MLH_FinalProject (4).ipynb

{
 "cells": [
  {
   "cell_type": "code",  
   "execution_count": 24, 
   "metadata": {}, 
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream", 
     "text": [
      "Requirement already satisfied: nltk in /opt/conda/lib/python3.8/site-packages (3.6.2)\n",
      "Requirement already satisfied: joblib in /opt/conda/lib/python3.8/site-packages (from nltk) (0.17.0)\n",
      "Requirement already satisfied: tqdm in /opt/conda/lib/python3.8/site-packages (from nltk) (4.54.0)\n",
      "Requirement already satisfied: regex in /opt/conda/lib/python3.8/site-packages (from nltk) (2021.4.4)\n",
      "Requirement already satisfied: click in /opt/conda/lib/python3.8/site-packages (from nltk) (7.1.2)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package stopwords to /home/jovyan/nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "!pip install nltk\n",
    "import nltk\n",
    "nltk.download(\"stopwords\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import scipy as sp\n",
    "import scipy.stats as stats\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns; sns.set()\n",
    "import re\n",
    "from nltk.corpus import stopwords\n",
    "from sklearn.model_selection import GridSearchCV #for my gridsearch\n",
    "plt.rcParams[\"figure.figsize\"] = [12,8] #setting the size of graphs\n",
    "hamColor = \"blue\"\n",
    "spamColor = \"orange\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "******PROJECT DESCRIPTION ******SMS Spam Detection Machine Learning ******Classification ******Mitchell Henderson CSPB 3022 \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Description of data: The SMS Spam Collection is a public set of SMS labeled messages that have been collected for mobile phone spam research.\n",
    "\n",
    "Unicamp location: https://www.dt.fee.unicamp.br/~tiago/smsspamcollection/\n",
    "\n",
    "Kaggle location: https://www.kaggle.com/uciml/sms-spam-collection-dataset \n",
    "\n",
    "Source:\n",
    "Tiago A. Almeida (talmeida@ufscar.br)\n",
    "Department of Computer Science\n",
    "Federal University of Sao Carlos (UFSCar)\n",
    "Sorocaba, Sao Paulo - Brazil\n",
    "\n",
    "Jose Maria Gomez Hidalgo (jmgomezh@yahoo.es)\n",
    "R&D Department Optenet\n",
    "Las Rozas, Madrid - Spain\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hello my name is Mitch Henderson. This is my final project for CSPB 3022, Intro to Data Science Algorithms.  Here in my github repo, you will find both a Jupyter notebook and a CSV file. When you run my Jupyter notebook,  you will see that I import a CSV of messages that are either spam or ham. I use machine learning to perform a classification.\n",
    "\n",
    "With the preponderance and growing issue of spam text messages I thought this problem will be a good topic to explore. More and more organizations are requiring cell phone numbers to stay in contact up-to-date or even use them as primary communication method. The goal is to help consumers filter out legitimate messages from the unsolicited spam. \n",
    "\n",
    "Therefore, I developed a model that can take a message and predict whether it is spam of ham. I will be using a model called “Bag of Words” – effectively this turns any piece of text into a vector.\n",
    "\n",
    "The database that I am using comes from Kagle and the links to the original data set are all provided here with the original URL location and the two authors of the dataset. This dataset consists of roughly 5000 text messages, out of these 5000, roughly 750 of these messages are spam.\n",
    "\n",
    "To start, let me introduce the concept of “cleaning”.  I have all of these words in my text messages that will have a mix of upper and lower case, various punctuations, white space, and I need to make the messages consistent since many of these abnormalities might not hold any predictive value.  All machine learning problems require the first step to be to import the data and then to clean it.  \n",
    "\n",
    "First, here, with this function, I’m importing the data I am beginning my exploratory data analysis also known as EDA and Please note this particular message “I HAVE A DATE ON SUNDAY WITH WILL!!”  - this message is in all caps and I want to turn it into all lower case to make sure all messages are in a consistent format, this is setting us up to be able to use our bag of words model.\n"
   ]
  },
  {
   "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>v1</th>\n",
       "      <th>v2</th>\n",
       "      <th>Unnamed: 2</th>\n",
       "      <th>Unnamed: 3</th>\n",
       "      <th>Unnamed: 4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ham</td>\n",
       "      <td>Go until jurong point, crazy.. Available only ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ham</td>\n",
