94pw7bXkOsnv/Pufc2QqviIiEIRVWEZHwZYHnbEcsc9i2aB5lyWo0UAq43czOA7oDc5xzP5zhPdfjH6Vbks26pfgLb50gjv1GYD9ZS9iJ15lvu1MfiAROXEN6ygO4LbBdsNeFnhd4zwH8M+UC4JzbAqwHGgZOjc3O2qwLnHPH8Z9OXDabY7Uzs/fNbLeZHT9xCx38xbwIUD6IyG8Ah/m9ZBMoun2AffhP2T1hVmDbYWY2zcz6m9k1gZmqgzEZ/5/hDDN7w8zuMrPLg3yviIiEKZ0SLCISvnYD1YHYILY9MbL6c97FOcUs4Ef8I6LRQAnOMNlSQGlgr3PuWNYVgdOafwEuONuBnXM7zWwR0MrMajjnPjezC4A2wIbAKOQJJyYKqh945KTk2Y4bcAf+z/paYIQ2s/H4S3l/wJfNe3O6dVAa/lJ9kpn9BXgFf6lcAOzAf0qtwz9Lcy2CuCbUOXfIzN4CEs2smXNuMf5rTCsBL2f+DM6578ysAf4ZptsAtwRWfW9m/+ece/Usx1ptZnHAY/hPL74z8Fm+BP7lnJt0trwiIhJ+NMIqIhK+lgeeW55pIzOLBBICL9flZaATAqfcvg40wl9QduKffOhMDgDnZ3cqq5lF4R8xPBhkhBOjqCdGVe/A/0vcCVm2O3HrlZecc3aGR7Mgj3vilN8BJ0Y8M418nphtt5eZFQtyf6cJfC/+hf8XAtc457o75x5yzv0jcEuin85xlyNOZA48nzLZUmbOuc+dc93xF/16+Cd+igBeMbN7z3Yg59xHzrn2+EeMbwT+F/8EW2+b2Rl/jkVEJDypsIqIhK8T1zd2MbNrzrDdPcCFwF5OnWwor43BP+J3MTDOOZd+lu0/wf//rexusxKPf5Qx2PuYvoe/3PYy/+1weuP/XmWdsGk1/tOH44Lcb44CsyLXwT8j89gcHpvwjyTflsNuglEeKIN/QqTdWTKUxD+KGzTn3Cb8EzV1MbOG+H8Bssw59/kZ3pPmnFvnnPsPcOLWP53P4ZipzrmVzrkngb8EFnc6l9wiIhIeVFhFRMKUc+5b4N/4T7mdZWZXZ93GzDrjP3UU/LPg5ttMrM65r/GfOtoF/y1QzmZc4Pn/mVnxEwsDXz8beJnTDLtZj30U//WXF+G/D20tYK5zbk+W7fYAE4F6ZvZEYPTyFGZ2eWDW3LM5MTL5inOub3YP/DMFZ972j9iD//TfuoGCeiJnNP4/62CuXc1qBP6Jqqbhv955ZNYNzKyBmVXM5r0nlp3xZ8vM4sys9B99v4iIhCddwyoiEt6ewn/N5EPARjP7L/57ZEYDNwANA9s955wbk9/hnHMfnMO2b5tZJ/yjj5+a2Qx+vybzMmCKc27iORx+Av5raP9fptfZ8QFX4v9e3mlmy/GfVnsh/smW6uMfRfwmpwMFimMP/KO4OR0H/DP4bgduMLNrnHOfBv1pApxzGWb2Kv7TcTeb2Uz8ZbMZ/tvOLA58fS7exX8v14uAX/CPUGfVE0gys6XANvzXz16Of5KnVPy3BzqTB4HWgVssbcc/gdM1+G+/s49sTkEWEZHwp8IqIhLGnHMOeNjM3gWSgKb4T+k8MeHObuAu59wZ79UaQnrgnxH4Hn6/pvJz4AV+v9YyKM655Wa2DbgC/+nQs3PY7qCZNcU/6tkT6Ip/NuWfgK34R2gXBJG7JDDdOZfjNaTOOWdmY4GnA8e7/1w+UyZP4J9Aqy/+79OBQMbH8V/fek6cc8cC94odDLzunEvNZrNJ+H+ubsB/2nExYBfwDvBCYCbkMxmOv5g2xH/9ahT+a5uHB97/3bnmFhGRgs/8/5YREZHCxMxK4Z+U6WrgVufcDI8jSYgLjHzGA9Wcc1s9jiMiIoWErmEVESmEnHOHgPb4R+Emm1kbjyNJCAvcrqYp8F+VVRERyU8aYRURKcTMrBb+SY9S8N9X87R7nErhZWYD8V+3ejf+e682ds6t9jaViIgUJiqsIiIiki0z+xb/bYe2A/90zmW97Y+IiEie0inBIiIiki3nXBXnXJRz7iqVVRGRgsvMxpnZHjPbkmnZ+Wa2wMy2Bp7L5vDe3oFttppZ7/xL7afCKiIiIiIiEt5ex3/v88z+B1jknLsSWBR4fQozOx/4B/4Z3BsA/8ip2OaVAn9bm4iICFesWDGvY4iIiIiIiHgiJSXFOedyHIx0zi0zsypZFncCEgJfTwCWAI9k2eYmYIFzbi+AmS3AX3wn/enQQSrwhbVYsWIcOXLE6xgiIiIiIiKeMLOjf+BtFZ1zuwGcc7vN7IJstrkI+D7T652BZfmmwBdWERERERGRQi7KzNZmej3KOTcqF/Zr2SzL11l7VVhFREREREQKtjTnXL1zfM9PZlY5MLpaGdiTzTY7+f20YfDPHL/kj0X8YzTpkoiIiIiISOEzCzgx629vYGY22/wXaG1mZQOTLbUOLMs3KqwiIiIiIiJhzMwmAR8B1cxsp5ndCzwLtDKzrUCrwGvMrJ6ZjQEITLb0v8CawOOpExMw5Vt25/L1FORcV6JECadJl0REREREpLAysxTnXAmvc+QFjbCKiIiIiIhISFJhFRERERERkZCkwioiIiIiIiIhSYVVREREREREQlLIFVYzK2NmU83sCzP73Mwae51JRERERERE8l+U1wGy8Qow3znXzcxigOJeBxIREREREZH8F1KF1czOA+KBPgDOuWPAMS8ziYiIiIiIiDdC7ZTgqsDPwHgz+8TMxphZWN5PSERERERERM4s1AprFHA9MMI5Vwc4AvxP1o3MrL+ZrTWztWlpafmdUSQsHDhwgC5dunDgwAGvo4iIiIiIZCvUCutOYKdz7uPA66n4C+wpnHOjnHP1nHP1oqJC6qxmkQJj1qxZzJgxg/fff9/rKCIiIiIi2Qqpwuqc+xH43syqBRa1AD7zMJJI2Bo3btwpzyIiIiIioSakCmvAfcBEM9sE1Aae8TiPSFho2bIlZnbysXLlSgBWrFhxyvKWLVt6nFRERERExM+cc15n+FNKlCjhjhw54nUMkZC3ePFi2rdvT0pKSo7bFC9enDlz5pCQkJB/wURERETkTzGzFOdcWE5Wq8IqEm5e+1uOqxZ/uYP2w6aTcuz0ycqKx0QxJ+kWEqpdkvO+B/xfbiQMXWf43uWKcP/+iYiIiCfCubCG4inBIpJHmlxxEQ2rVD5tedGoSCb3bX/msioiIiIiks9UWEUKiaPHjtP1tVks/up7YiIjiDCjSFQkAA7YfzTV24AiIiIiIlmosIoUAgePptJ2yHvM3rydKy8oQ1pGBrUursCsQZ0pFh1Falo641Zs8TqmiIiIiMgpdBNTkTD386EU2g55j407f2biPTczec2XJMbXYnDzukREGEO6N6fvWx9wLD3d66giIucuL68913XnIiKeU2GV0KOJb3LN93sP0uqVqXy39xAzB3bi5uuq0qN+jVO26dmgOg9PX0al88LyOn0RERERKcB0SrBImPryx73c+Pw77D5whA/+0pWbr6ua7XbFYqLpe+N1zNi4je/3HsznlCIiIiIiOVNhFQlD63f8RNwL7/BbWhpL/nobcVdefMbtB8bXIsM5XkvelE8JRURERETOToVVJMws27qTZi9OoVh0FMv/djt1Yiue9T1Vypemw3WXM2r5JlKPn36PVhERERERL6iwioSROXPmcNOr07iwTElWPNSDqyqeH/R7fQm1+fnQUd5d/1UeJhQRERERCZ4mXRIJExMnTqRPnz7Uvqgc8+67hfIli5/T+1tUv5RqFcsydMkGejW8Oo9SishpNNGceEU/e3+cvnci+UYjrCJhYNiwYfTq1YsmTZqwaPCt51xWASIijKSE2nz8zW7WfPtjHqQUERERETk3GmEVKcCcc/z73//mySefpGPHjkyePJmiEx7/w/vr3ega/j5jOcOWbOD1Pm1yMamIiIgIGp2Wc6YRVpECKiMjg7/+9a88+eST3HXXXUybNo2iRYv+qX2eV6wIdzW6mnfWfsHPh1JyKamIiIiIyB+jwipSAKWlpXHPPffw8ssvc//99zN+/HiionLnhImkpnVITUtn7IrNubI/EREREZE/SoVVpID57bff6NatGxMmTOCpp57ipZdeIiIi9/4qX31hOZpXi2XEso2kpWfk2n5FRERERM6VCqtIAXLo0CFuvvlmZs6cyZAhQ3jiiScws1w/ji+hNjv2HmL25q9zfd8iIiIiIsFSYRUpIH755ReaN2/OsmXLeOutt/D5fHl2rA41L+eSsqUYumRDnh1DRERERORsNEtwXtEMaJKLdu7cSevWrfnmm2+YPn06HTp0yNPjRUVGMLBpLf4+Yzmf7/6VGpXL5enxRERERESyoxFWkRC3detWmjRpws6dO5k/f36el9UT+t54HTFRkQzTKKuIiIiIeESFVSSEbdiwgSZNmpCSksKSJUto2rRpvh27Qqni3F6vGhNWfcrBo6n5dlwRERERkRNUWEVC1PLly0lISKBIkSIkJydz/fXX53sGX0IdDqce541Vn+X7sUVEREREVFhFQtDcuXNp3bo1lSpVYvny5VSrVs2THPWrVKJBlUoMXfIJzjlPMoiIiIhI4aXCKhJiJk2aRKdOnahRowbJycnExsZ6mseXUIcvf9rHoi92eJpDRERERAofzRIsEkJGjBhBUlIS8fHxzJo1i/POO8/rSNxa9yoenLaEoUs+oaXXYSR0aWZ0ERERyQMaYRUJAc45nnnmGQYNGkT79u2ZN29eSJRVgKLRUfS7sSbvb9rOt99+63UcERERESlEVFhFPOac46GHHuKxxx6jV69eTJs2jWLFinkd6xSJ8TUBGDlypMdJRERERKQwUWEV8VBaWhp9+/blhRde4L777mPChAlER0d7Hes0l5x/Hp1rX8GYMWM4evSo13FEREREpJBQYRXxSGpqKt27d2fcuHH84x//4JVXXiEiInT/SvoSavPrr78yefJkr6OIiIiISCERuv86Fgljhw4dol27drz33nu8/PLL/POf/8TMvI51RglXXcLVV1/NkCFDdIsbEREREckXKqwi+ezXX3+lZcuWLFmyhAkTJnD//fd7HSkoZobP52P9+vV8/PHHXscRERERkUJAhVUkH+3atYv4+Hg2btzIe++9x1133eV1pHNy5513ct555zF06FCvo4iIiIhIIaDCKpJPtm3bRpMmTdixYwfz5s2jY8eOXkc6ZyVLlqRPnz5MmTKFn376yes4IiIiIhLmVFhF8sGmTZto0qQJhw4dYvHixTRr1szrSH/YoEGDOH78OGPGjPE6ioiIiIiEORVWkTy2cuVKmjZtSnR0NMnJydSrV8/rSH9KtWrVaN26NSNGjCAtLc3rOCIiIiISxlRYRfLQ/PnzadmyJRUqVGD58uXUqFHD60i5wufzsWvXLmbOnOl1FBEREREJYyqsInlk8uTJdOzYkWrVqrF8+XIuvfRSryPlmptvvpkqVapo8iURERERyVMqrCJ5YNSoUfTo0YNGjRqxZMkSLrjgAq8j5arIyEgGDRrEkiVL2LJli9dxRERERCRMqbCK5LJnn32WAQMG0LZtW+bPn0/p0qW9jpQn7rnnHooWLcqwYcO8jiIiIiIiYUqFVSSXOOd4+OGHefTRR+nZsyczZsygePHiXsfKM+XKlaNnz5688cYb7N+/3+s4IiIiIhKGVFhFckF6ejr9+vXj+eefZ9CgQbz55ptER0d7HSvPJSUlkZKSwoQJE7yOIiIiIiJhSIVV5E9KTU3l9ttvZ+zYsTz++OMMHTqUiIjC8Vfr+uuv54YbbmDYsGFkZGR4HUdEREREwkzh+Fe1SB45fPgwHTp0YOrUqbz44ov87//+L2bmdax85fP52Lp1KwsWLPA6ioiIiIiEGRVWkT9o7969tGrVikWLFjFu3DgeeOABryN5omvXrlSsWFG3uBERERGRXKfCKvIH7N69m6ZNm7J+/XqmTp3K3Xff7XUkz8TExDBgwADmzJnD9u3bvY4jIiIiImFEhVXkHG3fvp0mTZrwzTffMHfuXLp06eJ1JM8NGDCAyMhIRowY4XUUEREREQkjKqwi52Dz5s00adKE/fv38+GHH9KiRQuvI4WECy+8kFtuuYWxY8eSkpLidRwRERERCRMqrCJBWrVqFU2bNsXMSE5OpkGDBl5HCik+n499+/YxadIkr6OIiIiISJhQYRUJwoIFC2jRogXlypVjxYoVXH311V5HCjlNmjShZs2aDB06FOec13FEREREJAyosIqcxdSpU2nXrh1XXHEFycnJVKlSxetIIcnM8Pl8bNiwgZUrV3odR0RERETCgAqryBmMGTOG7t2706BBA5YuXUqlSpW8jhTSevbsSZkyZXSLGxERERHJFSqsIjl4/vnn6devH61bt+aDDz6gTJkyXkcKeSVKlOCee+5h6tSp7N692+s4IiIiIlLAqbCKZOGc49FHH+Xhhx+me/fuzJw5k+LFi3sdq8AYOHAg6enpjBo1yusoIiIiIlLAqbCKZJKenk5iYiLPPvssiYmJTJw4kZiYGK9jFShXXHEFbdu2ZeTIkRw7dszrOCIiIiICmNn9ZrbFzD41s8HZrE8wswNmtiHweNKLnFmpsIoEHDt2jJ49ezJq1Cj+/ve/M3z4cCIjI72OVSD5fD5+/PFHpk+f7nUUERERkULPzK4F+gENgFpAezO7MptNk51ztQOPp/I1ZA5UWEWAI0eO0LFjR6ZMmcLzzz/P008/jZl5HavAuummm7j88ss1+ZKIiIhIaKgBrHLOpTjn0oClQBePMwVFhVUKvX379tG6dWsWLFjAmDFj+Nvf/uZ1pAIvIiKCQYMGsXz5cjZs2OB1HBEREZHCbgsQb2blzKw4cDNwSTbbNTazjWY2z8yuyd+I2VNhlULtxx9/pGnTpqxdu5YpU6Zw7733eh0pbNx9990UK1aMYcOGeR1FREREJNxFmdnaTI/+mVc65z4H/gMsAOYDG4G0LPtYD1zqnKsFDAFm5EPus1JhlULrm2++oUmTJmzfvp05c+bQtWtXryOFlbJly9KrVy8mTpzI3r17vY4jIiIiEs7SnHP1Mj1Ou12Dc26sc+5651w8sBfYmmX9Qefc4cDXc4FoMyufL+nPQIVVCqVPP/2UG2+8kb1797Jo0SJatmzpdaSwlJSUxNGjRxk/frzXUUREREQKNTO7IPAcC9wCTMqyvpIFJnExswb4u+Kv+Z0zKxVWKXQ+/vhj4uPjAVi2bBkNGzb0OFH4qlWrFnFxcQwfPpz09HSv44iIiIgUZtPM7DPgfSDJObfPzBLNLDGwvhuwxcw2Aq8CtzvnnFdhT1BhlUJl4eff0aJFC8qUKcPy5cu59tprvY4U9nw+H9u3b2f+/PleRxEREREptJxzcc65q51ztZxziwLLRjrnRga+HuqcuyawvpFzbqW3if1UWKXQeO+TrbQbNp2qVauyfPlyqlat6nWkQqFLly5UrlxZt7gRERERkXOmwiqFwrgVm7l11PvUjb2ApUuXUrlyZa8jFRrR0dEkJiYyf/58tm7devY3iIiIiIgEqLBK2HthwVruffMDWtaIZcH9t1K2bFmvIxU6/fv3Jzo6muHDh3sdRUREREQKEBVWCVvOOR6bsZy/TVvKrddfxayBnSlRJNrrWIVSpUqV6NatG+PHj+fw4cNexxERERGRAkKFVcJSekYGgyYt4pn5H9OvyXVM6tuOItFRXscq1Hw+HwcOHGDixIleRxERERGRAkKFVcLOsbR0eo2by8hlG3nkpvq8dkcrIiP0o+61xo0bU6dOHYYOHUoIzJAuIiIiIgWA/hUvBdqBo6l0GTGTA0dTAUg5dpzOI2byztovebZLHM92iSdw/2PxmJnh8/nYsmULy5Yt8zqOiIiIiBQAKqxSoM3a+DUzNm7j/U1fsz/lN1q/MpX5n33DqDta8chNDbyOJ1n06NGD888/X7e4EREREZGgqLBKgTZu5WYARi7bSMKLU1j97Y9M7tuefnE1PU4m2SlWrBj33nsv06dPZ+fOnV7HEREREZEQp8IqBUrLl9/FEl84+Vj59Q8ArPj6Bzbu/Jnj6RncNno2LV9+1+OkkpOBAweSkZHBa6+95nUUEREREQlxKqxSoDzWtiHFY36f7fdYesZp2xSPieLxto3yM5acg8suu4z27dszatQoUlNTvY4jIiIiIiFMhVUKlGbVYpmd1OWU0ppZ8Zgo5iTdQkK1S/I5mZwLn8/Hnj17mDp1qtdRRERERCSEqbBKgdOsWiz/aNf4tOVFoyOZ3Le9ymoB0LJlS6666ipNviQiIiIiZ6TCKgXOws+/44lZKzAgwoxi0VFEmBEVEcH+ozrFtCCIiIggKSmJVatWsXbtWq/jiIiIiEiIUmGVAmXmhm20Gzad6MgIMKh1cQVmDuxErYsrkHLsOONWbPE6ogSpd+/elChRgmHDhnkdRURERERCVPYXAnrMzCKBtcAu51x7r/NIaJi05nPuHD+PurEVKVeiKC1rXMrg5nWJiDCaV4/l5UXrSN62y+uYEqTSpUtz1113MW7cOJ5//nnKly/vdSQRERERCTGhOsJ6P/C51yEkdIxO3sQd4+YSd8XFLBx8K3Pv68pfW9YjIsIAiIyI4MFW9ZkxsLPHSeVcJCUlkZqaytixY72OIiIiIiIhKOQKq5ldDLQDxnidRULDiwvX0n/iAtpcfRlzfV0oVTTG60iSS6655hqaNWvG8OHDSU9P9zqOiIiIiISYkCuswMvAw8DpN9iUQsU5x79mr+TBqUvpdv1VzBjYiWIx0V7Hklzm8/nYsWMHs2fP9jqKiIiIiISYkCqsZtYe2OOcW3eW7fqb2VozW5uWlpZP6SQ/Oed4aNpS/jn7I/o0voZJ97YjJirS61iSBzp27MjFF1+sW9yIiIiIyGlCqrACNwIdzexb4B2guZm9lXUj59wo51w951y9qKiQnDdK/oT0jAwS317ICwvX4Uuozdg7byIqMtR+VCW3REVFMXDgQBYuXMjnn+vSdRERERH5XUi1AOfco865i51zVYDbgQ+dc708jiX56Pjx49w1fh6jkjfxaJsGvNq9+cmJlSR89e3bl5iYGIYPH+51FBEREREJISFVWKVwS01N5dZbb+XtNV/wTOcmPNM5DjOV1cLgggsuoHv37rz++uscPHjQ6zgiIiIiEiJCtrA655boHqyFx5EjR+jQoQMzZynPt90AACAASURBVM5kSPfmPNqmodeRJJ/5fD4OHz7Mm2++6XUUEREREQkRIVtYpfA4cOAAN910E4sWLWL8+PH4mtXxOpJ4oEGDBtSvX5+hQ4finPM6joiIiIiEABVW8dQvv/xC8+bNWb16NZMnT6ZPnz5eRxIP+Xw+vvjiCz788EOvo4iIiIhICFBhFc/88MMPNG3alM8++4wZM2bQrVs3ryOJx2677TbKly+vW9yIiIiICKDCKh759ttviYuLY8eOHcybN4+bb77Z60gSAooWLUq/fv2YNWsW3333nddxRERERMRjKqyS77744guaNGnC3r17WbhwIQkJCV5HkhCSmJgIwMiRIz1OIiIiIiJeU2GVfLVx40bi4+M5fvw4S5cupWFDzQYsp4qNjaVTp06MHj2a3377zes4IiIiIuIhFVbJN6tWrSIhIYEiRYqwbNkyatas