       "      <td>Ok lar... Joking wif u oni...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>spam</td>\n",
       "      <td>Free entry in 2 a wkly comp to win FA Cup fina...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ham</td>\n",
       "      <td>U dun say so early hor... U c already then say...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ham</td>\n",
       "      <td>Nah I don't think he goes to usf, he lives aro...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>spam</td>\n",
       "      <td>FreeMsg Hey there darling it's been 3 week's n...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>ham</td>\n",
       "      <td>Even my brother is not like to speak with me. ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>ham</td>\n",
       "      <td>As per your request 'Melle Melle (Oru Minnamin...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>spam</td>\n",
       "      <td>WINNER!! As a valued network customer you have...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>spam</td>\n",
       "      <td>Had your mobile 11 months or more? U R entitle...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>ham</td>\n",
       "      <td>I'm gonna be home soon and i don't want to tal...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>spam</td>\n",
       "      <td>SIX chances to win CASH! From 100 to 20,000 po...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>spam</td>\n",
       "      <td>URGENT! You have won a 1 week FREE membership ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>ham</td>\n",
       "      <td>I've been searching for the right words to tha...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>ham</td>\n",
       "      <td>I HAVE A DATE ON SUNDAY WITH WILL!!</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>spam</td>\n",
       "      <td>XXXMobileMovieClub: To use your credit, click ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>ham</td>\n",
       "      <td>Oh k...i'm watching here:)</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>ham</td>\n",
       "      <td>Eh u remember how 2 spell his name... Yes i di...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>ham</td>\n",
       "      <td>Fine if thatÂ’s the way u feel. ThatÂ’s the wa...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>spam</td>\n",
       "      <td>England v Macedonia - dont miss the goals/team...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      v1                                                 v2 Unnamed: 2  \\\n",
       "0    ham  Go until jurong point, crazy.. Available only ...        NaN   \n",
       "1    ham                      Ok lar... Joking wif u oni...        NaN   \n",
       "2   spam  Free entry in 2 a wkly comp to win FA Cup fina...        NaN   \n",
       "3    ham  U dun say so early hor... U c already then say...        NaN   \n",
       "4    ham  Nah I don't think he goes to usf, he lives aro...        NaN   \n",
       "5   spam  FreeMsg Hey there darling it's been 3 week's n...        NaN   \n",
       "6    ham  Even my brother is not like to speak with me. ...        NaN   \n",
       "7    ham  As per your request 'Melle Melle (Oru Minnamin...        NaN   \n",
       "8   spam  WINNER!! As a valued network customer you have...        NaN   \n",
       "9   spam  Had your mobile 11 months or more? U R entitle...        NaN   \n",
       "10   ham  I'm gonna be home soon and i don't want to tal...        NaN   \n",
       "11  spam  SIX chances to win CASH! From 100 to 20,000 po...        NaN   \n",
       "12  spam  URGENT! You have won a 1 week FREE membership ...        NaN   \n",
       "13   ham  I've been searching for the right words to tha...        NaN   \n",
       "14   ham                I HAVE A DATE ON SUNDAY WITH WILL!!        NaN   \n",
       "15  spam  XXXMobileMovieClub: To use your credit, click ...        NaN   \n",
       "16   ham                         Oh k...i'm watching here:)        NaN   \n",
       "17   ham  Eh u remember how 2 spell his name... Yes i di...        NaN   \n",
       "18   ham  Fine if thatÂ’s the way u feel. ThatÂ’s the wa...        NaN   \n",
       "19  spam  England v Macedonia - dont miss the goals/team...        NaN   \n",
       "\n",
       "   Unnamed: 3 Unnamed: 4  \n",
       "0         NaN        NaN  \n",
       "1         NaN        NaN  \n",
       "2         NaN        NaN  \n",
       "3         NaN        NaN  \n",
       "4         NaN        NaN  \n",
       "5         NaN        NaN  \n",
       "6         NaN        NaN  \n",
       "7         NaN        NaN  \n",
       "8         NaN        NaN  \n",
       "9         NaN        NaN  \n",
       "10        NaN        NaN  \n",
       "11        NaN        NaN  \n",