6XUkCVE+n49ff/2VyZMnex1FRERERDykwir54sMPP6Rly5aUK1eO5ORkqlWr5nUkCWHNmjWjRo0aDBkyRLe4ERERESnEVFglz82ZM4ebb76ZKlWqkJycTJUqVbyOJCHOzPD5fKxbt47Vq1d7HUdEREREPKLCKnlqypQpdO7cmWuvvZYlS5ZQuXJlryNJAXHnnXdSqlQp3eJGREREpBBTYZU8M27cOHr06EGjRo1YtGgR5cuX9zqSFCClSpWiT58+TJkyhZ9++snrOCIiIiLiARVWyROvvvoq9957Ly1btmT+/PmULl3a60hSAA0aNIhjx44xZswYr6OIiIiIiAdUWCXXPfPMM9x///106dKFWbNmUaJECa8jSQFVvXp1WrVqxYgRI0hLS/M6joiIiIjkMxVWyTXOOR599FEee+wxevXqxZQpUyhSpIjXsaSA8/l87Nq1i5kzZ3odRURERETymQqr5IqMjAzuu+8+nn32WRITE5kwYQJRUVFex5Iw0K5dOy699FJNviQiIiJSCKmwyp+WlpbGPffcw7Bhw/jb3/7G8OHDiYjQj5bkjsjISAYNGsSSJUvYsmWL13FEREREJB+pVcifcuzYMW6//XYmTJjAU089xXPPPYeZeR1Lwsy9995L0aJFGTZsmNdRRERERCQfqbDKH5aSkkKnTp2YNm0aL730Ek888YTKquSJcuXK0aNHD9544w3279/vdRwRERERyScqrPKHHDx4kLZt2/Lf//6X0aNHM3jwYK8jSZjz+XykpKQwYcIEr6OIiIiISD5RYZVztnfvXlq2bMnKlSt5++236du3r9eRpBC4/vrrady4McOGDSMjI8PrOCIiIiKSD1RY5Zz8+OOPNG3alE2bNvHee+9x++23ex1JChGfz8fWrVtZsGCB11FEREREJB+osErQduzYQXx8PNu3b2fOnDl06NDB60hSyHTr1o2KFSvqFjciIiIihYQKqwRl69atxMXFsWfPHhYsWECLFi28jiSFUExMDP3792fOnDls377d6zgiIiIiksdUWOWsNm/eTFxcHCkpKSxevJgbbrjB60hSiA0YMICIiAhGjBjhdRQRERERyWMqrHJGa9asISEhgcjISJYtW0adOnW8jiSF3EUXXcQtt9zC2LFjSUlJ8TqOiIiIiOQhFVbJ0bJly2jRogWlS5cmOTmZGjVqeB1JBPBPvrRv3z4mTZrkdRQRERERyUMqrJKt+fPn06ZNGy666CKSk5OpWrWq15FEToqLi+O6665j6NChOOe8jiMiIiIieUSFVU4zbdo0OnbsSLVq1Vi2bBkXXXSR15FETmFm+Hw+NmzYwMqVK72OIyIiIiJ5RIVVTvHGG29w2223Ua9ePRYvXkyFChW8jiSSrTvuuIPSpUvrFjciIiIiYUyFVU4aPnw4vXv3plmzZnzwwQeUKVPG60giOSpRogT33HMPU6dOZffu3V7HEREREZE8oMIqADz33HMkJSXRoUMHZs+eTcmSJb2OJHJWgwYNIi0tjVGjRnkdRURERETygAprIeec4/HHH+eRRx7h9ttvZ9q0aRQtWtTrWCJBueKKK2jbti2vvfYax48f9zqOiIiIiOQyFdZCLCMjg8GDB/P000/Tt29f3nrrLaKjo72OJXJOfD4fu3fvZvr06V5HEREREZFcpsJaSKWnp9OvXz9effVVHnjgAUaNGkVkZKTXsUTOWZs2bahataomXxIREREJQyqshdCxY8fo2bMn48aN48knn+SFF17AzLyOJfKHREREkJSURHJyMhs3bvQ6joiIiIjkIhXWQubo0aPccsstTJkyheeee45//etfKqtS4N19990UK1aMYcOGeR1FRERERHKRCmshcvjwYdq1a8fcuXMZMWIEDz30kNeRRHJF2bJl6dWrF2+99Rb79u3zOo6IiIiI5BIV1kJi3759tGrVimXLlvHGG2+QmJjodSSRXJWUlMTRo0cZP36811FEREREJJeosBYCe/bsoVmzZqxbt453332XXr16eR1JJNfVqlWLJk2aMGzYMDIyMryOIyIiIiK5QIU1zO3cuZP4+Hi++uor3n//fbp06eJ1JJE84/P52L59O/Pnz/c6ioiIiIjkAhXWMPb1118TFxfHDz/8wH//+19uuukmryOJ5KkuXbpQuXJl3eJGREREJEyosIapzz77jLi4OA4ePMiHH35IXFyc15FE8lxMTAwDBgxg3rx5bNu2zes4IiIiIvInqbCGofXr1xMfH49zjqVLl1KvXj2vI4nkm/79+xMVFcXw4cO9jiIiIiISMszsfjPbYmafmtngbNabmb1qZtvMbJOZXe9FzqxUWMPMihUraNasGSVKlCA5OZlrr73W60gi+apy5cp069aNcePGceTIEa/jiIiIiHjOzK4F+gENgFpAezO7MstmbYErA4/+wIh8DZkDFdYwsnDhQlq3bk2lSpVYvnw5V1xxhdeRRDzh8/k4cOAAEydO9DqKiIiISCioAaxyzqU459KApUDW2Vg7AW84v1VAGTOrnN9Bs1JhDRMzZ86kXbt2XH755SxbtoxLLrnE60ginrnhhhuoXbs2Q4cOxTnndRwRERERr20B4s2snJkVB24GshaGi4DvM73eGVjmKRXWMDBp0iS6du1K7dq1WbJkCRUrVvQ6koinzAyfz8fmzZtJTk72Oo6IiIhIXosys7WZHv0zr3TOfQ78B1gAzAc2AmlZ9mHZ7Nfz3/yrsBZwo0eP5o477iAuLo6FCxdy/vnnex1JJCT06NGDsmXL6hY3IiIiUhikOefqZXqMyrqBc26sc+5651w8sBfYmmWTnZw66nox8EPeRQ6OCmsB9uKLL9K/f3/atGnD3LlzKVWqlNeRREJG8eLFuffee3nvvffYtWuX13FEREREPGVmFwSeY4FbgElZNpkF3BWYLbgRcMA5tzufY55GhbUAcs7xr3/9iwcffJBu3boxY8YMihUr5nUskZAzcOBAMjIyeO2117yOIiIiIuK1aWb2GfA+kOSc22dmiWaWGFg/F9gObANGA4M8ynmKKK8DyLlxzvHQtKW8sHAdffr0YfTo0URF6Y9RJDtVq1alXbt2vPbaazz22GMUKVLE60giIiIinnDOxWWzbGSmrx2QlK+hgqAR1gIkPSODxLcX8sLCdfh8PsaOHauyKnIWPp+PPXv2MG3aNK+jiIiIiMg5UmEtII6np3PX+HmMSt7Eo20a8OqrrxIRoT8+kbNp1aoVV155pSZfEhERESmA1HgKgNTjadw66n3eXvMFz3RuwjOd4zDLbtZpEckqIiKCpKQkPvroI9atW+d1HBERERE5ByqsIe5I6nE6DJ/BzI1fM6R7cx5t09DrSCIFTu/evSlRogTDhg3zOoqIiIiInAMV1hB24GgqN706lUVf7GD8XTfha1bH60giBVKZMmW48847efvtt/n111+9jiMiIiIiQVJhDVG/HE6h+UtTWP3tj0zu254+N1zrdSSRAi0pKYnU1FTGjh3rdRQRERERCZIKawj6Yf9hmr4whc9272VGYie61b3K60giBd61115LQkICw4cPJz093es4IiIiIhIEFdYQcOBoKl1GzOTA0VS+/eUAcf/3Djv2HmSe7xZuvq6q1/FEwobP5+O7775jzpw5XkcRERERkSCosIaAWRu/ZsbGbby2bCNN/u8d9qWksnDwrSRUu8TraCJhpVOnTlx88cW6xY2IiIhIAaHCGgLGrdwMwOOzVnA8PYMlf72NhpdV9jiVSPiJiooiMTGRBQsW8MUXX3gdR0RERETOIsrrAIVRy5ffZdEXO06+jo70/97geHoGew6lUOvfbwDQonosCwff6klGkXDVr18/nnrqKYYPH86rr77qdRwREREROQONsHrgsbYNKR7z++8KjqdnnLZN8ZgoHm/bKD9jiRQKF1xwAbfddhuvv/46hw4d8jqOiIiIiJyBCqsHmlWLZXZSl1NKa2bFY6KYk3SLrmEVySM+n49Dhw7x5ptveh1FRERERM5AhdUjzarFMrlve4pGR56yvGh0JJP7tldZFclDDRo0oF69egwdOhTnnNdxRERERCQHKqwe2n80laiICCLMKBYdRYQZURER7D+a6nU0kbBmZvh8Pj7//HMWL17sdRwRERERyYEKq4fGrthMyrHj1Lq4AjMHdqLWxRVIOXaccSu2eB1NJOx1796dcuXK6RY3IiIiIiFMswR7qHTRIjzftSmDm9clIsJoXj2WlxetI3nbLq+jiYS9okWL0q9fP5577jl27NhBbGys15FEREREJAuNsHpoxqDO/LVlPSIiDIDIiAgebFWfGQM7e5xMpHBITEwEYOTIkR4nEREREZHsqLCKSKF16aWX0rFjR0aPHs1vv/3mdRwRERERySLowmpmNc1sspl9bWapZnZ9YPnTZtY27yKKiOQdn8/HL7/8wpQpU7yOIiIiIiJZBFVYA4V0HVAJeAOIzrQ6Fbgv96OJiOS95s2bU716dU2+JCIiIhKCgh1h/X/A6865psDTWdZtAGrnaioRkXxy4hY3a9asYfU3u72OIyIiIiKZBFtYqwOTA1+7LOsOAufnWiIRkXx21113UapUKYYu2eB1FBERERHJJNjCugeomsO6a4AduRNHRCT/lSpVit69ezN53ZfsOZjidRwRERERCQi2sL4DPGVmTTItc2Z2FfAIMDHXk4mI5KOkpCSOpaUzZsVmr6OIiIiISECwhfUJYC2wlN9HU2cCW4BNwDO5H01EJP9Ur16dltVjGbFsA2npGV7HERERERGCLKzOuVTnXHugNTABGAO8DbRzzrV3zh3Pw4wiIvnCl1CHnfsOM2vT115HEREREREg6mwbmFkR4G/AbOfcImBRXoUxs0vw3zanEpABjHLOvZJXxxMRyax9zarEnl+KoUs+4ZY6V3odR0RERKTQO+sIq3MuFXgMKJP3cUgDHnTO1QAaAUlmdnU+HFdEhMiICAY1rc3iL7/n0x9+8TqOiIiISKEX7DWsHwN18zIIgHNut3NufeDrQ8DnwEV5fVwRkRPuvfFaikRFMky3uBERERHxXLCF9WFgoJn5zKyqmZUws+KZH7kdzMyqAHXwl+Ws6/qb2VozW5uWlpbbhxaRQqx8yeL0qF+dNz7+jANHU72OIyIiIlKoncsI6+XAq8BW4CBwKMsj15hZSWAaMNg5dzDreufcKOdcPedcvaios16GKyJyTnwJdTiSepwJH33qdRQRkZBw4GgqXUbM1C/yRCTfBdv27gFcXgY5wcyi8ZfVic659/LjmCIimdW9tCKNLqvMsKUb8CXUISLCvI4kIuKpWRu/ZsbGbby/6Wt6NdT0IiKSf4IqrM651/M4BwBmZsBY4HPn3Iv5cUwRkez4EurQa/xcFn7xHa2vruJ1HBERT41budn/vGKLCquI5KtzOp/WzC4EGgPnA3uBj5xzP+RinhuBO4HNZnZixpO/O+fm5uIxRETOqtv1V/LXqcUZuuQTFVYRKXRavvwui77YcfJ1TKT/KrIVX+/CEl84ubxF9VgWDr413/OJSOERVGE1s0hgCNAPiMy0Kt3MRgH3Oecy/mwY59xyQOfeiYjnikRH0T+uJk/PW8U3vxzgsvKlvY4kIpJvHmvbkI+2/0DKMf/klsfSM055BigeE8XjbRt5kk9ECo9gJ136F/7rWP8OVAGKBZ7/Hlj+z9yPJiLirQFxNYkwY8RS3eJGRAqXZtVieX9QZ6Iisv+nYvGYKOYk3UJCtUvyOZmIFDbBFta7gMedc88753Y451IDz88DTwB98iyhiIhHLi5bii61r2TMii2kHDvudRwRkXxzLC2d1z/6jLSMDCLt1JPfikZHMrlve5VVEckXwRbWC4BNOazbFFgvIhJ2fAm12ZfyG++s+cLrKCIi+WJ/ym+0GTKNNz/+jK51rqRYTBQRZiev2YqKiGC/bm8jIvkk2ML6FXB7DutuB77MnTgiIqEl/sqLufbC8gxZ8gnO5cvdvUREPPPdrwe58fl3WL5tF2/e3ZZfjxwl5dhxal1cgTsb+WcHPnLsOONWbPE4qYgUFsHOEvxv4B0ziwWmAj/hH1W9FWhGzmVWRKRAMzN8CbVJfHshH23/gRsuv8jrSCIieWL9jp9oN3Q6R4+n8d+/dKVZtVimrvuK57s2ZXDzuvx06AgTP/6cG6+4iPOKxXgdV0RCnJmNO8PqDOAgsAF4zzl3OKcNg70P6xQz249/8qVXgGjgOLAOaOOcWxBscBGRguaOBjV4ZHoyQ5dsUGEVkbA0d/N2bhszm3IlirJw8O1cc2F5AGYM6nxym8qlS9KlzpV8+OUOdj07wKuoIlJwXAdcgn+g8yfgZ6ACUBHYAxwAfMDTZtbCOfdVdjsJ9pRgnHMfOOca458huBJQzDl3g8qqiIS7kkVjuLvxNby77it2H8jxF4AiIgXSa8s20mH4DKpVLMuqR3qeLKvZSYyvyd4jv/Huumz/XSkiktmTwH6goXOusnOupnOuMtAIf1l9CKgGHAKez2knQRVWMytlZpUBnHMZzrk9J+67amaVzazkn/ssIiKhbVDT2qRlZDB6+Wavo4iI5IqMDMf/TF9G4tsLaXNNFZb+tTuVS5/5n3TNq8VyVcWyjFim232JyFk9B/zDObcm80Ln3Gr8t0X9j3PuG+BZID6nnQQ7wjoWeCqHdf8ExgS5HxGRAunKimVpc00VRi7byPH0dK/jiIj8Kampqdwxbg7/+e8aEuNrMXNgZ0oWPft1qWZGYlwtPtq+m4079+RDUhEpwK4AjuawLgWoEvj6O6BITjsJtrDGA3NyWDeXMzRiEZFw4Uuow+4DR5j+yTavo4iI/GF79+6lVatWvLP2S/7TJY7hPVoQFRn0VWL0bnw1RaOjGLkspzseiogA8AnwDzOrlHlh4Mzdf+CfDwngUuCHnHYS7H+dSuNvwdn5DSgb5H5ERAqsNtdUoWr50gxd8gkHjqbSZcRMDuhehCJSgGzfvp0bbriBjz/+mHf6tuPhmxpgZmd/YybnlyhG97rVeOvjzzj027E8SioiYSARuBD41sxWmNkMM1sBfIt/TqSBge0uBEbntJNgC+tWoF0O624Gvg5yPyIiBVZkRASDmtYmedsuhi7+hBkbt/H+Jv3nT0QKhtWrV9O4cWP27NnDwoUL6V6v+h/eV2J8TQ6nHmfi6s9zMaGIhBPn3CagKvAg8BX+036/Ah4ALnfObQ5s96xz7j857SfYwjoE8JnZ82Z2jZmdH3h+DkjCf6sbEZGwd/cN11AsOophS/wTjoxbscXjRCIiZzdz5kwSEhIoUaIEK1euJC4u7k/tr+Fllal9SQVGLN2Icy6XUopIuHHOHXXODXPO3e2caxt4Hu6cy+na1tMEex/W0WZWEXgU+GumVb8BjzvnchzCFREJBy1ffpdFX+w4+fro8TQAVny9C0t84eTyFtVjWTj41nzPJyKSk6FDh/KXv/yFevXq8f7771OxYsU/vU8zY2B8bQZMXMCqb3bTuOqFuZBURMKVmUUBp83s5pzL6bLTk4IqrIGd/dvMhgCNgXLAr8BHzrkD55BVRAqwemvzdv9r6+Xt/v+Mx9o25KPtP5ByLO2U5cfSM05+XTQ6kkfbNMjvaCIi2crIyOChhx7ixRdfpFOnTkycOJESJUrk2v571q/O36YtZcTSjSqsInIaMzsPeAa4BbgAyO6C+ciz7SfowgoQKKfzz+U9IiLhoFm1WGYndaH9sOmnldYTfjueTqcRM7n+kgtoUKUS9atUokGVylQpd945T2oiIvJnHD16lDvvvJNp06Zx33338dJLLxEZedZ/F56TkkVjuLNhDcau2MJLtyZQrmSxXN2/iBR4rwHt8d8C9TPgD83SFlRhNbOuQBnn3NjA68uAicDVwCLgXufc/j8SQESkoGhWLZbJfdtz6+j3+e347/diLRodyYvdEihZJIY13/3I6m9/ZOiSDaSm+bcpV6Kov7z+UIL69etTv379XDklT0QkOz///DOdOnVi1apVvPjiiwwePDjPfmmWGF+L4Us38vpHn/JgqxA+TUZEvHAT8IBzbsyf2UmwI6yPA29kej0EKA88CwwAnsY/+ZKISFjbfzSVqIgIIiyDIlGRpKalExURQamiMfRqeDV3NroagGNp6Wze9UugwO5mzbc/8cG//01Ghv8U4tjYWBo0aHCywNatW5fzzjvPy48mImFg27ZttG3blp07d/Luu+/StWvXPD3edRdV4MbLL2Tkso080KIuERE6myQ3FOZLcCSsHAF2/tmdBFtYqwKbAcysNNAa6OKcm2NmO/AXVxVWEQl7Y1dsJuXYcWpdfAH/6RLHI9OT2bhzD+NWbKFXw6tPbhcTFUndSytS99KKJMbXAuDwHf9k/fr1rFmzhjVr1rB69WqmTp0K+CcwqV69+ikltlatWhQpUsSTzykiBc/KlSvp2LEjZsaHH35I48aN8+W4A+Nr02v8XBZ98R2trq6SL8cUkQLhBWCQmX3gnMs469Y5OJdrWE/MWd4USAcWBl7vBCr80QAiIgVJ6aJFeL5rUwY3948kNK8ey8uL1pG8bddZ31uyZEni4+OJj48/ueyXX345WWDXrFnDvHnzmDBhAgDR0dHUqlXrlBJbvXr1XL8OTUQKvmnTpnHHHXdwySWXMG/ePK644op8O3a3669k8LvFGJm8SYVVRDK7CKgFfGlmi4Gsl5A659wjZ9tJsIV1I3CHma0C+gKLnXOpgXWxwJ4g9yMiUqDNGNT5lNeRERE82Ko+D7aq/4f2V758edq2bUvbtm0BcM7x/fffs3r16pMl9s0332T48OGAv/TWrVv3ZIlt0KABsbGxmtRJpJByzvHyyy/z4IMP0qhRI2bNmkX58uXzNUOR6CjubnwNLy5ax659h7iobKl8Pb6IhKxuQAb+ztkqm/UOyLXC+nfgfaA3cBj/KcEndAY+DnI/IiJyBmZGbGwssbGxdOvWDfDfmuLLL788WWJXr17NK6+8wrFj/sn2KlSocLK8nhiJrVBBJ76IhLv09HQeeOABhgwZQteuXXnzzTcpVsybmXoHxNfi+QVrGbNiM/9of4MnGUQktDjnLsuN/QRVWJ1zy80sFrgK+DrLjMDjgG25EUZERE4XERFBjRo1qFGjBr179wYgNTWVzZs3nzISO2/ePJzzX71RpUqVU0ps3bp1KVmypJcfQ0RyUUpKCj179mTmzJk8+OCDPPfcc0RERHiW5/IKZbjp6iqMXr6Zx9o2IirSuywiEl6CvobVOXcIWJfN8rmZX5tZJP577NR3zq3/0wlFROQ0RYoUoV69etSr9/tUj4cOHWL9+vWnlNh3330X+L30Zi6xNWvWJCYmxquPICJ/0E8//USHDh1Yt24dQ4YMwefzeR0JgMT4mnQZOYvZm7+mc+0rvY4jIh4ws5uB5c65g4Gvzyhrl8zOuUy6dC50MZWISD4rVaoUTZs2pWnTpieX7dmzh7Vr154ssbNnz+b1118HICYmhtq1a59yPexVV13l6SiNiJzZl19+Sdu2bfnxxx+ZPn06HTt29DrSSe2vu5yLypRk5P9n776jorr2948/exi6YsWKBaNGY40xaow9GBv2gsaGSpkZkxvTzddeclO8SW4SBUSxa8RKNNhjb1E0iHrtJYoNG6D0