       "12        NaN        NaN  \n",
       "13        NaN        NaN  \n",
       "14        NaN        NaN  \n",
       "15        NaN        NaN  \n",
       "16        NaN        NaN  \n",
       "17        NaN        NaN  \n",
       "18        NaN        NaN  \n",
       "19        NaN        NaN  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Importing the data. Beginning the EDA. Using the describe function which gives us a summary of our data.\n",
    "#First 20 messages\n",
    "\n",
    "df = pd.read_csv(\"spam_XLS_utf8.csv\")\n",
    "df.head(20)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I’m using this describe function here which gives us a summary of the first 20 messages or so. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>v1</th>\n",
       "      <th>v2</th>\n",
       "      <th>Unnamed: 2</th>\n",
       "      <th>Unnamed: 3</th>\n",
       "      <th>Unnamed: 4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>5572</td>\n",
       "      <td>5572</td>\n",
       "      <td>50</td>\n",
       "      <td>12</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>2</td>\n",
       "      <td>5169</td>\n",
       "      <td>43</td>\n",
       "      <td>10</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>ham</td>\n",
       "      <td>Sorry, I'll call later</td>\n",
       "      <td>bt not his girlfrnd... G o o d n i g h t . . .@\"</td>\n",
       "      <td>GE</td>\n",
       "      <td>GNT:-)\"</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>4825</td>\n",
       "      <td>30</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          v1                      v2  \\\n",
       "count   5572                    5572   \n",
       "unique     2                    5169   \n",
       "top      ham  Sorry, I'll call later   \n",
       "freq    4825                      30   \n",
       "\n",
       "                                               Unnamed: 2 Unnamed: 3  \\\n",
       "count                                                  50         12   \n",
       "unique                                                 43         10   \n",
       "top      bt not his girlfrnd... G o o d n i g h t . . .@\"         GE   \n",
       "freq                                                    3          2   \n",
       "\n",
       "       Unnamed: 4  \n",
       "count           6  \n",
       "unique          5  \n",
       "top       GNT:-)\"  \n",
       "freq            2  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()\n",
    "#describing the data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here is the size of the data set and then here I am creating a function I go through and get rid of the special characters, also by the nature of my data set I don’t have any NaNs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "ham     4825\n",
       "spam     747\n",
       "Name: v1, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.v1.value_counts()\n",
    "#Size of data\n",
    "#Shows us the number of ham vs spam messages"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here I will take each message and split it by word, this is a process that I call “tokenizing” in my code."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Creating a helper function\n",
    "#Cleaning data\n",
    "\n",
    "def cleanToken(w):\n",
    "    w = w.lower()\n",
    "    w = re.sub(r\"\\W\", \"\", w) #removes One non-word character\n",
    "    w = re.sub(r\"\\s\", \"\", w) #removes One whitespace\n",
    "    return w"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here is our frequency dictionary – this is still part of my cleaning.  What we are first going to do to build out our bag of words data representation, the first time around we are going to go through and count the number of times each unique word appears across all messages, and we’ll refer to this as our frequency dictionary which stores our unique words and their frequency of appearances.\n",
    "\n",
    "Just about any message will contain words like “the”, “and”, “or”, we refer to these words as “stop words”  as they contain no predictive value. They aren’t very helpful in classifying to ham or spam and I use the python library NLTK to get rid of them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9420\n"
     ]
    }
   ],
   "source": [
    "#Creating our frequency dictionary\n",
    "#Cleaning data\n",
    "#Cleaning Stop words\n",
    "#tokenizing\n",
    "\n",
    "freq = {} \n",
    "stop_words = set(stopwords.words('english')) #getting rid of our \"stop words\" such as \"the\", \"or\", \"and\", etc... NLTK python library\n",
    "\n",
    "for index,row in df.iterrows():\n",
    "    \n",
    "    for w in row[\"v2\"].split(\" \"):\n",
    "        w = cleanToken(w)\n",
    "        \n",
    "        if w != \"\" and w not in stop_words: #getting rid of empty strings and stopwords before they get added\n",
    "            if w in freq: #if key in dictionary\n",
    "                freq[w] += 1 #when I've already seen a word and want to increment it in my dictionary\n",
    "            else:\n",
    "                freq[w] = 1 #each time I see a word for the first time\n",
    "                \n",
    "\n",
    "for k, v in freq.items(): #iterable object that goes through our dictionary\n",
    "    #print(k,v)\n",
    "    pass\n",
    "\n",
    "print(len(freq))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Creating our list of frequencies\n",
    "#visualization\n",
    "\n",
    "y = [] #creating our empty list\n",
    "top_words = [] #our top most occuring words\n",
    "for word in sorted(freq, key = freq.get, reverse = True):\n",
    "    y.append(freq[word])\n",
    "    top_words.append(word) #capturing the top words, non-inclusive of the stop words\n",
    "    #print(word) #uncomment to view\n",
    "#print(top_words)\n",
    "#print(y)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have our frequency dictionary and we know how often each words shows up, let’s take a look at our distribution.  Here is a visualization of our distribution of words and I index them by the nth most common word. As you can see it’s a pretty sharp drop from the first ten or so. We can see that the most occurring words show up a LOT more than the less common words. It would appear that there is an exponential decay in the number of times a word appears. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Our distribution of words\n",
    "#We are exploring data here, part of the EDA\n",
    "#visualization plot\n",
    "\n",
    "plt.plot(y[0:75]) #we can change this value to see the distribution of word frequencies\n",
    "plt.xlabel(\"Index of the nth most common word\")\n",
    "plt.ylabel(\"Number of occurrences\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The first step was to count the number of times each word appears. The second step is to now take each individual message and convert them into a vector.  This vectorized message holds a count for the top occurring words within that message.  As an example, if I had the phrase “Mary had a little lamb\", imagine if after we did our frequency count, we found the top occurring words were maybe [“Mary”, \"little\", \"lamb\", \"the\", \"and\", \"a\"], the first thing in this vector would be to count the number of times “Mary” shows up, and the how many times the word “little“ shows up and so on. The python implementation of this is in my helper function where I vectorize messages, tokenize a message with top words, and return a dictionary."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "#helper function - vectorize\n",
    "#vectorize some message (take in a message, top words, and return a dictionary)\n",
    "\n",
    "def vectorize(message, top_words):\n",
    "    vector = dict.fromkeys(top_words, 0)\n",
    "    for w in message.split(\" \"):\n",
    "        w = cleanToken(w) #reassigning w to be the \"cleaned version\" of itself. \n",
    "        if w in vector:\n",
    "            vector[w] += 1 #grabbing the key value pair for this specific word and incrementing its value by 1\n",
    "    return vector\n",
    "    #print(vector) #test message\n",
    "#vectorize(\"Mary had a little lamb\", [\"little\", \"lamb\", \"the\", \"and\", \"a\"]) #small test case #uncomment to view\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I hope to end up with a high number of words that have high frequency that have strong predictive value of what words are indicative of spam and which ones are indicative of ham.\n",
    "\n",
    "I have a for loop that that goes through the messages and applies the vectorization, building out the bag of words dataset. \n",
    "\n",
    "In addition to the bag of words, we also feature engineer the length and the word count of each message."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>u</th>\n",
       "      <th>call</th>\n",
       "      <th>2</th>\n",
       "      <th>im</th>\n",
       "      <th>get</th>\n",
       "      <th>ur</th>\n",
       "      <th>4</th>\n",
       "      <th>dont</th>\n",
       "      <th>go</th>\n",
       "      <th>ok</th>\n",
       "      <th>...</th>\n",
       "      <th>never</th>\n",
       "      <th>special</th>\n",
       "      <th>meeting</th>\n",
       "      <th>job</th>\n",
       "      <th>draw</th>\n",
       "      <th>18</th>\n",
       "      <th>5</th>\n",
       "      <th>classification</th>\n",
       "      <th>length</th>\n",
       "      <th>wordCount</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>ham</td>\n",