sn9/QPzZHZCZPcw8r7VmBaac85ZlkA/nzD47YziwEtmu3wE0A3Ag92OJZ8+GEsALV5I01cBKREQWoEyZMujcuTM6d875JaeUEn///feD98IePHgQc+fOxfTp0wEAbm5uTyzq5OHhwUWdiCzArl270L17d9jb22P79u1o0qSJ6qRHaO008G9RDxN/34dzN+PxiuogIlLBE8C1hz5+aRxYiYhsiBACVatWRdWqVdG3b18AOQu3nDx58pEh9ocffkBGRgYAoGzZsk8s6lSqVKl8NySkpMF33gbM8+2IYs68ziyRMZYuXYqhQ4fC09MT69evh6dngfwcWOD83q6HKev2Y+bOI/hOdQwRmZ2U8u+HPwVwTUqZ8fjzhBBaABWM2SYHViIiG2dnZ4c6deqgTp068PX1BZCzqNORI0ceGWIjIyMfLOpUrVq1RwbYRmkZcHW0N2p/a46cQ8SRs1gbcw6Dmr5mqj8WkVWQUuK7777D6NGj0apVK6xevRolS5ZUnfVMFUsURfcG1TFn7zFMSUuDoyN/KUVkwy4AeAs5pwc/rkHu/TwlmIiI8s7R0RFNmjRBkyZNMHLkSABAYmIiDh069GCI3bt3L5YuXQoA0AiBOhVKoUnVcnizSjm8WbUc6lUsDXu7J/8dmrP3aM5/9xzjwEr0HJmZmXj//fcxc+ZM9O/fH/PmzSsUA6CuZX2s+usMVqxYgYEDB6rOISJ1nvd+IicAacZshAMrEREZxc3NDW3btkXbtm0f3Hfjxo2cFYlDvsKBi9cREX0WYXuOAQCc7LVo6OGOK/H3cPnu/Qevcci93MWec1cgdN8/uP+dWpWxZVRfM/1piCzb/fv34ePjV+ug9QAAIABJREFUg3Xr1mH06NH46quvCs2CaO/UqoLq7sURHBzMgZXIxggh6gNo+NBdnYUQtR57mhOAfgBOG7NNUwysEsAOAPdMsG0iIrIgZcuWhbe3N7yvbAeQc/rihVsJOPj3dRy8eB0HLl5H3L3kR16TnpX9yH8BwMVBi7Gdmpmtm8iSXbt2Dd7e3oiOjkZISAgCAwNVJ+WJRiOga9UAn67cgaNHj6JevXqqk4gIgBDiIwB+yJnXjgIYJqVMfehxXwDTAFzJvWu6lHJ2HnfTE8CE3I8lgPHPeN4FAEZ9czNqYBVCLASwEznX1DnxvOdKKbMBtH3ec4iIyDoJIVDNvTiquReHT+OcX6hmZWdj/r7j0C/Z8siQ+g8XBy0iR/ZCm1crmTuXyOIcP34cnTt3xu3bt7F27doHK3wXNr5v1cGY3/cjJCQEM2bMUJ1DZPOEEBUB/AvAa1LKFCHEMgD9Acx77KnhUsqXubjzvwH8BzmnAycCaAfg4GPPSX/aQkzPYuy5JcUBfAPgmBDiphBitRDiYyHEm0KIF75RloiIbJedRoPhb9fDysBucLJ/9J8MJ3s7hPt5c1glArBt2za8/fbbSE9Px86dOwvtsAoApYo4o1+/fli4cCHu37//4hcQkTloATjnrtDrAuBqQe9ASpkhpUySUt6XUmqklNtzP3/4ZvSwChg5sEopu0opSwF4HcAk5LxB9hMA+wHECyE25/UPQ0REtiU+JQ1ajQYaIR6swqDVaBCfYtSaC0RWbdGiRejQoQMqVqyI/fv3o1GjRqqTXpper8e9e/ewZMkS1SlENk9KeQU5Rz4vIec6qQlSyk1PeWpvIUSMEGKFEKJAfpsshPAQQrQTQnR+/GbM6/P07n0pZYyUcjqA0QC+BLALgCtyDvUSERE9U9ieo0hOz0ADD3cMbVYHAJCUnoE5uYs0EdkiKSWmTJmCwYMHo0WLFtizZw+qVKmiOqtANGvWDPXr10dwcPCDS2IRkclohRBRD90CHn5QCFECQHcAnsi5/qmrEGLQY9tYC6CqlLI+gC0A5r9MkBCiqBBiPYC/AWwG8Hvube1DtxcyamAVQtQVQuiFEEuEELEAzgD4EEA0gL4Ayuf9j0BERLakmJMjpvVujagvB+Hrni2h1Qi0eMUDbs4OqtOIlMjIyICfnx/Gjx+PwYMHY8OGDShevLjqrAIjhIBer0d0dDQOHHjaZRiJqABlSikbP3QLfexxLwAXpJQ3c0/JXQWg+cNPkFLellL+c9rTLABvvGTT1wAqA2iJnPe09gTQBkAYchZdMmq1RWOPsMYg5xDyHQAjAJSUUr4hpRwlpVwppYzLWzsREdmaCEMPfOzVGBqNQLlirujTqCaOXr2JJSO6qE4jMrvExER4e3tjzpw5GDduHObPnw8HB+v75c3AgQNRpEgRBAcHq04hsnWXADQTQrgIIQSAdwA8spiuEOLhg5DdHn88HzoD+ArAn7mfX5VS7pRSBgD4DcBnxmzE2IE1HDnDqg45Kz9NEUL0EkKUzlszERFRjpFtGiI+OQ2/HjypOoXIrGJjY9GyZUts3boVYWFhmDx5MnJ+frQ+RYsWxaBBgxAeHo47d+6oziGyWVLKPwGsAHAYOZe00QAIFUJMFkJ0y33av4QQx4UQR5CzorDvS+62LIDLUsosAEkASj702DoA7xqzEWMXXRogpawEoAaAnwEUQc6qwTeEEP8TQvDXZkRElCdvv1IR9SqWxozt0Xx/G9mMmJgYNGvWDBcuXEBkZCSGDx+uOsnkdDodUlNTMX/+S70djohekpRygpSylpSyrpRysJQyTUo5Xkq5JvfxL6WUdaSUDaSUbaWUL/sb5csA/jnAeQaA90OPNQWQ+sQrniKviy5dkFLOB/Df3NsuALUABDz3hURERI8RQsDQuiH+uhyHPy9cU51DZHKbNm1CixYtAAC7d+/Gu+8adXCh0GvQoAHeeusthISE8JdTRLZlM3LeOwsAPwIYKYTYK4TYBmAqgAXGbMTYRZeaCSE+E0KsEULcBnAEOZe3SUDOucdv5bWeiIhoYJPaKOrkgKAdR1SnEJnU3Llz0aVLF3h6emL//v2oX7++6iSz0uv1OH36NLZu3ao6hYjM5wsAEwBASrkQQG/kLLZ0F8DI3MdfyNgjrHuRcx7zfQBjAdSXUrpLKbtLKb+XUnLpNyIiyrOiTg4Y2uw1hB86hVv3k1XnEBU4KSXGjx+P4cOHo23btti1axc8PDxUZ5ld3759UbJkSYSEhKhOISIzkVImSylvPfT5ainlQCllLwCnkPM+1hcydmB9RUpZSUr5npQyWEp5PB/NRERET9C3aoj0zCxej5WsTnp6OoYOHYopU6Zg+PDhiIyMhJubm+osJZycnDBs2DBERETg2jW+BYDImgkhigsh+ueeodtbCGH/0GN9hRBRyDld2NOY7Rm76NKF3B2UEEK0FEK8l3vxWQghnIQQeXovLBER0T9eq1AKbWpWQvDOI8jKzladQ1Qg4uPj0alTJyxcuBBTpkzB7NmzYW9v/+IXWrHAwEBkZmZi9uzZqlOIyESEEPWQczmcJQC+BbAcwD4hRBUhxB7kXH3GEcAgAK8Zs01j38NqJ4T4DkAsgB0AFuL/T8QrkXtuMhERUX6MbN0QF28nYsPxi6pTiF7apUuX0KJFC+zatQsLFizA2LFjrfayNXlRo0YNtG/fHqGhocjMzFSdQ0Sm8W8AichZ48gFQG3kXB71IIC6AIZIKetJKX+VUhr1W2pjj4z+G4A/gPcBVAPw8Hfd3wB0NXI7RERET+je8BWUL+aKoB3RqlOIXsrhw4fRtGlTxMbGYsOGDRg8eLDqJIui0+kQGxuLdeuMeusaERU+jQGMk1L+KaVMlVKeAqBHzuVtPpFSLsrrBo0dWIcAGC2lnIuc6+k87BxyhlgiIqJ8sbezQ0CL+lh//ALO34xXnUOUL+vWrUOrVq3g4OCAPXv2oF27dqqTLE7Xrl1RoUIFLr5EZL3KArj42H3/fJ6vSwIYO7AWR85g+jQOAOzys3MiIqJ/+LeoB40QCNnJS9xQ4RMaGopu3brh1Vdfxf79+1GnTh3VSRbJ3t4efn5+2LBhAy5cuKA6h4hM41kXXM7XewGMHViPAej+jMc6ATicn50TERH9o2KJoujZsAbC9h5DSnqG6hwio2RnZ+PLL79EYGAgOnTogB07dqB8+fKqsyyav78/NBoNZs6cqTqFiExjoxAi7p8bgH+WBv/j4ftzH3shrZE7nQpgpRDCGTkrPUkADYUQPQEEAuiWxz8EERHREwytG2DF4dNYfvg0hjTjESqybGlpafD19cXSpUsRGBiI6dOnQ6s19kcr2+Xh4YGuXbtizpw5mDRpEhwdHVUnEVHBmVTQGzTqu6qU8jchxHsAvgMwPPfu2QCuABgspdxY0GFERGR72tSshNrlSmLG9mgOrGTR7ty5g549e2Lnzp345ptv8Pnnn3Ml4DzQ6XSIiIjAqlWrMGDAANU5RFRApJQFPrAaff1UKeUyKWVVALUAtEDOdXMqSymXFXQUERHZJiEEDK0b4sDF64j6+7rqHKKnunDhApo3b479+/fj119/xRdffMFhNY/at2+PV155BcHBwapTiMjCGT2w/kNKeVpKuVdKeVJK+aw31BIREeXL4GavwdXRHsE7uPgSWZ6DBw+iWbNmiIuLw+bNm9G/f3/VSYWSRqNBYGAgdu3ahePHj6vOISILZvTAKoSoIIQIEEJMFkJ899jtW1NGEhGR7Sjm7IhBTWpjycGTuJOUojqH6IE1a9agdevWcHV1xd69e9GqVSvVSYWar68vHBwceIkbInouowbW3MWVzgOYAWAEgL5PuRERERUIQ+uGSM3IxLx9PPJClmHGjBno2bMn6tati3379qFWrVqqkwo9d3d39O3bFwsWLEBSUpLqHCKyUMYeYf03gE0AykopK0opPR+7VTNhIxER2Zj6Hu5oUb0ignccQXY2331C6mRnZ+PTTz/F+++/D29vb2zbtg1ly5ZVnWU19Ho9EhMT8euvv6pOISILZezAWgnAz1LKO6aMISIi+oehdQOcvRmPLSf/Vp1CNiolJQU+Pj74/vvv8cEHH2DVqlVwdXVVnWVVmjdvjrp16yI4OBhcGoWInsbYgXUvgFdNGUJERPSwXg1roExRF8zYHq06hWzQrfvJ8PLywsqVK/HDDz/gp59+gp2dneosqyOEgF6vx+HDhxEVFaU6h4gskLED68cAAoQQQ3MXX3J5/GbKSCIisj2O9lr4t6iH34+ex99/8ygrmc/ZuLt469tfcejQISxbtgwfffQRL1tjQoMGDYKrqysvcUNET2XswBoDoB6AuQAuA7j3lBsREVGBCmhZHwAQGhqquIRsxb7zV/HWd7/ibnIqtm7dij59+qhOsnpubm4YOHAgli5dirt376rOISILY+zAOhzAsNzb8GfciIiIClTlkm7oWr8aZs2ahbS0NNU5ZOVWHj6Ndj8uR3FnR+z7/D00b95cdZLN0Ol0SElJwYIFC1SnEJGFMWpglVLOk1LOf97N1KFERGSbDK0b4ubNm1i5cqXqFLJSUkr8uOUQ+s5ai9cruWPv5wNQo2wJ1Vk25fXXX0fTpk0REhLCxZeI6BFaY58ohPAB4A+gJgCnxx+XUpYpwC4iIiIAgFetKqhRowaCgoLw3nvvqc4hK5OVnY2Plm/HL9v+Qu/Xa2DhsE5wdrAHADQ28RpAUY1Nu/3CRq/Xw9fXF9u3b0fbtm1V5xCRhTDqCKsQ4j0A8wGcBeABYA2A33NfnwhguqkCiYjItmk0OauI7tmzB0eOHFGdQ1YkOT0DvWeuwS/b/sLHXm9gmX/XB8MqmV+/fv1QokQJhISEqE4hIgti7HtYPwMwBcDI3M+DpJTDAXgCuAUg2QRtREREAABfX184OzsjKChIdQoVUgkpaegZ/BsSUnLeCx2XmIy2PyzDmphz+NmnLb7v0wYaDVcCVsnZ2Rm+vr5YtWoVrl+/rjqHiCyEsQNrDQB7pJRZALIAuAGAlPIegG8BvG+aPCIiIqBEiRJ47733sGjRIiQkJKjOoUJozZFziDhyFmtjzuHU9Tto9t0SHL1yC6sDu+ODto1U51EunU6HzMxMzJkzR3UKEVkIYwfWBACOuR9fAVD7occEgFIFGUVERPQ4g8GA5ORkriJK+TJn71EAwA9bDqH5tF+RlJaB7R/3Q/eG1RWX0cNq1qyJd955BzNnzkRWVpbqHCKyAMYOrFEA6ud+vAbAeCGEvxBiKIBpAP40RRwREdE/GjVqhKZNmyIoKIiriNILef13OYTu+we3veeuAgD+uhyHO0mpiLuXjKbfLoHXf5crLqXH6XQ6XLp0CevXr1edQkQWwNiB9WsAl3I/Hg/gAIAgAHOR8x7WgIJPIyIiepTBYMDJkyexbds21Slk4cZ0agoXh/9/MYT0rOwnnuPioMXYTs3MmUVG6N69O8qXL8/Fl4gIgJGXtZFS7gewP/fjeADdhRCOAByllIkm7CMishqmvESGrVweo1+/fvj4448RFBSEdu3aqc4hC9b21cr4fWRPeM9YjeT0zCced3HQInJkL7R5tZKCOnoee3t7+Pn5YerUqbh48SKqVq2qOomIFDL2COsTpJRpHFaJiMicnJycMGLECERERODKlSuqc8jCtX21MsL9vGFv9+iPO072dgj38+awasH8/f0hhEBoaKjqFCJSLN8DKxERkQo6nQ7Z2dn8QZaMcicpFVnZOacDO9troRECWo0G8bmXtyHLVKlSJXh7eyMsLAzp6emqc4hIIQ6sRERUqHh6eqJz584IDQ1FRkaG6hyycN9sPIBsCVQt5Ybf9N3RwMMdyekZmLPnmOo0egGdToe4uDisXr1adQoRKWTUe1iJiIgsicFgQJcuXRAREYG+ffuqziELlZGVhb/vJKJCMVecmTQCWq0G7WpVxn//OIRdZ3lKuaXr0KEDPD09ERwcDB8fH9U5VEiYcr0IwHbWjLAkPMJKRESFzj8/yM6YMUN1Clmw+fuOIzk9EyED20OrzfmRx06jwSft30SEvofiOnoRjUaDwMBA7NixAydOnFCdQ0SKcGAlIqJCx87ODjqdDjt27MDx48dV55AFSsvIxOTI/WhStRy861VTnUP5NGzYMNjb2/MSN0Q2jAMrEREVSsOHD4ejoyOCg4NVp5AFCt0dg8t37+Gr7i0ghFCdQ/lUpkwZ9OnTB/Pnz0dycrLqHCJSwOIGViFERyHEKSHEWSHEaNU9RERkmUqXLg0fHx8sWLAA9+7dU51DFiQ5PQNfrf8TrWt44J1alVXn0EvS6/VISEjA0qVLVacQkQIWNbAKIewAzADQCcBrAAYIIV5TW0VERJbKYDDg3r17WLRokeoUsiDTt/2FG4nJmNr9bR5dtQItWrRAnTp1eDYFkY2yqIEVQBMAZ6WU56WU6QCWAuiuuImIiCxUkyZN8MYbbyAoKAhSStU5ZAESU9Lw7aaD6FinKlpU91CdQwVACAGdToeoqChERZl4CVgisjiWNrBWBHD5oc9jc+8jIiJ6ghACBoMBx44dw+7du1XnkAX479bDuJOUiind3ladQgVo8ODBcHFx4eJLRDZIWNJvpIUQfQF0kFL65X4+GEATKeUHjz0vAEAAADg4OLyRlpZm9lbVeI2p/OPXjqjwed7/t9mpyYjpXBFuzTqg2r/z9x43a/7/1pa+5925cweenp5o164dVq9erTrH5hX0372/p/rj9obFqL/+KqLbFi/YjRMVckKIZCmlq+oOU7C0I6yxACo99LkHgKuPP0lKGSqlbCylbKzVas0WR0RElkfj5IJSXYfh7taVyLh1XXUOKTRt2jTcu3cPkydPVp1CJlC6tw4yLQV31i1UnUJEZmRpA+tBADWEEJ5CCAcA/QGsUdxEREQWzr23HsjKxK3fZqtOIUVu3LiBn3/+Gf3790e9evVU55AJuNZ+Ay6vvYmbK4P5nnUiG2JRA6uUMhPA+wA2AjgBYJmUkleEJyKi53KqXANuzd7FzVUzITMzVeeQAl9//TXS0tIwceJE1SlkQu699Ui9cAI7d+5UnUJEZmJRAysASCnXSSlrSilfkVJ+pbqHiIgKB/c+BmTExSJ+11rVKWRmly9fRnBwMIYOHYqaNWuqziETKvmuD+yKFufiS0Q2xOIGViIiovwo1sIbDuUq4+aKINUpZGZfffUVpJQYN26c6hQyMY2TC0p1GYqVK1fixo0bqnOIyAw4sBIRkVUQdnYo3SsQ9w5sQerFU6pzyEzOnz+PsLAw+Pv7o2rVqqpzyAzce+uQkZGBuXPnqk4hIjPgwEpERFajdLcREFp73FwZrDqFzGTSpEnQarUYM2aM6hQyE6eqtdC2bVvMnDkTWVlZqnOIyMQ4sBIRkdWwL1UWxd/pg9u/z0NWSpLqHDKxEydOYNGiRRg5ciQqVKigOofMSKfT4eLFi9i4caPqFCIyMQ6sRERkVcr0HYms+wm4u/FX1SlkYhMmTICLiwu++OIL1SlkZj169EDZsmW5+BKRDeDASkREVsW1fnM416iPuOUzeK1GKxYdHY3ly5dj1KhRcHd3V51DZubg4AA/Pz9ERkbi0qVLqnOIyIQ4sBIRkVURQsC9jwEpp6ORdHS/6hwykfHjx6N48eL45JNPVKeQIv7+/pBSIjQ0VHUKEZkQB1YiIrI6JTsOhMbVjZe4sVL79+/H2rVr8emnn6J48eKqc0iRKlWqoEuXLpg9ezYyMjJU5xCRiXBgJSIiq2PnUgSlvIfi7pZlyLh7U3UOFbBx48ahdOnS+PDDD1WnkGI6nQ43btxARESE6hQiMhEOrEREZJXce+shM9Jx+7cw1SlUgLZv344tW7bgyy+/RJEiRVTnkGIdO3ZElSpVEBzMS1kRWSsOrEREZJWcPWujaOO2uLkqBJLXarQKUkqMHTsWFSpUgF6vV51DFsDOzg6BgYHYtm0bTp48qTqHiEyAAysREVkt974jkX7tbyTsXa86hQrAxo0bsWfPHowdOxbOzs6qc8hCDB8+HPb29pg5c6bqFCIyAQ6sRERktYq36gZ79wq4uXyG6hR6Sf8cXa1SpQpGjBihOocsSNmyZdGrVy/MmzcPKSkpqnOIqIBxYCUiIqsltPYo3TMAifs2IC32nOocegkRERE4dOgQJkyYAAcHB9U5ZGH0ej3i4+MRHh6uOoWIChgHViIismqle/gDdlrcXBmiOoXyKSsrC+PGjUPNmjUxePBg1TlkgVq1aoXatWtz8SUiK8SBlYiIrJqDewWUaNsTt9bOQXYqTxcsjJYtW4bjx49j0qRJ0Gq1qnPIAgkhoNPpcODAARw+fFh1DhEVIA6sRERk9dz7GJCVcAd3NvN0wcImMzMTEyZMQL169dCvXz/VOWTBhgwZAmdnZ4SE8GwKImvCgZWIiKxekUat4eT5Gm6uCFKdQnm0YMECnDlzBlOmTIFGwx9b6NmKFy+OAQMGYPHixUhISFCdQ0QFhN/5iYjI6gkh4N7HgOT/HUTS8YOqc8hIaWlpmDRpEho3boxu3bqpzqFCQK/XIzk5GYsWLVKdQkQFhG8EIZsT1Vh1ARGpUKrzYFyZ/gVurgiCa525qnPICLNnz8alS5cQGhoKIYTqHCoEGjdujDfeeAPBwcEwGAz8e0NkBXiElYiIbIJdETeU7DwYdzYvRWb8bdU59ALJycmYOnUqWrZsiXfffVd1DhUier0ex48fx+7du1WnEFkUIcRHQojjQohjQohfhRBOjz3uKIQIF0KcFUL8KYSoqqb0URxYiYjIZpTpY4BMS8Xt3+epTqEXCA4OxvXr1zF16lQeJaM86d+/P4oVK8bFl4geIoSoCOBfABpLKesCsAPQ/7GnjQBwV0pZHcCPAL41b+XTcWAlIiKb4Vy9Hoq83hI3VwZDZmerzqFnuHfvHr755hu0b98erVq1Up1DhYyrqyuGDBmCFStW4ObNm6pziCyJFoCzEEILwAXA1cce7w5gfu7HKwC8IyzgN4YcWImIyKa49zEgLfYcEvdvUp1Cz/DTTz/h1q1bmDp1quoUKqR0Oh3S09Mxdy7fr042QyuEiHroFvDwg1LKKwD+A+ASgGsAEqSUj/9DWBHA5dznZwJIAFDK9OnPx4GViIhsSvG2vaAtWZaXuLFQd+/exX/+8x9069YNTZo0UZ1DhdRrr72G1q1bY+bMmcjm2RRkGzKllI0fuoU+/KAQogRyjqB6AqgAwFUIMeixbTztaKo0Ta7xOLASEZFN0dg7oHQPfyTs/h1pVy+qzqHH/Oc//0FCQgImT56sOoUKOZ1Oh/Pnz2PTJp5NQQTAC8AFKeVNKWUGgFUAmj/2nFgAlQAg97ThYgDumLXyKTiwEhGRzXHvGQAIgVurZqpOoYfExcXhp59+go+PDxo0aKA6hwq5Xr16oUyZMlx8iSjHJQDNhBAuue9LfQfAiceeswbA0NyP+wDYKqXkEVYiIiJzcyhXCcVbdcOt32YjOz1NdQ7l+vbbb5GSkoKJEyeqTiEr4ODggBEjRmDt2rW4fPmy6hwipaSUfyJnIaXDAI4iZw4MFUJMFkJ0y31aGIBSQoizAD4GMFpJ7GM4sBIRkU1y7zMSmfG3cPePFapTCMCVK1cwY8YMDB48GLVq1VKdQ1YiICAAUkrMmjVLdQqRclLKCVLKWlLKulLKwVLKNCnleCnlmtzHU6WUfaWU1aWUTaSU51U3AxxYiYjIRhV9sx0cK9fk4ksW4quvvkJWVhYmTJigOoWsSNWqVdGpUyfMnj0bGRkZqnOIKB84sBIRkU0SGg3ce+uRFLMX0dHRqnNs2oULFzB79mz4+fnB09NTdQ5ZGZ1Oh2vXrmHNmjWqU4goHziwEhGRzSrlPRTC0RlBQTzKqtLkyZOh0WgwZswY1SlkhTp37ozKlSsjODhYdQoR5QMHViIisllatxIo1XEgFi9ejPj4eNU5NunUqVNYsGABDAYDPDw8VOeQFbKzs0NAQAD++OMPnD59WnUOEeURB1YiIrJp7n30SE5Oxvz581Wn2KSJEyfCyckJo0dbxGKUZKVGjBgBrVaLmTN5KSuiwoYDKxER2TSXWo3QrFkzBAUFwQIuN2dTYmJisHTpUnz44YcoU6aM6hyyYuXKlUPPnj0xb948pKSkqM4hojzgwEpERDZv5MiROH36NLZu3ao6xaaMHz8exYoVw2effaY6hWyAXq/HnTt3sHz5ctUpRJQHHFiJiMjm9enTB6VLl8aMGTNUp9iMgwcP4rfffsMnn3yCEiVKqM4hG9CmTRu8+uqrXHyJqJDRqg6g/IlqrLqAiMh6ODk5YcSIEZg2bRpiY2O5+I8ZjB07FqVKlcKHH36oOoVshBACOp0OH330EaKjo9GwYUPVSURkBB5hJSIiAhAYGAgpJUJDQ1WnWL2dO3di06ZNGD16NNzc3FTnkA0ZOnQonJycEBISojqFiIzEgZWIiAiAp6cnunTpglmzZiE9PV11jtWSUmLs2LEoV64cDAaD6hyyMSVKlED//v2xaNEiJCYmqs4hIiNwYCUiIsplMBhw/fp1rF69WnWK1dq8eTN27dqFMWPGwMXFRXUO2SC9Xo+kpCQsXrxYdQoRGUEU9iX8XV1dZVJSkuoMIiIyocZRpt3+P+sCZGdno0aNGvDw8MCOHTtMu1MzMdfXzhhSSjRt2hQ3btzA6dOn4ejoaLowUs6Uf/deZi0PKSUaN26MjIwMHDlyBEKIggsjUkQIkSyldFXdYQo8wkpERJRLo9FAr9dj586dOHbsmOocq7N27VocPHgQ48eP57BKyvyz+NLRo0exd+9e1TlE9AIcWImIiB4ybNgwODo6IigoSHWKVcnOzsa4ceNQvXp1DBkyRHUO2bgBAwbAzc2Niy8RFQIcWImIiB5SqlQp9O/fHwsXLuSiLAVo+fLliImJwaRJk2Bvb686h2w1RK/WAAAgAElEQVRckSJFMHjwYCxbtgy3bt1SnUNEz8GBlYiI6DEGgwH379/HokWLVKdYhczMTEyYMAF16tSBj4+P6hwiADmLL6Wnp2PevHmqU+gZEhIS0LNnTyQkJKhOIYU4sBIRET2mSZMmaNy4MYKCglDYFye0BIsWLcKpU6cwefJk2NnZqc4hAgDUqVMHLVu2REhICLKzs1Xn0FOsWbMGERERWLt2reoUUogDKxER0VMYDAYcP34cO3fuVJ1SqKWnp2PSpElo1KgRevbsqTqH6BE6nQ7nzp3Dli1bVKfQU8yZM+eR/5Jt4sBKRET0FD4+PihRogQXX3pJc+bMwcWLFzF16lRePoQsTu/eveHu7s7FlyyEl5cXhBAPbv+s4rxnz55H7vfy8lJcSubEgZWIiOgpXFxcMHz4cKxatQrXrl1TnVMopaSkYMqUKWjevDk6duyoOofoCY6Ojhg+fDjWrFmDK1euqM6xeWPGjIGLi8uDz9PT0x/5L5DzvXns2LFmbyN1RGF/b46rq6tMSkpSnUFERFbo7NmzqFGjBiZNmoTx48erzsmXxlGm3X5U42c/9uOPP+Ljjz/Gtm3b0KZNG9OGkMUx5d+95/29y6vz58+jevXqGD9+PCZOnFhwG6Z82bZtG7y9vZGcnPzEYy4uLoiMjOT3k6cQQiRLKV1Vd5gCj7ASERE9Q/Xq1dGhQwfMnDkTGRkZqnMKlfv37+Prr7/GO++8wx8uyaJVq1YNHTp0wKxZs5CZmak6x+Y1adIEb7/99hP3Ozk5ITw8nN9PbBAHViIioucwGAy4evUqV6nMo59//hk3b97E1KlTVacQvZBer+f/5xbg8OHDaNSoETZv3gx7e3toNBo4OztDo9FAq9UiPj5edSIpwFOCiYiIniMrKwvVqlVDjRo1CuVKoipOCY6Pj4enpydatGjBAYAKhczMTHh6eqJ27drYtGmT6hybk52djR9//BFffvkl3N3d4e7ujqNHj6JBgwb49ttv8cUXX+DIkSNo3bo1tm7dqjrXIvGUYCIiIhtlZ2cHnU6HP/74AydPnlSdUyj88MMPiI+Px+TJk1WnEBlFq9UiICAAmzdvxtmzZ1Xn2JTr16+jU6dO+PTTT9G5c2fExMSgatWqmDZtGqKiotC+fXscPHgQ3333Hdzc3FTnkgI8wkpERPQCcXFxqFSpEnQ6HX766SfVOXli7iOsN2/eRLVq1dCxY0csX77ctDsnKkBXr15F5cqV8dFHH2HatGmqc2zCunXr4Ovri3v37uHHH39EYGAgL3+VTzzCSkREZMPKlCmDvn37Yt68eeAvSZ/vu+++Q3JyMo+uUqFToUIF9OjRA3PnzkVqaqrqHKuWmpqKUaNGoUuXLihXrhyioqKg0+k4rNJTcWAlIiIygsFgQGJiIhYvXqw6xWJdu3YN06dPx8CBA1G7dm3VOUR5ptfrcfv2baxYsUJ1itX63//+h2bNmuGnn37CBx98gAMHDqBOnTqqs8iC8ZRgIiIiI0gp8frrr0NKiejo6EJzJMCcpwS///77mDlzJk6ePIlXXnnFtDsmMoHs7GzUqlUL7u7u2LNnj+ocqyKlRGhoKD766CO4urpi3rx56NKli+osq8FTgomIiGycEAIGgwExMTHYt2+f6hyL8/fffyM0NBTDhw/nsEqFlkajgU6nw969exETE6M6x2rcvn0bvXv3hk6nQ4sWLRATE8NhlYzGgZWIiMhIAwcOhJubG4KCglSnWJwpU6ZACIGxY8eqTiF6Kb6+vnB0dERISIjqFKuwfft2NGjQAL///jumTZuGDRs2oHz58qqzqBDhwEpERGQkV1dX+Pr6Yvny5YiLi1OdYzHOnDmDefPmQafToVKlSqpziF5KyZIl4ePjg4ULF+LevXuqcwqtjIwMjBkzBu3atYOLiwv27duHTz/9FBoNxw/KG/6NISIiygO9Xo/09HSEhYWpTrEYEydOhKOjI7788kvVKUQFQq/X4/79+1iyZInqlELp/PnzaNmyJf79739j2LBhOHz4MN544w3VWVRIcdElIiKiPPLy8sKZM2dw/vx52NnZqc55LlMvujTP6Rjq16+Pzz//HN98841pd0ZkJlJKNGrUCFJK/PXXX4VmkTVLsGjRIhgMBmg0GoSGhqJfv36qk2wCF10iIiKiBwwGAy5duoTIyEjVKcqNHz8eRYsWxWeffaY6hajACCGg0+lw5MgR7N+/X3VOoZCYmIjBgwdj8ODBqF+/Po4cOcJhlQoEB1YiIqI86tatGypUqGDziy8lnTiE1atX4+OPP0apUqVU5xAVqPfeew9Fixbl4ktG+PPPP/H6669jyZIlmDhxIrZv344qVaqoziIrwYGViIgoj7RaLQIDA7Fx40acPXtWdY4yV0PGoWTJkhg1apTqFKICV7RoUQwaNAjh4eG4ffu26hyLlJWVha+//hotWrRAZmYmduzYgQkTJkCr1apOIyvCgZWIiCgf/P39odVqbfboy/0je5C4dz0+//xzFCtWTHUOkUno9XqkpaVh/vz5qlMsTmxsLLy8vPB///d/6NWrF44cOYIWLVqoziIrxEWXiIiI8snHxwebN29GbGwsXFxcVOc8lSkWXZJS4rSuLVIvnkT8pXNwdbXKdT6IAAAtWrRAXFwcTp48yUuy5IqIiMCIESOQlpaGX375Bb6+vlyYSjEuukRERERPMBgMuHv3LsLDw1WnmNW9g1tx//AOlB/2fxxWyerpdDqcOXMGW7duVZ2iXHJyMnQ6HXr27ImqVavi8OHDGDZsGIdVMikeYSUiIsonKSXq1asHZ2dnHDx4UHXOUxX0EVYpJU4NfwvpcVdQd9UZHH7bqWB3QGRhUlNT4eHhgTZt2mDFihWqc5Q5cuQIBgwYgBMnTuDTTz/FV199BQcHB9VZlItHWImIiOgJQggYDAZERUXhwIEDqnPMImF3JJKO/YnyfuOhceSwStbPyckJw4cPR0REBK5evao6x+yklPj555/RpEkT3L17F5s2bcK0adM4rJLZcGAlIiJ6CYMGDUKRIkVs4hI3MjsbV0PGwaFiNZTu6qs6h8hsAgICkJWVhdmzZ6tOMau4uDh4e3vjww8/RPv27RETE4P27durziIbw4GViIjoJbi5uWHw4MFYunSp1V/6In7rSqScjkaFgIkQWnvVOURmU716dbz77ruYNWsWMjMzVeeYxaZNm1C/fn388ccf+Pnnn7F27Vq4u7urziIbxIGViIjoJRkMBqSlpWHu3LmqU0xGZmXh6szxcPKsjZId3lOdQ2R2er0esbGxiIyMVJ1iUunp6fj000/RoUMHlCpVCgcOHMAHH3zAhZVIGQ6sREREL6lu3bpo1aoVgoODkZ2drTrHJO5sXILUiydRIXAyhJ2d6hwis/P29kbFihURHBysOsVkTp06hbfeegvff/899Ho9Dh48iPr166vOIhvHgZWIiKgAGAwGnD9/Hhs3blSdUuBkZgauhk6Ec82GKN62l+ocIiW0Wi38/f2xceNGnD9/XnVOgZJSIiwsDI0aNcLFixcRERGBoKAgi72+NNkWixlYhRDThBAnhRAxQojVQojiqpuIiIiM1bNnT5QrV84qF1+6tWYu0q+cR0X9VAiNxfzoQGR2fn5+sLOzw8yZM1WnFJi7d+/Cx8cHfn5+aNq0KWJiYtC9e3fVWUQPWNK/OpsB1JVS1gdwGsCXinuIiIiM5uDgAH9/f0RGRuLChQuqcwpMdloqroVNgWu9ZnB7u7PqHCKlKlasiG7dumHOnDlIS0tTnfPSdu/ejYYNG2L16tX4+uuvsXnzZlSsWFF1FtEjLGZglVJuklL+s+zafgAeKnuIiIjyKiAgABqNxqqOvtxcNRMZcbGooJvKRVeIkLP40q1bt7By5UrVKfmWmZmJCRMmoHXr1tBqtdizZw9Gjx4NO74/nSyQkFKqbniCEGItgHAp5aIXPdfV1VUmJSWZoYqIiOjFevXqhV27duHy5ctwcnJSnYPGUfl/bVZKEo71qAbnanVQM3jrU58T1Tj/2ycqjLKzs1GzZk2UL18eu3btUp2TZxcvXsTAgQOxd+9eDBkyBL/88gvc3NxUZ9FLEkIkSyldVXeYglmPsAohtgghjj3l1v2h54wBkAlg8XO2EyCEiBJCRNnKtbCIiKhwGDlyJG7duoUVK1aoTnlpN5dNR+adOFTQTVGdQmQxNBoNdDoddu/ejWPHjqnOyZPw8HA0bNgQR48exeLFizF//nwOq2TxLOoIqxBiKAAdgHeklMnGvIZHWImIyJJIKVG7dm2UKFEC+/btU52T7yOsWfcTcLS7J1zrNkONn9Y983k8wkq26NatW/Dw8ICfnx+mT5+uOueF7t+/jw8++ADz5s1D06ZNsWTJElSrVk11FhUgHmE1AyFERwBfAOhm7LBKRERkaYQQ0Ov12L9/Pw4fPqw6J99uLPkRWYl3eXSV6ClKly6Nvn37YsGCBbh//77qnOeKiopCo0aNMH/+fIwZMwa7du3isEqFisUMrACmAygKYLMQIloIEaI6iIiIKD+GDh0KFxcXBAcHq07Jl8z427ix5AcUb9sLrrXfUJ1DZJH0ej3u3buHX3/9VXXKU2VnZ2PatGlo3rw5UlJSsG3bNkydOhX29vaq04jyxKJOCc4PnhJMRESWKCAgAIsWLcKVK1dQokQJZR35OSU49pcvcGPhNLz261E4v1Lnuc/lKcFkq6SUaNiwIezs7HDo0CGLWkX72rVrGDJkCLZs2YJevXph1qxZKFmypOosMiGeEkxERER5YjAYkJKSgvnz56tOyZOMW9cRF/4LSnZ474XDKpEtE0JAp9Phr7/+woEDB1TnPLB27VrUr18fe/bsQWhoKFasWMFhlQo1DqxEREQm0LBhQ7z11lsICgpCdna26hyjXZ/3NWRmOsr7T1CdQmTxBg0ahCJFiiAkRP072VJSUvDBBx+gW7duqFixIg4dOgR/f3+LOvJLlB8cWImIiExk5MiROHPmDP744w/VKUZJv34JN1eFoJS3L5wq11CdQ2TxihYtioEDB2Lp0qW4c+eOso5jx46hSZMmmD59OkaNGoX9+/ejdu3aynqIChIHViIiIhPp06cPSpcujaCgINUpRrkWNhUAUMFvvOISosJDr9cjNTUVCxYsMPu+pZQICgrCm2++ibi4OKxbtw4//vgjnJyczN5CZCocWImIiEzE0dERfn5+WLNmDS5fvqw657lSL5/FrbVzULpnABzKVVadQ1RoNGjQAG+99RZCQkJgzsVMb926hR49emDkyJFo3bo1YmJi0KlTJ7Ptn8hcOLASERGZkE6ng5QSoaGhqlOe69qsSRBaB5Qf9n+qU4gKHZ1Oh1OnTmHbtm1m2d/WrVvRoEEDrF+/Hj/88APWrVuHsmXLmmXfRObGgZWIiMiEqlSpAm9vb8yaNQvp6emqc54q5fz/cGfDYpTp9z7sS5dXnUNU6PTr1w8lS5Y0+eJLGRkZGD16NLy8vFC0aFH8+eef+Oijj6DR8Ed6sl78201ERGRiBoMBN27cwKpVq1SnPNXV0AnQOLui3JDPVacQFUpOTk4YNmwYVq9ejWvXrplkH2fPnsXbb7+Nb7/9FiNGjMChQ4fw+uuvm2RfRJaEAysREZGJvfvuu3jllVcscvGl5FN/If6PFSg74CNoi5dWnUNUaAUGBiIzMxNhYWEFul0pJebPn4/XX38dZ86cwfLlyzFr1iy4uroW6H6ILBUHViIiIhPTaDTQ6/XYtWsXjh49qjrnEVdDxsHOrQTKDPxYdQpRoVajRg14eXkhNDQUWVlZBbLNhIQEDBw4EL6+vmjUqBFiYmLQp0+fAtk22RYhxKtCiOiHbolCiFGPPaeNECLhoedYxJLxHFiJiIjMwNfXF05OThZ1lPV+zD4k7I5E2UGfQVu0uOocokJPr9fj8uXLWLdu3Utva9++fWjYsCGWLVuGyZMnY+vWrahUqVIBVJItklKeklI2lFI2BPAGgGQAq5/y1F3/PE9KOdm8lU/HgZWIiMgMSpUqhf79+2PhwoVITExUnQMAuBo8FtqSZVDG5wPVKURWoWvXrihfvjyCg4PzvY2srCxMnToVLVu2BADs3LkT48aNg52dXUFlEr0D4JyU8m/VIcbgwEpERGQmI0eORFJSEhYuXKg6BYkHt+Je1FaUG/ol7FyKqM4hsgr29vbw9/fHhg0bcOHChTy//vLly2jXrh3GjRuHvn37Ijo6Gs2bNzdBKVkhrRAi6qFbwHOe2x/Ar8947C0hxBEhxHohRB0TdOaZMOcFjk3B1dVVJiUlqc4gIiIySpMmTXD//n0cP34cQgiT769x1JP3SSlxasTbSL9xCXVXnYXG0Snf249q/BJxRFYoNjYWVapUweeff46vv/7a6NetXLkSfn5+yMzMxIwZMzB48GCzfI8g6yCESJZSvnAlLiGEA4CrAOpIKW889pgbgGwp5X0hRGcAP0kpa5im2Hg8wkpERGRGBoMBJ06cwI4dO5Q1JO5dj6Sj+1B++NiXGlaJ6EkeHh7o2rUrwsLCjLr2clJSEgICAtCnTx9Ur14df/31F4YMGcJhlUylE4DDjw+rACClTJRS3s/9eB0AeyGE8uXjObASERGZkY+PD0qWLKls8SUpJa4Ej4VDBU+U6jZcSQORtdPr9bh58+YLr70cHR2Nxo0bY/bs2fjiiy+wZ88eVK9e3UyVZKMG4BmnAwshyonc35QIIZogZ1a8bca2p+LASkREZEbOzs4YPnw4Vq9ejatXr5p9//HbViHl1F+o4D8BGnsHs++fyBa0b98e1apVQ3BwMBISEtCzZ08kJCQ8eDw7Oxs//vgjmjZtioSEBGzevBnffPMNHBz4/ySZjhDCBUB7AKseuk8nhNDlftoHwDEhxBEAPwPoLy3g/aMcWImIiMxMp9MhMzMTs2bNMut+ZVYWroaMh2OVV1Gy40Cz7pvIlmg0GgQGBmLnzp0IDg5GREQE1q5dCwC4ceMGunTpgo8//hgdO3ZETEwM3nnnHcXFZAuklMlSylJSyoSH7guRUobkfjxdSllHStlAStlMSrlXXe3/x4GViIjIzF555RV07NgRoaGhyMjIMNt+72xaitQL/0OFwMkQWq3Z9ktki4YNGwYHBwdMnz4dADBnzhxs2LAB9evXx/bt2zFjxgxERESgdGnlbxEksmgcWImIiBQYOXIkrl69ijVr1phlfzIzA9dCJ8C5Rn2UeKePWfZJZIu8vLwghECZMmWQnp6OK1euAAB27NiBTp06IS4uDqmpqVi1ahUXViIyAgdWIiIiBTp16oQqVapgxowZZtnf7d/nIy32HCoEToHQ8J9/IlMZM2YMXFxcnrg/Ozv7wccuLi4YO3asObOICi3+i0VERKSAnZ0ddDodtm3bhhMnTph0X9npabg6ezJc6jRBsVZdTbovIlvXtm1b/P77708dWoGcYTUyMhJt2rQxbxhRIcWBlYiISJERI0bAwcEBwcHBJt3PrdWhyLhxGRX1U3kKIpEZtG3bFuHh4XByevQ6x05OTggPD+ewSpQHHFiJiIgUcXd3R79+/TB//nzcv3/fJPvITk3GtTlfocjrrVC0iZdJ9kFET4qPj4dWq4VGo4GzszM0Gg20Wi3i4+NVpxEVKhxYiYiIFDIYDEhMTMTixYtNsv245TOQeecGKvDoKpFZhYWFITk5GQ0aNMBvv/2GBg0aIDk5GXPmzFGdRlSoCAu4FuxLcXV1lUlJSaoziIiI8kVKiUaNGiE7OxvR0dEFOlQmJiaiZGVPuNZ+EzV+2VBg231YVGOTbJao0OvRowdatWqFUaNGQaPRICsrC//973+xa9cuREREqM4jKyOESJZSuqruMAUOrERERIrNnj0b/v7+2L17N95+++0C2+7kyZMxYcIE1Jp3AK513iyw7T6MAysRkXrWPLDylGAiIiLFBgwYgGLFihXoJW7u3LmD77//HsXb9DDZsEpERGRqHFiJiIgUc3V1ha+vL1asWIEbN24UyDanTZuGe/fuoULg5ALZHhERkQocWImIiCyAXq9HRkYGwsLCXnpbN27cwM8//wwfHx84V69XAHVERERqcGAlIiKyAK+++iq8vLwQEhKCzMzMl9rWN998g9TUVEyaNKmA6oiIiNTgwEpERGQhDAYDLl++jMjIyHxvIzY2FsHBwRg6dChq1qxZgHVERETmx4GViIjIQnTt2hUeHh4ICgrK9zamTp2K7OxsjB8/vgDLiIiI1ODASkREZCG0Wi0CAwOxadMmnDlzJs+vP3/+PMLCwuDv74+qVasWfCAREZGZcWAlIiKyIH5+ftBqtQgODs7zaydNmgStVosxY8aYoIyIiMj8OLASERFZkHLlyqF3796YO3cukpOTjX7diRMnsGjRIhgMBlSoUMGEhURERObDgZWIiMjCGAwGxMfHY+nSpUa/ZuLEiXB2dsbo0aNNWEZERGReHFiJiIgsTMuWLVG3bl3MmDEDUsoXPj86OhrLli3DqFGj4O7uboZCIiIi8+DASkREZGGEEDAYDDh8+DAOHDjwwuePHz8exYoVwyeffGKGOiIiIvPhwEpERGSBBg0ahKJFi77wEjd//vkn1q5di88++wwlSpQwUx0REZF5cGAlIiKyQEWLFsWQIUMQHh6OW7duPfN5Y8eORenSpfGvf/3LjHVERETmwYGViIjIQun1eqSlpWHOnDlPfXz79u3YsmULRo8ejaJFi5q5joiIyPSEMYs5WDJXV1eZlJSkOoOIiMgk2rRpg0uXLuHMmTOws7N7cL+UEq1atcK5c+dw7tw5ODs7P/X1jaNM2xfV2LTbJyKiFxNCJEspXVV3mAKPsBIREVkwg8GACxcuYOPGjY/cv3HjRuzevRtjx4595rBKRERU2PEIKxERkQXLyMhA5cqV0ahRI0RGRgLIObr65ptv4tatWzh9+jQcHBye+XoeYSUisn48wkpERERK2NvbIyAgAOvXr8eRI0fQs2dPLFmyBIcOHcKECROeO6wSEREVdjzCSkREZOGuXLmCKlWqoGPHjoiMjISHhwecnZ3xv//9D1qt9rmv5RFWIiLrZ81HWJ//rxwREREpV7FiRfTo0QO//fYbACA2NhZLlix54bBKRERU2PGUYCIiIgvl5eUFIQSEEFi5ciUyMzMBAEIIvPfeew8e8/LyUlxKRERkGhxYiYiILNSYMWPg4uLyxP0Pv53HxcUFY8eONWcWERGR2fA9rERERBZs27Zt8Pb2RnJy8hOPubi4IDIyEm3atDF/GBERWQxrfg8rj7ASERFZsLZt2yI8PBxOTk6P3O/k5ITw8HAOq0REZNU4sBIREVm4+Ph4aLVaaDQaODs7Q6PRQKvVIj4+XnUaERGRSXFgJSIisnBhYWFITk5GgwYN8Ntvv6FBgwZITk7GnDlzVKcRERGZFAdWIiIiC1esWDFMmzYNUVFRaN++PQ4ePIjvvvsObm5uqtOIiIhMiosuERERERERFWJcdImIiIiIiIjIzDiwEhERERERkUXiwEpEREREREQWiQMrERERERERWSQOrERERERERGSROLASERERERGRReLASkRERERERBaJAysRERERERFZJA6sREREREREZJE4sBIREREREZFF4sBKREREREREFokDKxEREREREVkkDqxERERERERkkTiwEhERERERkUUSUkrVDS9FCJENIEV1RwHQAshUHVFI8WuXf/zavRx+/fKPX7v849cu//i1ezn8+uUfv3b5x6+dcZyllFZ5MLLQD6zWQggRJaVsrLqjMOLX7v+1c/+hd9V1HMefr+373TTnmDGmczMLrExIac6Yldsw+mXSghAUy+H8RxhRin9oGCEWoVRIiMggl0ZJWUYz1LFKWeg0/Yo/Nr/zB1E6HX4z+zGJ1OW7Pz7v5e3bSO+553vO+d77esDhe+7nnvvh83nxveeezzmfc6pzdoNxftU5u+qcXXXObjDOrzpnV52zs6EchZuZmZmZmdns5wGrmZmZmZmZdZIHrN2xqe0GzGLOrjpnNxjnV52zq87ZVefsBuP8qnN21Tm7Eed7WM3MzMzMzKyTfIXVzMzMzMzMOskD1hki6QZJU5J29pSdJGmHpMck3SZpYZbPk7Q5yx+RtLbnMydn+dOSvitJLXSncTXmd2eW7ZJ0vaS5LXSnUc6uuhqzu1vSE5IezmVJC91plLMbTI35zZO0SdKTknZL+lwL3WlUHdlJOrznf+5hSS9KuqalLjVG0jGS7pI0mfv6L2X52yVtk/RU/j0iy5XHIk9LelTSiiw/VtJEZrdL0oVt9qsJdWWX712ddUxqBI71KmR3fH6fX5F0yZvVY0MoIrzMwAKsBlYAO3vKHgDW5PoG4Mpc3whszvUlwAQwJ1//DjgVEHAH8Km2+zbL8luYfwX8DDi77b45u+4uNWZ3N7Cy7f44u9mz1JjfFcDXc30OsLjtvs2W7KbVOQGsbrtvDWS3FFiR64cDTwInAFcDl2b5pcBVuX4G5VhEwCrg/iyfB8zP9QXAH4Cj2+7fLMnuQ8A9wNxcdgBr2+5fx7JbApwCfAO45M3qabt/XupffIV1hkTEduClacXvBbbn+jbgwJnvE4Bf5+emgL8CKyUtpQwadkT5Nt4EfHam294FdeSXr/+e24xRflCH/qZtZ1ddXdmNImc3mBrz2wB8M997PSJenMFmd0Ld/3uS3k05QP7tDDW5MyJib0Q8lOv7gElgGbAOuDE3u5E3jj3WATdFcR+wSNLSiHg1Il7JbeYzAjP46sqO8tt6CDnoB8aBFxrrSAv6zS4ipiLiAeC1t1iPDZmh36F0zE7gM7l+FnBMrj8CrJM0JuldwMn53jJgT8/n9zDaX8R+8wNA0lZgCtgH/LS55naKs6uuUnbA5pwe99Vhn971fzi7wfSVn6RF+f6Vkh6SdIukI5ttcmdU/d8DOAf4cZ4oHmKZ6/AAAAOHSURBVBmS3gl8ALgfODIi9kIZFFAG8FCOQZ7t+dh/jktyeuaj+f5VEfF8My1v3yDZRcQO4C5gby5bI2KymZa37y1m1289NmQ8YG3WBmCjpAnK1IVXs/wGyo7rQeAa4F5gP2XayHQj9QM6Tb/5ARARn6BMG5kPnN5kgzvE2VVXJbtzI+L9wGm5fKHRFneHsxtMv/mNAcuBeyJiBWVq4beabnRHVNrnpbOBmxtqZydIWkC59ePLPbNrDrrpQcoCICKejYgTgeOA9aNysmTQ7CQdB7yP8t1dBpwuaXX9Le2ePrJrpB7rrrG2GzBKImI38HEASe8BPp3l+4GLDmwn6V7gKeAvlB3YAcuBkTljOV2F/Ho/+09JWyjTTbY11eaucHbVVckuIp7Lv/sk/Qj4IGVK/0hxdoOpkN+fgX8AP8+3bgEuaLDJnVF1nyfpJGAsIiYabXCLJI1TDvZ/GBG3ZvELOdV3b05bncryPfz3Fen/OS6JiOcl7aKccBrqmTk1Zfd54L6IeDnrvINyj+t2hlif2fVbjw0ZX2FtkPJpl5LmAJcD1+frt0k6LNc/BuyPiMdzOsQ+SatyWtx5wC/aaX37+s1P0oLc4SFpjPLAg92tNL5lzq66CtmNSVqc5ePAmZTpiSPH2Q2mwm9GALcBa7OKjwKPN93uLug3u56PnsMIXV3NY4vvAZMR8Z2et7YA63N9PW8ce2wBzlOxCvhbDi6WSzo06zwC+DDwRCOdaEld2QHPAGty/zcOrKHcizm0KmTXbz02ZDRit2g0RtLNlIOGxZSb579GeXLextzkVuCyiIicd78VeB14DrggIv6Y9awEvg8cSnm63BdH4b6aOvLL6Ui/pExnnQv8Brgoz7APLWdXXU3ZHUY5Mz5Oye5XwMUR8a/metI8ZzeYGn8zjgV+ACwC/gScHxHPNNaRFtSVXdb1e+CMvEI79CR9hPJwqccomQB8hXIf4E+Ad1AGVGdFxEs5QLgW+CTlav75EfFgDv6/TZkeLODaiNjUaGcaVmN2c4HrKE+7DuDOiLi40c40rEJ2R1Gm8S/M7V+mPEDtxIPVExG3N9QVa4gHrGZmZmZmZtZJnhJsZmZmZmZmneQBq5mZmZmZmXWSB6xmZmZmZmbWSR6wmpmZmZmZWSd5wGpmZmZmZmad5AGrmZmZmZmZdZIHrGZmZmZmZtZJHrCamZmZmZlZJ/0bmBqwP2NQc9cAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1080x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_qt = df[df.Type == \"QT\"]\n",
    "qt_sort =df_qt.sort_values(by=\"Year\",ascending=True)\n",
    "qt_sort\n",
    "\n",
    "qt_sen_anew = list(qt_sort[\"sentiment_score_anew\"])\n",
    "qt_sen_iv = list(qt_sort[\"sentiment_score_iv\"]*100)\n",
    "year_label = list(qt_sort[\"Year\"])\n",
    "\n",
    "qt_rating = list(qt_sort[\"Rating\"])\n",
    " \n",
    "x=year_label\n",
    "y1=qt_sen_anew\n",
    "y2=qt_sen_iv\n",
    "bar_width = 0.3\n",
    "plt.rcParams['figure.figsize'] = (15.0,8.0) \n",
    " \n",
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(111)\n",
    "plt.bar(x=range(len(x)), height=y1, color='coral', alpha=0.8, width=bar_width)\n",
    "plt.bar(x=np.arange(len(x)) + bar_width, height=y2, color='deepskyblue', alpha=0.8, width=bar_width)\n",
    "\n",
    "ax1.set_ylabel('anew_score',fontsize='15')\n",
    "ax1.set_title(\"QT Moive Analysis\",fontsize='20')  \n",
    "#plt.yticks(fontsize=15)\n",
    "plt.xticks(range(len(x)), x)\n",
    "#plt.xticks(fontsize=15)\n",
    " \n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(range(len(x)), qt_rating, 'black', marker='*', ms=10)\n",
    "ax2.set_ylabel('Rating',fontsize='15')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Rating')"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_dis = df[df.Type != \"QT\"]\n",
    "dis_sort =df_dis.sort_values(by=\"Year\",ascending=True)\n",
    "dis_sort\n",
    "\n",
    "dis_sen_anew = list(dis_sort[\"sentiment_score_anew\"])\n",
    "dis_sen_iv = list(dis_sort[\"sentiment_score_iv\"]*100)\n",
    "year_label = list(dis_sort[\"Year\"])\n",
    "\n",
    "dis_rating = list(dis_sort[\"Rating\"])\n",
    "dis_rating\n",
    " \n",
    "x=year_label\n",
    "y1=dis_sen_anew\n",
    "y2=dis_sen_iv\n",
    "bar_width = 0.3\n",
    "plt.rcParams['figure.figsize'] = (15.0,8.0) \n",
    " \n",
    "fig = plt.figure()\n",
    "ax1 = fig.add_subplot(111)\n",
    "plt.bar(x=range(len(x)), height=y1, color='coral', alpha=0.8, width=bar_width)\n",
    "plt.bar(x=np.arange(len(x)) + bar_width, height=y2, color='deepskyblue', alpha=0.8, width=bar_width)\n",
    "\n",
    "ax1.set_ylabel('anew_score',fontsize='15')\n",
    "ax1.set_title(\"Disney Moive Analysis\",fontsize='20')  \n",
    "plt.yticks(fontsize=15)\n",
    "plt.xticks(range(len(x)), x)\n",
    "plt.xticks(fontsize=15)\n",
    " \n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(range(len(x)), dis_rating, 'black', marker='*', ms=10)\n",
    "ax2.set_ylabel('Rating',fontsize='15')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Regression:\n",
    "### Topic Modelling:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ALL：\n",
    "from sklearn import feature_extraction  \n",
    "from sklearn.feature_extraction.text import TfidfTransformer  \n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn.feature_extraction.text import HashingVectorizer \n",
    "from sklearn.metrics.pairwise import cosine_similarity\n",
    "from sklearn.decomposition import NMF\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.metrics import precision_recall_fscore_support\n",
    "import numpy as np\n",
    "from sklearn.cluster import KMeans\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn import metrics\n",
    "from sklearn import preprocessing\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic #0: one like back two know see little go hand would\n",
      "Topic #1: drschultz big back like one two know nigger horse black\n",
      "Topic #2: know get like back got one int see go room\n",
      "Topic #3: back two get one like int see door room cut\n",
      "Topic #4: get like back one know two gonna got door youre\n",
      "Topic #5: japanese sword oren yuki cu one back black english go\n",
      "Topic #6: sex machine door back fuckin get pussy big chet vampires\n",
      "Topic #7: nazi french cinema back collanda one ltaldo major table donny\n",
      "Topic #8: drschultz horse nigger slave men big candyland slaves roy greenville\n",
      "Topic #9: know get back got day like gonna one tell youre\n",
      "Topic #10: like know get one back int got room youre gonna\n",
      "Topic #11: room door back one looks youre theodore wife like phone\n",
      "Topic #12: zzzzzz fourstory flowery flowers flower flow flourishcompletely flourish flour flounder\n",
      "Topic #13: bag know contd money got get looks back one black\n",
      "Topic #14: virgil get know day youre like got room nicholson shit\n",
      "Topic #15: wurlitzer camera day grace roger cu os deputies back cut\n",
      "Topic #16: japanese cu oren sword yuki english back hanzo one floor\n",
      "Topic #17: marsellus fuckin like gonna room shit youre int os one\n",
      "Topic #18: fuckin nice fuck know youre car got warehouse cu cops\n",
      "Topic #19: back money looks contd day bag get know closeup look\n",