       "      <td>111</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>ham</td>\n",
       "      <td>29</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>spam</td>\n",
       "      <td>155</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>ham</td>\n",
       "      <td>49</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>ham</td>\n",
       "      <td>61</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 203 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   u  call  2  im  get  ur  4  dont  go  ok  ...  never  special  meeting  \\\n",
       "0  0     0  0   0    0   0  0     0   1   0  ...      0        0        0   \n",
       "1  1     0  0   0    0   0  0     0   0   1  ...      0        0        0   \n",
       "2  0     0  1   0    0   0  0     0   0   0  ...      0        0        0   \n",
       "3  2     0  0   0    0   0  0     0   0   0  ...      0        0        0   \n",
       "4  0     0  0   0    0   0  0     1   0   0  ...      0        0        0   \n",
       "\n",
       "   job  draw  18  5  classification  length  wordCount  \n",
       "0    0     0   0  0             ham     111         20  \n",
       "1    0     0   0  0             ham      29          6  \n",
       "2    0     0   0  0            spam     155         28  \n",
       "3    0     0   0  0             ham      49         11  \n",
       "4    0     0   0  0             ham      61         13  \n",
       "\n",
       "[5 rows x 203 columns]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Bag of words\n",
    "#Iterate through original dataset, clean the words the exact same way we did with our freq dictionary\n",
    "#create our vectors\n",
    "#We should end up with new pandas dataframe with each frequently occuring word \n",
    "#together with a classification if the word is spam or ham\n",
    "\n",
    "vectors = [] #holding all of our vectorized rows\n",
    "for index,row in df.iterrows():\n",
    "    v = vectorize(row[\"v2\"],top_words[:200]) #grabbing the top words and vectorizing them. Set at 100 originally\n",
    "    v[\"classification\"] = row[\"v1\"]  #adding a key whether its ham or spam\n",
    "    v[\"length\"] = len(row[\"v2\"]) #looking at the length of each text message. This will give us a character count.\n",
    "    v[\"wordCount\"] = len(row[\"v2\"].split(\" \")) #adding the length of the text message in terms of words.\n",
    "    vectors.append(v) #appending onto our list. Now we can create our pandas dataframe\n",
    "\n",
    "BoW_dataframe = pd.DataFrame(vectors) #passing in vectors to this dataframe function, \n",
    "#converting our list of dictionaries to a panda dataframe\n",
    "BoW_dataframe.head() #printing our vectorized dataset. This is now \"machine model learning\" ready\n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we have our bag of words model, it would be nice to see if there are any noticeable trends. Here I’ve created a stacked bar chart of the top 10 words, as you can see these words like you, call, 2 and I’m are represented in the bar charts and show us the number of times they appear across all messages. \n",
    "\n",
    "We are getting interesting results, the word “u” appears more often in ham, whereas “call” tends to appear more often in spam. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Creating a stacked bar chart of top 10 words\n",
    "#visualization\n",
    "\n",
    "hamFrame = BoW_dataframe[BoW_dataframe.classification == \"ham\"] #this grabs only the entries in dataframe where its labeled ham\n",
    "hamCount = hamFrame[top_words[:10]].sum().values #list of top ten strings which are also in the hamFrame\n",
    "\n",
    "spamFrame = BoW_dataframe[BoW_dataframe.classification == \"spam\"] #this grabs only the entries in dataframe where its labeled spam\n",
    "spamCount = spamFrame[top_words[:10]].sum().values #list of top ten strings which are also in the spamFrame\n",
    "\n",
    "plt.bar(top_words[:10],hamCount, label= \"Ham\", color = hamColor) #total number across all ham messages\n",
    "plt.bar(top_words[:10],spamCount, bottom = hamCount, label=\"Spam\", color = spamColor) #total number across all spam messages\n",
    "plt.legend()\n",
    "\n",
    "plt.xlabel(\"Word\")\n",
    "plt.ylabel(\"Number of occurrences across all messages\")\n",
    "plt.show()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In our next bar chart, we show a stacked bar chart of average word count for both spam and ham messages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Further EDA\n",
    "#visualization\n",
    "#Why is this performed? Creating a stacked bar chart of average word count for both spam and ham messages\n",
    "\n",
    "averageHamLength = hamFrame.length.mean() #getting the average character count for ham messages\n",