      "\n"
     ]
    }
   ],
   "source": [
    "count = CountVectorizer()\n",
    "count_matrix = count.fit_transform(qt_cleaned_data)\n",
    "\n",
    "\n",
    "n_components = 20  \n",
    "n_top_words = 10\n",
    "\n",
    "def get_words(model, feature_names, n_top_words):\n",
    "    for topic_idx, topic in enumerate(model.components_):\n",
    "        message = \"Topic #%d: \" % topic_idx\n",
    "        message += \" \".join([feature_names[i]\n",
    "                             for i in topic.argsort()[:-n_top_words - 1:-1]])\n",
    "        print(message)\n",
    "    print()\n",
    "\n",
    "nmf = NMF(n_components=n_components, random_state=1,alpha=.1, l1_ratio=.5).fit(count_matrix)\n",
    "\n",
    "cv_feature_names = count.get_feature_names()\n",
    "\n",
    "get_words(nmf, cv_feature_names, n_top_words)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Topic #0: back looks door one like go night know int around\n",
      "Topic #1: contd back looks day like door love ext int see\n",
      "Topic #2: shanyu back crikee man chienpo khan looks sword cut look\n",
      "Topic #3: contd oh little clawhauser big right bunny flash well youre\n",
      "Topic #4: oh well time yes one see go good say father\n",
      "Topic #5: hey ha okay right oh get go dad gasps come\n",
      "Topic #6: michael oh time children step go chim sir diddle well\n",
      "Topic #7: thou night father close cont ext cut love car laurence\n",
      "Topic #8: oh get fish one youre gonna little got see look\n",
      "Topic #9: zut flicks flip flint flings flinging flinches flinch flimsy flight\n",
      "Topic #10: bill int contd kids house continuous plummer youre principal back\n",
      "Topic #11: contd int youre later oh little kids looks gonna know\n",
      "Topic #12: zut flicks flip flint flings flinging flinches flinch flimsy flight\n",
      "Topic #13: ha oh hey fish get mine dad okay swim gonna\n",
      "Topic #14: frozen sven lee contd snow arendelle castle right trolls okay\n",
      "Topic #15: carpet back lamp oh like looks one turns ali princess\n",
      "Topic #16: carpet looks lamp back head princess ali oh palace pulls\n",
      "Topic #17: lumire castle night contd int one village philippe garderobe see\n",
      "Topic #18: sebastion scuttle got la sea girl max oh know kiss\n",
      "Topic #19: contd oh little youre later big right hey okay gonna\n",
      "\n"
     ]
    }
   ],
   "source": [
    "count = CountVectorizer()\n",
    "count_matrix = count.fit_transform(disney_cleaned_data)\n",
    "\n",
    "\n",
    "n_components = 20\n",
    "n_top_words = 10\n",
    "\n",
    "def get_words(model, feature_names, n_top_words):\n",
    "    for topic_idx, topic in enumerate(model.components_):\n",
    "        message = \"Topic #%d: \" % topic_idx\n",
    "        message += \" \".join([feature_names[i]\n",
    "                             for i in topic.argsort()[:-n_top_words - 1:-1]])\n",
    "        print(message)\n",
    "    print()\n",
    "\n",
    "nmf = NMF(n_components=n_components, random_state=1,alpha=.1, l1_ratio=.5).fit(count_matrix)\n",
    "\n",
    "cv_feature_names = count.get_feature_names()\n",
    "\n",
    "get_words(nmf, cv_feature_names, n_top_words)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Machine Learning Analysis:\n",
    "### 1. Predict if a movie is a Tarantino movie or a Disney movie:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "# The whole data:\n",
    "all_data = qt_data + disney_data # QT + Disney total: 20\n",
    "# Split data into several small data:\n",
    "def sp_text(x):\n",
    "    sen = []\n",
    "    for i in range(len(x)):\n",
    "        sen += re.findall(r'.{200}',x[i]) # split the data by the length, each small string length is 200\n",
    "    return sen"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "qt_split_data = sp_text(qt_data)\n",
    "dis_split_data = sp_text(disney_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(12418, 2)"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "qt_sp_df = pd.DataFrame(qt_split_data)\n",
    "qt_sp_df.columns = ['Script']\n",
    "qt_sp_df.insert(0,'Type',\"0\")\n",
    "dis_sp_df = pd.DataFrame(dis_split_data)\n",
    "dis_sp_df.columns = ['Script']\n",
    "dis_sp_df.insert(0,'Type',\"1\")\n",
    "all_sp_df = qt_sp_df.append(dis_sp_df)\n",
    "all_sp_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "y_label = [int(yl) for yl in list(all_sp_df.Type)]\n",
    "\n",
    "#依次导入X， y， test_size split the textual data:\n",
    "X_train, X_test, y_train, y_test = train_test_split(all_sp_df.Script, y_label, test_size=0.3,stratify=all_sp_df.Type)\n",
    "# stratify 处理不均衡的数据集"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8692, 281499)"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "\n",
    "cv = CountVectorizer(ngram_range = (2,3),stop_words = 'english')\n",
    "\n",
    "#fit_transform是fit和transform两种方法的简写\n",
    "#fit方法用于构建特征空间（构建词典）\n",
    "#transform方法使用该空间将文本数据转化为特征矩阵\n",
    "X_train_cv = cv.fit_transform(X_train)\n",
    "X_train_cv.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.93      0.98      0.96      2318\n",
      "           1       0.96      0.88      0.92      1408\n",
      "\n",
      "    accuracy                           0.94      3726\n",
      "   macro avg       0.95      0.93      0.94      3726\n",
      "weighted avg       0.94      0.94      0.94      3726\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import classification_report\n",
    "\n",
    "#用学到的特征空间去对测试集数据进行转换为特征矩阵\n",
    "X_test_cv = cv.transform(X_test)\n",
    "#测试集特征矩阵的列数与训练集特征矩阵的列数是一致的，都是19985\n",
    "X_test_cv\n",
    "# Navie Bayes:\n",
    "# 多项式朴素贝叶斯：sklearn.naive_bayes.MultinomialNB(alpha=1.0, fit_prior=True, class_prior=None)\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "nb = MultinomialNB()\n",
    "#这里是有监督的训练模型\n",
    "#所以要同时输入X_train_cv和y_train\n",
    "nb.fit(X_train_cv, y_train)\n",
    "#训练好的模型，对测试集数据进行预测\n",
    "predicted_cv = nb.predict(X_test_cv)\n",
    "print(classification_report(y_test, predicted_cv, labels=[0,1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, '$ROC curve$')"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score,roc_curve\n",
    "\n",
    "fpr,tpr,thresholds=roc_curve(y_test,predicted_cv)  #计算fpr,tpr,thresholds\n",
    "auc=roc_auc_score(y_test,predicted_cv) #计算auc\n",
    "\n",
    "#画曲线图\n",
    "plt.figure()\n",
    "plt.plot(fpr,tpr)\n",
    "plt.title('$ROC curve$')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.80      1.00      0.89      2318\n",
      "           1       0.99      0.58      0.73      1408\n",
      "\n",
      "    accuracy                           0.84      3726\n",
      "   macro avg       0.90      0.79      0.81      3726\n",
      "weighted avg       0.87      0.84      0.83      3726\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# TF-IDF:\n",
    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
    "\n",
    "\n",
    "tv = TfidfVectorizer(ngram_range = (2,3),stop_words='english')\n",
    "X_train_tf = tv.fit_transform(X_train)\n",
    "X_test_tf = tv.transform(X_test)\n",
    "nb2 = MultinomialNB()\n",
    "nb2.fit(X_train_tf, y_train)\n",
    "predicted_tf = nb2.predict(X_test_tf)\n",
    "print(classification_report(y_test, predicted_tf, labels=[0,1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr,tpr,thresholds=roc_curve(y_test,predicted_tf)  #计算fpr,tpr,thresholds\n",
    "auc=roc_auc_score(y_test,predicted_tf) #计算auc\n",
    "\n",
    "#画曲线图\n",
    "plt.figure()\n",
    "plt.plot(fpr,tpr)\n",
    "plt.title('$ROC curve$')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.74      1.00      0.85      2318\n",
      "           1       1.00      0.42      0.60      1408\n",
      "\n",
      "    accuracy                           0.78      3726\n",
      "   macro avg       0.87      0.71      0.72      3726\n",
      "weighted avg       0.84      0.78      0.75      3726\n",
      "\n"
     ]
    }
   ],
   "source": [
    "## Logistic Regression: \n",
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.metrics import classification_report\n",
    "\n",
    "cv = CountVectorizer(ngram_range = (2,3),stop_words='english')\n",
    "X_train_counts = cv.fit_transform(X_train)\n",
    "X_test_counts = cv.transform(X_test)\n",
    "\n",
    "\n",
    "LR = LogisticRegression()\n",
    "LR.fit(X_train_counts, y_train)\n",
    "predicted_lr = LR.predict(X_test_counts)\n",
    "predicted_lr_pb = LR.predict_proba(X_test_counts)\n",
    "print(classification_report(y_test, predicted_lr, labels=[0,1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fpr,tpr,thresholds=roc_curve(y_test,predicted_lr)  #计算fpr,tpr,thresholds\n",
    "auc=roc_auc_score(y_test,predicted_lr) #计算auc\n",
    "\n",
    "#画曲线图\n",
    "plt.figure()\n",
    "plt.plot(fpr,tpr)\n",