    "#print(averageHamLength)\n",
    "\n",
    "averageSpamLength = spamFrame.length.mean() #getting the average character count for spam messages\n",
    "#print(averageSpamLength)\n",
    "\n",
    "plt.bar([\"Ham\", \"Spam\"], [averageHamLength, averageSpamLength], color = [hamColor, spamColor])\n",
    "plt.xlabel(\"Type of Message\")\n",
    "plt.ylabel(\"Average Character Count\")\n",
    "plt.show()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Next, we check messages for co-occurrence of words, here I am taking these and I look at probability of each word both in the ham and the spam data set, then I check for  spam over ham plus a very small value of epsilon, which I set to .0001 here. It is important for us to look at the words that have a high probability of appear in spam and not ham, and vice versa."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'observed probability')"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#EDA and visualization and analysis\n",
    "#Why is this performed? Exploring feature importance\n",
    "#Plotting nth word occurrence probability across ham and spam messages to check for co-occurrence of words\n",
    "\n",
    "conciseDataHamFrame = hamFrame.drop(columns = [\"classification\", \"length\", \"wordCount\"])\n",
    "conciseDataSpamFrame = spamFrame.drop(columns = [\"classification\", \"length\", \"wordCount\"])\n",
    "\n",
    "wordFreqHam = conciseDataHamFrame.astype(bool).sum(axis=0) / len(hamFrame)\n",
    "x = list(range(len(wordFreqHam)))\n",
    "\n",
    "wordFreqSpam = conciseDataSpamFrame.astype(bool).sum(axis=0) / len(spamFrame)\n",
    "y = list(range(len(wordFreqSpam)))\n",
    "\n",
    "fig,ax = plt.subplots(2,1)\n",
    "\n",
    "ax[0].bar(x,wordFreqHam.values)\n",
    "ax[0].set_title(\"Ham\")\n",
    "#ax[0].set_xlabel(\"index of the nth word\")\n",
    "ax[0].set_ylabel(\"observed probability\")\n",
    "\n",
    "ax[1].bar(y,wordFreqSpam.values)\n",
    "ax[1].set_title(\"Spam\")\n",
    "ax[1].set_xlabel(\"index of the nth word\")\n",
    "ax[1].set_ylabel(\"observed probability\")\n",
    "\n",
    "\n",
    "#print(x)\n",
    "#plt.bar(x, wordFreqHam.values)\n",
    "#print(wordFreqHam.index.values)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I am looking to see words that appear much more in spam and this here gives me a very strong list of words that appear overwhelmingly within spam such as “claim”, “prize”, “guaranteed”, and “18”.\n",
    "\n",
    "This is giving me very good data to determine where these words are appearing and whether or not they might be able to predict whether or not a message is indeed spam.  At this stage of the EDA, this is a hint at which model will work best, for example a decision tree should be able to find these words and classify each message, it’s a rule-based system that makes splits based on a set of heuristics of yes/no questions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "claim         1445.783133\n",
       "prize         1124.497992\n",
       "guaranteed     669.344043\n",
       "18             575.635877\n",
       "tone           548.862115\n",
       "16             213.490274\n",
       "urgent         161.316633\n",
       "nokia           90.882794\n",
       "service         77.812645\n",
       "txt             73.479042\n",
       "win             51.794213\n",
       "mobile          45.473892\n",
       "draw            44.769441\n",
       "customer        33.377203\n",
       "cash            31.564941\n",
       "contact         28.977650\n",
       "per             23.285612\n",
       "free            18.334070\n",
       "mins            17.593573\n",
       "stop            16.642554\n",
       "dtype: float64"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#EDA and visualization\n",
    "#Exploring feature importance analysis\n",
    "#Why is this performed? Looking at ratio of above two graphs\n",
    "#for each index I am plotting: spam / (ham + epsilon)\n",
    "#This is giving is words that appear much much more in spam text messages, than in ham text messages.\n",
    "#We find that some words have an extraordinarily higher ratio (such as \"claim\" and \"prize\") of appearing in \n",
    "#...Spam than in Ham text messages.\n",
    "\n",
    "epsilon = .0001\n",
    "ratio = wordFreqSpam / (wordFreqHam + epsilon)\n",
    "\n",
    "plt.bar(y,ratio.values)\n",
    "plt.title(\"Ratio\")\n",
    "plt.xlabel(\"Index of the nth word\")\n",
    "plt.ylabel(\"Ratio of Spam to Ham probability\")\n",
    "plt.show()\n",
    "#print(ratio)\n",
    "#print(ratio.columns)\n",
    "\n",
    "ratio.sort_values(ascending = False)[:20]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Project topic: (5 pts)\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split, cross_val_score\n",
    "from sklearn.utils import resample #Bringing in resample for my upsamples"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I create my test and train split of my data. This will be used to train my models and then test them accordingly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3733, 202)\n",
      "(1839, 202)\n"
     ]
    }
   ],
   "source": [
    "#Test / Train split of my data\n",
    "X = BoW_dataframe.drop(columns = \"classification\")\n",
    "y = BoW_dataframe.classification\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X,y,test_size=.33,random_state=42) #makes sure that we shuffle the data in the same way.\n",
    "print(X_train.shape)\n",
    "print(X_test.shape)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "When I was reviewing the data initially there were far more ham messages than spam messages. Because of this we will want to make sure we have a balanced dataset as that will work better for some of the models. Balance is a problem within machine learning - where there is a one example in the minority class for hundreds or thousands of cases in the majority class."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Upsampling\n",
    "df_Spam_Upsampled = resample(spamFrame, \n",
    "                             replace=True, #sampling with replacements\n",
    "                            n_samples= len(hamFrame), #getting my hamframe length to equal n_samples\n",
    "                            random_state=123) #setting my random number generator\n",
    "\n",
    "df_Upsampled = pd.concat([hamFrame, df_Spam_Upsampled]) #combining majority ham frame with upsampled spam minority class\n",
    "\n",
    "X_Upsample = df_Upsampled.drop(columns = \"classification\")\n",
    "y_Upsample = df_Upsampled.classification\n",
    "\n",
    "\n",
    "X_train_Upsample, X_test_Upsample, y_train_Upsample, y_test_Upsample = train_test_split(X_Upsample,y_Upsample,test_size=.33,random_state=42)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here I do my imports for my different machine learning models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import neighbors #For KNN\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.svm import SVC\n",
    "from sklearn.discriminant_analysis import LinearDiscriminantAnalysis\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Then I create this function for the grid search and cross validation. For every model that I am considering, I want to do a grid search for hyperparameter tuning and then cross validation. Since I will be doing this for every model I decided to write a function for "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "#this function is responsible for gridsearch, crossvalidation\n",
    "\n",
    "def trainModel(CLF,parameters,X,y,numFolds=10):\n",
    "\n",
    "    gridsearch = GridSearchCV(CLF, parameters, cv=numFolds) #can try different cv values here as well\n",
    "    gridsearch.fit(X,y)\n",
    "    print(\"******************\" + type(CLF).__name__ + \"******************\")\n",
    "    print(\"Here are our optimal parameters \",gridsearch.best_params_)\n",
    "    print(\"Here is our score:\",gridsearch.best_score_) #prints out parameters that did best and their respective scores.\n",
    "\n",
    "    bestCLF = gridsearch.best_estimator_ #returning our best classifier\n",
    "    results = cross_val_score(bestCLF, X, y, cv = numFolds) #returns the cross validation scores\n",
    "    print(f\"Here is the mean from the cross validation score from the \"+ type(CLF).__name__ + \" classifier: {results.mean()}\") #gives us the mean across all scores\n",
    "    print(f\"Here is the standard deviation from the cross validation results: {results.std()}\") #returning the standard dev\n",
    "    print(f\"Here are the actual results: {results}\") #numpy array of results\n",
    "    print()\n",
    "    print()\n",
    "    print()\n",
    "    return bestCLF\n",
    "    #Look up sklearn models for other classifiers. Code will look very similar\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here are my models and I will give a brief description of each:\n",
    "\n",
    "First is KNN, or K nearest neighbor, we use KNN to predict new values to make class predictions.  We are looking at geometric distance and finding the nth nearest neighbor resembling any new sample. Whatever class the majority of class the neighbors belong to then that will be the predicted class for the new sample.\n",
    "\n",
    "Logistic regression, this is essentially the classification version of linear regression.\n",
    "\n",