    "plt.title('$ROC curve$')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2oAAAHiCAYAAACZVz+XAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3dfZTedX3n/9c7k0xIQgKBBARCCAgoAUEwIlbbWkULni60v9pWurZ1i+Luqeu2dbtrqz/t0e62dburp1t7Y6u19ferSNufllosXVutaKsFRJFbSbmRcBswBELIzWQ+vz9miMMwSa7AZK7vzPV4nJPDXN/rM9f1Jud7Jnnme3NVay0AAAB0x7x+DwAAAMBTCTUAAICOEWoAAAAdI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINgE6oquVV1apqS1Vtraq7quriSWuWVNWvVdW/VtVjVXVTVb1l0pqlVfXfq2r9+Jo7qup3qmrlzP4fAcAzJ9QA6IoXJnmotXZwa21xkl9O8gdVtSIZC7kkX0pyfJJXJVmW5M1J3vdk0FXVoUmuSvL8JOe31pYm+d4kC5IcN53DVtXQdL4eAEwk1ADoihcm+dqEx/+YZCjJ8vHHv53kgSRvaK3d2cZ8Ocn/SvJz42s+kOQ7SV7XWrstSVprG1prb2mtXfPkC1fVRVV1bVVtHj8694qq+m9V9cEJa1ZV1eNVNW/88Zuq6u+q6iNVtSnJu8aPAK6Y8D1rq+qBqlo2/vjN40f9NlfVZ6vqiGn9HQNgzhJqAHTFmUmuTXYfGfv18cfrq2pNkp9M8iuttTbp+/41yfFVdWySn0ryztba6J7epKrenuRdGTsatzzJDye5M2Oh+I0JS89IcuOE1zo9yUuT/FWSw5P8RpINSdZO+J7/luQ3W2uPVtWvJPn3SS5IsjLJPUl+rbffCgAGnVADoCtemOQ/VdWjSTYlOSLJeeNh9pokd7fWvjbF9x2TZGOSc5NsbK39857eYPw6tfck+cnW2tdaa6OttW+21u4cf//rJyw/Y4rHv9Vau3z8+7YnuSHJKeOvfXaSs5J8aPzI2buSXNRaW99a25HkI0levJ+/JwAMKKEGQN9V1cKMBc/prbVlSV6X5JwkO8eXrMzY0aup/EiSK5McmeTb+3irc5N8s7U28cjZkwF3ZJIbJ2w+I089wnZ6kj+f9Ho35LtH1H49ya+OB9yrkgwn+ZeqeqSqHknyt0k272M+AEgi1ADohtOSbE9ye5K01v4yY9H1o+PP35HkuCevF3tSVb06yYsydp3at5McM3nNJIcleWSK7acmua21tm38decn+YGMH1GrquMydkOSWyZ93w1J1lbVuUmOSvKnE97nU621Qyf8OqS19oq9/i4AwDihBkAXnJnkhknXn12Rseu7kuTy8f/+WlUtrqqFVfWGJJcm+dnW2h1JPjO+5jeqallVLaiqF4zf/OPJW/Nfl+TlVXVGjTmpqk5JUkkWV9X88dB7f8aO4j156uMZGTsSN/natyePqP33jF0/t2t8+9eS/EBVnZUk4/NcWFX1bH6TABgcQg2ALph8fVgydqrgq6vqoNbaloydTnh6kruS3J/k3yb5odbaZUnSWns0ySuTnJzktiQPZyzkHmitbRxf808Zu6HHZ5I8luRTSRZl7Jb+12fsiNn/ydjRuQ2ttU3js5yR5OtTzH1Tkuck2dVa+/STG8evk3tvkr+sqi3j686bFKIAsEflzwwAAIBucUQNAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOmZ+v954xYoVbc2aNf16ewAAgL669tprH2qtrZzqub6F2po1a3LNNdf06+0BAAD6qqru2tNzTn0EAADoGKEGAADQMUINAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOEGgAAQMcINQAAgI4RagAAAB2zz1Crqo9W1YNVdcMenq+q+u2qWl9V11fVWdM/JgAAwODo5Yjax5Kct5fnz09y0vivS5L83rMfCwAAYHDN39eC1toXq2rNXpZcmORPW2styVeq6tCqOqq1dt80zQgAQEdtH9mVbTtH+z0G7NXQvMrBC/eZPp0yHdMek+TuCY83jG8TagDwDD2ydUduf+jxfo8xJz2weVu+ec/mLBgaO7Hoa9/elEe27szQvOrzZLPPtp27csv9j/V7DNinFx57aD79cy/r9xj7ZTpCbaqfam3KhVWXZOz0yKxevXoa3hqA2WTbzl354rc25pEnds74e//rxi25fePjGZ7/zO+jdfN9j+auh7dmeGhe5h/gv9Q/tn3kgL4+T/d9J6/s9wizzrJFC7LsoAV53nOWZs2KJf0eB/Zo5dKF/R5hv01HqG1IcuyEx6uS3DvVwtbah5N8OEnWrVs3ZcwBMPO27hjJN+7enDb1v7Pt9o27N+eRrTty032P5p5NT+x39HThX97XHL74mR85acmiBUP54TOPzvDQ0PQO9rS3alm+eDinrzrkgL7PoDr60EU5+cil/R4DYI+mI9QuT/LWqro0yUuSbHZ9GkA3bXp8Rx7b9vQjNRf/ydW57cEtPb/OvEpGW/IDz1u5+/SxXhxz6KKsXLowP/mS1TlsyXDP3zddDlm0IEsPWjDj7wsA+2ufoVZVn0jyiiQrqmpDkvckWZAkrbXfT3JFktcmWZ9ka5J/d6CGBZjtRkefesRq5+horr5jU3bu+u6F+KOt5Ypv3p+lB+3/v6X9/S0PZF5V5tXTjxg9tm0kD23Zvtfvv/SSc/b5HicdcXAOP3j2nUICALNJL3d9vGgfz7ckPzdtEwFMct/mJ55yFGjHyGiuuu2hLBia/muE/vr6+7LsGQRSL27f+HjueeSJ/fqeQxbt39GfkV2jeWLnrvzQ6UdP+fzWHSM5c/XyPGfZQU/ZPtpavufEFTnm0EX79X4AwIExu+5RCcwKu0Zbto/s2uuaBx7dnpvve/RpdyP659sfzo6R7x5d2rDpiXxp/UMHYMq9O2v1odP+mkcsW5iheZXzTntOlgx/98fvyOhoXvn8I1ITjoIND83LKUctfco2AGBwCDUYYJse35ENm757hGfbyNgd+RaO3yCiteRT192Tw5YMZ3964eo7Nz3r2Y5cNnZq3a7RZMnwUP799z83J6w8ePfzw/Pn5ezjD8t033ivavZ9zgoAMPf42wjMUdt27srN9z2aOx9+PF+67eHsGh3NZ66/L4uGh3bfVnzT1t5ukb5h0xNZt2Z5z+99zgmHZcnw/Jx9/GF7XXfYkuG8YIo72q05fEkOWnBg76gHANBlQg1mkfs3b8vdm7bmiR278o/f2phFk2LmC996MEkysqtNeRv0Yw9blNaSVz7/iCRj1yWtPPignHr0st1rFi8cyrrjDtt9BK2SzN+Pu/oBAPDsCTXogNseeCzv+vQNu29Xfs1dm/LQlu05aP7QU0453Lrj6dd9Tfw8qF3jdxR85fOPyOEHD+e5Kw/O9560MicdcXCOWb5ov26jDgBA/wg1mGG7Rlvef+Ut2bVrLKo+d/MDufPhrbufP/nIg3PoogXZMTKaH3vRqqeE2mgb+xyqk49cmoMPmp8zVh3iZhMAAHOQUGOg7dw1mv/nK3c97c6DB8pfX39fvnnP5t13NVwyPJSd48H2vgtPzU+9dM0MTQIAQJcJNQbOrtGWXaMtf/zlO/Lrn72lLzOcctSyXPaWc7L0oP37jCwAAAaDUGNOaa3l63c/ki/d9lAWzH/69VgPb9meP7zqjqdsW3fc8vz+T70oQzNwCuG8qhyyWJwBALB3Qo1Z58Z7N+fT192Tr9/9SG6459EsWTiUeeOR9eBj23t6jVevPTIvPPbQnHfac/LcCZ/NBQAAXSDU6JxHtu7IHQ89ntE2dqONhfPn5dq7NmXzEzszNK9y3bcf2b12eGheXrzmsBw64SjVo9tG8m/PXp0zV0/9uV/z5iUL5/uMLgAAukuo0VcPb9men//k13Po4uHd2/76G/fucf33nrQiLzvx8Pzgqc/Jj5x5jGu8AACYk4Qa06q1lo986Y48tGXH05774rc2ZrS13beTH9k1mtse3LL7+RNWLEmSrDl8cU4+cmkuesnqHDR/KC9es3z3Z4W5FT0AAINAqPGsbXp8R+58+PH8/c0P5nc+v3739uFJN/N48pb0r1575O5tRx26KGesOiRvf83zZmZYAACYBYQa++2bGzbn41+5M4uH5+fSq7+dbTtHn/L8uuOW5w9+6kU5/OCFfZoQAABmN6FGTx7esj1//Y178+ufvSXbR74bZouHx27K8V/Pe35OOuLgvPj4w3LIIteNAQDAsyHUmNITO3YlSa66bWMu+fi1T3v+4xefnZc9d0XmzXPNGAAATDehRpLklvsfzRXX35eFC4byR1fdnk1bdz7l+dWHLc7rzz42rztrVVYuXeimHgAAcAAJNfK5mx7Im/70mqdtf8f5z0+SvHjN8rzouMNmeiwAABhYQm3A/d2N9+8+tfFX/83aXPSS1alUFgyVo2YAANAnQm1Abd0xkvd95qZ84l/uzilHLcvbXnlizn/BUf0eCwAAiFAbSN+4+5H8/Ce/njsffjxv+f4T8vZXP+9pn3kGAAD0j1AbILtGW37vC+vzwc/dliOWLsyfvemcvPS5h/d7LAAAYBKhNiDu/s7W/MInv55r7tqUf3PG0fm1C0/LIYt93hkAAHSRUJvjWmv5/752T95z+Y2pJB/8iRfmh888pt9jAQAAeyHU5rBHtu7IOz99Q/7m+vty9prD8r9+4oysWr6432MBAAD7INTmqH9a/1B+8bJv5KEt2/Nfznte3vJ9z83QPLfbBwCA2UCozTHbR3blt668NX941R05YeWSfOqnX5YXrDqk32MBAAD7QajNId964LG87RPX5Zb7H8sbzlmdd752bRYND/V7LAAAYD8JtTmgtZaP/dOd+fXP3pKlC+fnIz+zLq865ch+jwUAADxDQm2We/DRbfnPf3F9vvitjXnl84/Ib/7o6Vm5dGG/xwIAAJ4FoTaLXXnj/XnHX16fJ3buyvt++LS84SWrU+WGIQAAMNsJtVlkdLTlvZ+5KZddc3eWHbQg9z+6Lacdsywf/Ikzc+IRB/d7PAAAYJoItY66+ztbs/mJnRltLb/059fnkEULcvP9j+axbSNJkhcdtzw/8z1rcvHLj8/w/Hl9nhYAAJhOQq1jRnaN5jUf/GJu3/j4054754TDsmNkNP/zx1+Y41cs6cN0AADATBBqHfOZ6+/bHWm/+aMvyPLFw1m4YCgvP3GFD6wGAIABIdT67KZ7H83//ofbMq8qf/PN+3Zv/8J/fkXWOGoGAAADSaj10YZNW/Pa374qSfKcZQflhJVL0lrysy9bI9IAAGCACbU++fR19+TnP/n1JMmPnHlMPvATL+zzRAAAQFcItT54/9/ekt/9wr8mSQ5bMpz3v+70Pk8EAAB0iVCbYa213ZH2uV/8vpx4xNI+TwQAAHSND+CaYd+8Z3OS5JSjlok0AABgSkJtBm3ZPpILfufLSZJf+sGT+zwNAADQVUJtBv3jrRuTJFXJy09c2edpAACArhJqM+jRbTuTJP/nF74vw/P91gMAAFNTCzPosmvuTpIsHnYPFwAAYM8UwwzY/MTO/Njv/1O+9cCWJMnRhy7q80QAAECXCbUD7HM3PZA3/ek1ux9/4CfO6OM0AADAbCDUDqDR0bY70t74PWvyK689xbVpAADAPqmGA+ThLdtzwq9ckSQ57vDF+dULThVpAABAT5TDAXLL/Y8lSVYuXZj/900v6fM0AADAbCLUDpDtI7uSJB/6ybOyavniPk8DAADMJkLtALj2ru/kZz82dm3a4uGhPk8DAADMNkLtALj86/cmSV56wuE59ehlfZ4GAACYbdz1cZpt3TGSP/nnu7JowVA+cck5/R4HAACYhRxRm0Zfvf3hrH33lUmSVct9qDUAAPDMCLVp9Hc3PZAkeeXzj8ifvdnRNAAA4Jlx6uM0qiTD8+flo298cb9HAQAAZjGhNk0e27Yzf/SlO/o9BgAAMAc49XGaXPftR5IkZ6w6pM+TAAAAs51QmwZfuPXB/PRH/yVJ8p4LTu3zNAAAwGwn1J6lTY/vyBv/+OokY3d6PGv18j5PBAAAzHauUXuWHnxse5LkP7ziufmv5z2/z9MAAABzgSNqz9LDW8ZC7QXHuDYNAACYHkLtWfqNv70lSXLksoV9ngQAAJgrhNqzsG3nrly/YXOS5EXHHdbnaQAAgLlCqD0L//itjUmS/+usY/o8CQAAMJcItWfhlvseS5Jc/PLj+zwJAAAwlwi1Z2Fejf33eUcu7e8gAADAnCLUnoW/+sa9/R4BAACYg3oKtao6r6purar1VfWOKZ5fXVWfr6rrqur6qnrt9I/aPesf3JIkmT+kdwEAgOmzz8KoqqEkH0pyfpK1SS6qqrWTlr0ryWWttTOTvD7J7073oF3zR1fdniQ5+3h3ewQAAKZXL4eCzk6yvrV2e2ttR5JLk1w4aU1Lsmz860OSzOlzAr9020P5tb+5OUnyjvOf3+dpAACAuWZ+D2uOSXL3hMcbkrxk0ppfTfJ3VfUfkyxJcu60TNdBrbW84SNfTZL8/htelLNWL+/zRAAAwFzTyxG1mmJbm/T4oiQfa62tSvLaJB+vqqe9dlVdUlXXVNU1Gzdu3P9pO+C+zdt2f33eac/p4yQAAMBc1UuobUhy7ITHq/L0UxsvTnJZkrTW/jnJQUlWTH6h1tqHW2vrWmvrVq5c+cwm7rM3/vG/JEl+68fO6PMkAADAXNVLqF2d5KSqOr6qhjN2s5DLJ635dpJXJUlVnZKxUJudh8z2YuuOkXzrgbE7PV5wxtF9ngYAAJir9hlqrbWRJG9NcmWSmzN2d8cbq+q9VXXB+LK3J3lzVX0jySeSvLG1Nvn0yFlvy/aRJMnPvPS4DM93S34AAODA6OVmImmtXZHkiknb3j3h65uSvGx6R+uev7z2niTJSUcu7fMkAADAXOaw0H74q6+PhdprTj2yz5MAAABzmVDbD4csWpBjD1uUI5Ye1O9RAACAOUyo7aejD1nU7xEAAIA5TqgBAAB0jFADAADoGKHWo9s3bslX7/hORufepw4AAAAdI9R69LZLr0uSnHXc8j5PAgAAzHVCrQePbduZG+55NEnyC+ee3OdpAACAuU6o9WB0/GzH//uH1uagBUP9HQYAAJjzhBoAAEDHCDUAAICOEWq9cKNHAABgBgm1Hvz9LQ/0ewQAAGCACLUebB8ZTZKce8oRfZ4EAAAYBEKtB7vGb/u4cL47PgIAAAeeUOvBuz59Q5JkwVD1eRIAAGAQCLV9eHz7yO6vDz94YR8nAQAABoVQ24cnb/j4ztee0tc5AACAwSHUAAAAOkaoAQAAdIxQ24d/uOXBfo8AAAAMGKG2Fzt3jeZtn7guSbL68MV9ngYAABgUQm0vnvz8tDecszo/eOpz+jwNAAAwKIRaD4451NE0AABg5gg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCbS+u37A5STLaWp8nAQAABolQ24sNm7YmSc489tA+TwIAAAwSodaDY5Yv6vcIAADAABFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOEGgAAQMcINQAAgI4Ranvxzk/dkCSZV9XnSQAAgEEi1PZgZNdonti5K0myavmiPk8DAAAMEqG2B9fetSlJ8rZXnphyRA0AAJhBQm0P7n90W5Lke05c0edJAACAQSPU9mHl0oX9HgEAABgwQm0PvnDrxn6PAAAADCihNoXWWj513T1JkqMOOajP0wAAAINGqO3FD556ZBYPz+/3GAAAwIARantxylHL+j0CAAAwgIQaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOEGgAAQMcINQAAgI4RagAAAB0j1AAAADpGqAEAAHSMUAMAAOgYoQYAANAxQg0AAKBjhBoAAEDHCDUAAICOEWpTaK3fEwAAAINMqE3hy//6UJJkZJdiAwAAZl5PoVZV51XVrVW1vqresYc1P15VN1XVjVX1Z9M75sx69ImRJMm5a4/s8yQAAMAgmr+vBVU1lORDSV6dZEOSq6vq8tbaTRPWnJTkl5O8rLW2qaqOOFADz6TFw0P9HgEAABhAvRxROzvJ+tba7a21HUkuTXLhpDVvTvKh1tqmJGmtPTi9Y86sz938QL9HAAAABlgvoXZMkrsnPN4wvm2ik5OcXFVfrqqvVNV50zVgP+zYNZokWXP4kj5PAgAADKJ9nvqYpKbYNvkuG/OTnJTkFUlWJbmqqk5rrT3ylBequiTJJUmyevXq/R52Jj135ZIMz3evFQAAYOb1UiIbkhw74fGqJPdOseavWms7W2t3JLk1Y+H2FK21D7fW1rXW1q1cufKZzgwAADCn9RJqVyc5qaqOr6rhJK9PcvmkNZ9O8gNJUlUrMnYq5O3TOSgAAMCg2GeotdZGkrw1yZVJbk5yWWvtxqp6b1VdML7syiQPV9VNST6f5Jdaaw8fqKEBAADmsl6uUUtr7YokV0za9u4JX7ckvzj+CwAAgGfB3TIAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOE2hT+5vr7Mtr6PQUAADCohNoeHLxwfr9HAAAABpRQm8K8Sl7xvJX9HgMAABhQQg0AAKBjhBoAAEDHCDUAAICOEWoAAAAdI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOEGgAAQMcINQAAgI4RagAAAB0j1AAAADpGqAEAAHSMUAMAAOgYoQYAANAxQg0AAKBjhNok3354a0Zb0lq/JwEAAAaVUJvkpvseTZKsPnxxnycBAAAGlVDbg9OOPqTfIwAAAANKqAEAAHSMUAMAAOgYoQYAANAxQg0AAKBjhBoAAEDHCDUAAICOEWoAAAAdI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY4QaAABAxwi1ST5/y4P9HgEAABhwQm2SbSO7kiQnrFzS50kAAIBBJdSmsObwxTlowVC/xwAAAAaUUAMAAOgYoQYAANAxQm2S7zy+I6Ot31MAAACDTKhNctVtD+Xx7SP9HgMAABhgQm0KLznhsH6PAAAADDChNsnw0Lwcd7hb8wMAAP0j1AAAADpGqAEAAHSMUAMAAOgYoQYAANAxPYVaVZ1XVbdW1fqqesde1r2uqlpVrZu+EQEAAAbLPkOtqoaSfCjJ+UnWJrmoqtZOsW5pkrcl+ep0DwkAADBIejmidnaS9a2121trO5JcmuTCKda9L8n7k2ybxvkAAAAGTi+hdkySuyc83jC+bbeqOjPJsa21z0zjbAAAAAOpl1CrKba13U9WzUvygSRv3+cLVV1SVddU1TUbN27sfUoAAIAB0kuobUhy7ITHq5LcO+Hx0iSnJflCVd2Z5Jwkl091Q5HW2odba+taa+tWrlz5zKcGAACYw3oJtauTnFRVx1fVcJLXJ7n8ySdba5tbaytaa2taa2uSfCXJBa21aw7IxAAAAHPcPkOttTaS5K1Jrkxyc5LLWms3VtV7q+qCAz0gAADAoJnfy6LW2hVJrpi07d17WPuKZz8WAADA4OrpA68BAACYOUINAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOEGgAAQMcINQAAgI4RagAAAB0j1AAAADpGqAEAAHSMUAMAAOgYoQYAANAxQg0AAKBjhBoAAEDHCDUAAICOEWoAAAAdI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOEGgAAQMcINQAAgI4RagAAAB0j1AAAADpGqAEAAHSMUAMAAOgYoQYAANAxQg0AAKBjhBoAAEDHCDUAAICOEWoAAAAdI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOEGgAAQMcINQAAgI4RagAAAB0j1AAAADpGqAEAAHSMUAMAAOgYoQYAANAxQg0AAKBjhBoAAEDHCDUAAICOEWoAAAAdI9QAAAA6pqdQq6rzqurWqlpfVe+Y4vlfrKqbqur6qvr7qjpu+kcFAAAYDPsMtaoaSvKhJOcnWZvkoqpaO2nZdUnWtdZOT/IXSd4/3YMCAAAMil6OqJ2dZH1r7fbW2o4klya5cOKC1trnW2tbxx9+Jcmq6R0TAABgcPQSasckuXvC4w3j2/bk4iSffTZDAQAADLL5PaypKba1KRdWvSHJuiTfv4fnL0lySZKsXr26xxEBAAAGSy9H1DYkOXbC41VJ7p28qKrOTfLOJBe01rZP9UKttQ+31ta11tatXLnymcx7QK1/8LHs2DWa0dEpOxQAAGBG9BJqVyc5qaqOr6rhJK9PcvnEBVV1ZpI/yFikPTj9Y86MOx8au8xu7dHL+jwJAAAwyPYZaq21kSRvTXJlkpuTXNZau7Gq3ltVF4wv+x9JDk7y51X19aq6fA8vNyucsOLgfo8AAAAMsF6uUUtr7YokV0za9u4JX587zXMBAAAMrJ4+8BoAAICZI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoTXDljfcnSar6PAgAADDQhNoEu1pLkpx85NI+TwIAAAwyoTbJquWLMjzfbwsAANA/igQAAKBjhBoAAEDHCDUAAICOEWoAAAAdI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY4QaAABAxwg1AHnS8ugAAAeBSURBVACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOEGgAAQMcINQAAgI4RagAAAB0j1AAAADpGqAEAAHSMUAMAAOgYoQYAANAxQg0AAKBjhBoAAEDHCDUAAICOEWoAAAAdI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOEGgAAQMcINQAAgI4RagAAAB0j1AAAADpGqAEAAHSMUAMAAOgYoQYAANAxQg0AAKBjhBoAAEDHCDUAAICOEWoAAAAdI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY3oKtao6r6purar1VfWOKZ5fWFWfHH/+q1W1ZroHBQAAGBT7DLWqGkryoSTnJ1mb5KKqWjtp2cVJNrXWTkzygSS/Od2DAgAADIpejqidnWR9a+321tqOJJcmuXDSmguT/Mn413+R5FVVVdM3JgAAwODoJdSOSXL3hMcbxrdNuaa1NpJkc5LDp2NAAACAQTO/hzVTHRlrz2BNquqSJJckyerVq3t465l13GFLsmXbSL/HAAAABlwvobYhybETHq9Kcu8e1myoqvlJDknynckv1Fr7cJIPJ8m6deueFnL99p/OPanfIwAAAPR06uPVSU6qquOrajjJ65NcPmnN5Ul+Zvzr1yX5h9Za50IMAABgNtjnEbXW2khVvTXJlUmGkny0tXZjVb03yTWttcuTfCTJx6tqfcaOpL3+QA4NAAAwl/Vy6mNaa1ckuWLStndP+Hpbkh+b3tEAAAAGU08feA0AAMDMEWoAAAAdI9QAAAA6RqgBAAB0jFADAADoGKEGAADQMUINAACgY4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DFCDQAAoGOqtdafN67amOSuvrz53q1I8lC/h2DOsn9xINm/ONDsYxxI9i8OpK7uX8e11lZO9UTfQq2rquqa1tq6fs/B3GT/4kCyf3Gg2cc4kOxfHEizcf9y6iMAAEDHCDUAAICOEWpP9+F+D8CcZv/iQLJ/caDZxziQ7F8cSLNu/3KNGgAAQMc4ogYAANAxAxtqVXVeVd1aVeur6h1TPL+wqj45/vxXq2rNzE/JbNXD/vWLVXVTVV1fVX9fVcf1Y05mp33tXxPWva6qWlXNqrtc0V+97F9V9ePjP8NurKo/m+kZmd16+DNydVV9vqquG/9z8rX9mJPZp6o+WlUPVtUNe3i+quq3x/e966vqrJmecX8MZKhV1VCSDyU5P8naJBdV1dpJyy5Osqm1dmKSDyT5zZmdktmqx/3ruiTrWmunJ/mLJO+f2SmZrXrcv1JVS5O8LclXZ3ZCZrNe9q+qOinJLyd5WWvt1CQ/P+ODMmv1+DPsXUkua62dmeT1SX53ZqdkFvtYkvP28vz5SU4a/3VJkt+bgZmesYEMtSRnJ1nfWru9tbYjyaVJLpy05sIkfzL+9V8keVVV1QzOyOy1z/2rtfb51trW8YdfSbJqhmdk9url51eSvC9j/wCwbSaHY9brZf96c5IPtdY2JUlr7cEZnpHZrZd9rCVZNv71IUnuncH5mMVaa19M8p29LLkwyZ+2MV9JcmhVHTUz0+2/QQ21Y5LcPeHxhvFtU65prY0k2Zzk8BmZjtmul/1roouTfPaATsRcss/9q6rOTHJsa+0zMzkYc0IvP79OTnJyVX25qr5SVXv712uYrJd97FeTvKGqNiS5Isl/nJnRGAD7+3e0vprf7wH6ZKojY5Nvf9nLGphKz/tOVb0hybok339AJ2Iu2ev+VVXzMna69htnaiDmlF5+fs3P2GlDr8jY2QBXVdVprbVHDvBszA297GMXJflYa+1/VtVLk3x8fB8bPfDjMcfNqr/fD+oRtQ1Jjp3weFWeflh995qqmp+xQ+97O5QKT+pl/0pVnZvknUkuaK1tn6HZmP32tX8tTXJaki9U1Z1JzklyuRuK0KNe/3z8q9baztbaHUluzVi4QS962ccuTnJZkrTW/jnJQUlWzMh0zHU9/R2tKwY11K5OclJVHV9Vwxm7UPXySWsuT/Iz41+/Lsk/NB86R2/2uX+Nn5r2BxmLNNd3sD/2un+11ja31la01ta01tZk7BrIC1pr1/RnXGaZXv58/HSSH0iSqlqRsVMhb5/RKZnNetnHvp3kVUlSVadkLNQ2zuiUzFWXJ/np8bs/npNkc2vtvn4PtScDeepja22kqt6a5MokQ0k+2lq7sarem+Sa1trlST6SsUPt6zN2JO31/ZuY2aTH/et/JDk4yZ+P36Pm2621C/o2NLNGj/sXPCM97l9XJnlNVd2UZFeSX2qtPdy/qZlNetzH3p7kD6vqFzJ2Wtob/WM5vaiqT2TstOwV49c4vifJgiRprf1+xq55fG2S9Um2Jvl3/Zm0N2W/BwAA6JZBPfURAACgs4QaAABAxwg1AACAjhFqAAAAHSPUAAAAOkaoAQAAdIxQAwAA6BihBgAA0DH/PzUy6bCkGzYjAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_0=list(predicted_lr_pb[:,1])\n",
    "fpr,tpr,thresholds=roc_curve(y_test,y_0)  #计算fpr,tpr,thresholds\n",
    "auc=roc_auc_score(y_test,y_0) #计算auc\n",
    "\n",
    "#画曲线图\n",
    "plt.figure()\n",
    "plt.plot(fpr,tpr)\n",
    "plt.title('$ROC curve$')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Predict if a movie is a Action movie or a Musical movie:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Type</th>\n",
       "      <th>Script</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>gsavjgecaladdin complete script compiled ben s...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>ntense barbaric heyit home wind back sun west ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>an moons fool could fall fall hard dunes ah sa...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>y enchantment finest merchandise side river jo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>s tupperware never seen one intact famous dead...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3405</td>\n",
       "      <td>1</td>\n",
       "      <td>uch wipes face dad would proud bawls ca keep s...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3406</td>\n",
       "      <td>1</td>\n",
       "      <td>ght trained failure option detail small best r...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3407</td>\n",
       "      <td>1</td>\n",
       "      <td>uns sing solve problem like maria von trapp ki...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3408</td>\n",
       "      <td>1</td>\n",
       "      <td>y duck new family adorable ducklings ever saw ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3409</td>\n",
       "      <td>1</td>\n",
       "      <td>s armed forces huuah audience stares sits back...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>4588 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Type                                             Script\n",
       "0       0  gsavjgecaladdin complete script compiled ben s...\n",
       "1       0  ntense barbaric heyit home wind back sun west ...\n",
       "2       0  an moons fool could fall fall hard dunes ah sa...\n",
       "3       0  y enchantment finest merchandise side river jo...\n",
       "4       0  s tupperware never seen one intact famous dead...\n",
       "...   ...                                                ...\n",
       "3405    1  uch wipes face dad would proud bawls ca keep s...\n",
       "3406    1  ght trained failure option detail small best r...\n",
       "3407    1  uns sing solve problem like maria von trapp ki...\n",
       "3408    1  y duck new family adorable ducklings ever saw ...\n",
       "3409    1  s armed forces huuah audience stares sits back...\n",
       "\n",
       "[4588 rows x 2 columns]"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mus_df=all_df[all_df['Genres'].str.contains('musical')]\n",
    "act_df=all_df[all_df['Genres'].str.contains('action')]\n",
    "\n",
    "mus_sen = mus_df['Script'].tolist()\n",
    "act_sen = act_df['Script'].tolist()\n",
    "\n",
    "mus_split_data = sp_text(mus_sen)\n",
    "act_split_data = sp_text(act_sen)\n",
    "\n",
    "mus_sp_df = pd.DataFrame(mus_split_data)\n",
    "mus_sp_df.columns = ['Script']\n",
    "mus_sp_df.insert(0,'Type',\"0\")\n",
    "act_sp_df = pd.DataFrame(act_split_data)\n",
    "act_sp_df.columns = ['Script']\n",
    "act_sp_df.insert(0,'Type',\"1\")\n",
    "all1_sp_df = mus_sp_df.append(act_sp_df)\n",
    "all1_sp_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "all1_sp_df.to_csv('a_m_sp.csv') #action musical split sentences dataframe "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Precision: 0.9053995855848536, Recall: 0.8180635865788585, F1_score: 0.8502393871688478\n",
      "Precision: 0.9062820430658806, Recall: 0.8262466550020584, F1_score: 0.8567507181615065\n",
      "Precision: 0.9135802469135803, Recall: 0.8438188213856894, F1_score: 0.870661861074705\n",
      "Precision: 0.9216480252471342, Recall: 0.8328296149760047, F1_score: 0.8649847940983839\n",
      "Precision: 0.9071182174156256, Recall: 0.8251489203276248, F1_score: 0.8555015072622636\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import KFold\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn.metrics import precision_recall_fscore_support\n",
    "\n",
    "\n",
    "dataset = pd.read_csv('a_m_sp.csv')\n",
    "X = dataset.Script\n",
    "y = dataset.Type\n",
    "\n",
    "#cross-validation 打乱顺序，10等分\n",
    "kf = KFold(n_splits=5, shuffle=True)\n",
    "for train_index, test_index in kf.split(X):\n",
    "    X_train1, X_test1, y_train1, y_test1 = X[train_index], X[test_index], y[train_index], y[test_index]\n",
    "    cv = CountVectorizer(ngram_range = (2,3),stop_words='english')\n",
    "    X_train_counts = cv.fit_transform(X_train1)\n",
    "    X_test_counts = cv.transform(X_test1)\n",
    "    clf = MultinomialNB()\n",
    "    clf.fit(X_train_counts, y_train1)\n",
    "    predicted_clf = clf.predict(X_test_counts)\n",
    "    p, r, f1, _ = precision_recall_fscore_support(y_test1, predicted_clf, average='macro')\n",
    "    print(\"Precision: {0}, Recall: {1}, F1_score: {2}\".format(p, r, f1))\n",
    "    \n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "## 有问题！！\n",
    "fpr,tpr,thresholds=roc_curve(list(y_test1),predicted_clf)  #计算fpr,tpr,thresholds\n",
    "auc4=roc_auc_score(list(y_test1),predicted_clf) #计算auc\n",
    "\n",
    "#画曲线图\n",
    "plt.figure()\n",
    "plt.plot(fpr,tpr)\n",
    "plt.title('$ROC curve$')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.9650846468879472"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "auc4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Not useful:\n",
    "# kmeans:\n",
    "from sklearn.cluster import KMeans\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.feature_extraction.text import CountVectorizer\n",
    "from sklearn import metrics\n",
    "from sklearn import preprocessing\n",
    "\n",
    "\n",
    "X1_train, X1_test, y1_train, y1_test = train_test_split(all1_sp_df.Script, all1_sp_df.Type, test_size=0.3)\n",
    "\n",
    "# 这里将特征空间压缩到200维，也就是只要200个词语作为特征词\n",
    "cv1 = CountVectorizer(max_features=200)\n",
    "X1_train_counts = cv1.fit_transform(X1_train)\n",
    "X1_test_counts = cv1.transform(X1_test)\n",
    "X1_train_counts\n",
    "\n",
    "\n",
    "km=KMeans(n_clusters=2).fit(X1_train_counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}