    "Then we have decision tree predictive modeling approach where the observations are the branches and the conclusions about our target value are in the leaves. This decision tree model is a rule-based approach where the model is responsible for finding splits in the data and continuously making them until there is a high accuracy in the prediction.\n",
    "\n",
    "Then we have random forest. Random forest is a collection of decision trees, each is slightly different than the other, where we might likely get better results here as opposed to just a single decision tree. \n",
    "\n",
    "Support vector machines are when we draw a line for classification purposes, the samples on one side are classified as spam and the other side is ham.\n",
    "\n",
    "And then we have linear discriminate analysis. It is a linear classifier that takes into account empirical probabilities from the dataset.  Because LDA takes into account empirical probabilities, its decision making process depends heavily on how well balanced the data is.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "#List of classifiers\n",
    "classifiers = []\n",
    "\n",
    "#knn\n",
    "knnCLF = neighbors.KNeighborsClassifier()\n",
    "knnParams = {'n_neighbors':[5,25,57], 'p':[1,2]} #parameters must be appropriate to model being evaluated\n",
    "knnDictionary = {\"classifier\": knnCLF, \"params\": knnParams} #dictionary of key value pairs and parameters\n",
    "#print(knnDictionary[\"params\"])\n",
    "classifiers.append(knnDictionary) #data structure to hold classifier and parameters\n",
    "#print(classifiers)\n",
    "\n",
    "#logistic regression\n",
    "lrCLF = LogisticRegression(max_iter=1000, random_state=0) #creating logistic regression model\n",
    "lrParamGrid = {\"penalty\": [\"l2\"], \"C\": [1.0,0.9,0.7]}  #hyperparameters\n",
    "lrDictionary = {\"classifier\": lrCLF, \"params\": lrParamGrid}\n",
    "classifiers.append(lrDictionary)\n",
    "\n",
    "#Decision Tree\n",
    "dtcCLF = DecisionTreeClassifier(random_state=0)\n",
    "dtcParamGrid = {\"max_depth\": [5,20,250]} #hyperparameters\n",
    "dtcDictionary = {\"classifier\": dtcCLF, \"params\": dtcParamGrid}\n",
    "classifiers.append(dtcDictionary)\n",
    "\n",
    "#Random Forest\n",
    "rfcCLF = RandomForestClassifier(max_depth=60, random_state=0) \n",
    "#Changing the max_depth hyperparameter will change results\n",
    "rfcParamGrid = {}\n",
    "rfcDictionary = {\"classifier\": rfcCLF, \"params\": rfcParamGrid}\n",
    "classifiers.append(rfcDictionary)\n",
    "\n",
    "#Support Vector Machine - Upsampled\n",
    "svmCLF = SVC() #baseline\n",
    "svmParamGrid = {}\n",
    "svmDictionary = {\"classifier\": svmCLF, \"params\": svmParamGrid}\n",
    "classifiers.append(svmDictionary)\n",
    "\n",
    "#LinearDiscriminantAnalysis\n",
    "ldaCLF = LinearDiscriminantAnalysis()\n",
    "ldaParamGrid = {}\n",
    "ldaDictionary = {\"classifier\": ldaCLF, \"params\": ldaParamGrid}\n",
    "classifiers.append(ldaDictionary)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And this is our for a loop that goes through each classifier and trains each one, we are given series of outputs.  We train our models on both data sets, one on non-up sampled and one on up sampled data - but then we will test inly on non up-sample data. \n",
    "\n",
    "\n",
    "We will be given optimal parameters. And then we are running a cross validation and given an accuracy score, mean, standard deviation.\n",
    "\n",
    "We are given the actual results from the cross validation.\n",
    "\n",
    "Random Forest has the highest accuracy score, again looking at cross validation. It outperformed all the others. The random forest classifier comes in with an accuracy score of .98 score.\n",
    "\n",
    "That’s not to say the others still come very close in fact they do- Decision Tree comes in at .974 and Logistic Regression at .97.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training on non-upsampled data: \n",
      "******************KNeighborsClassifier******************\n",
      "Here are our optimal parameters  {'n_neighbors': 25, 'p': 1}\n",
      "Here is our score: 0.936237473297874\n",
      "Here is the mean from the cross validation score from the KNeighborsClassifier classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.009977857183224994\n",
      "Here are the actual results: [0.93582888 0.95454545 0.94385027 0.93565684 0.9383378  0.93297587\n",
      " 0.93029491 0.93297587 0.91420912 0.94369973]\n",
      "\n",
      "\n",
      "\n",
      "Training on upsampled data:\n",
      "******************KNeighborsClassifier******************\n",
      "Here are our optimal parameters  {'n_neighbors': 5, 'p': 1}\n",
      "Here is our score: 0.9279168919662553\n",
      "Here is the mean from the cross validation score from the KNeighborsClassifier classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.00872379361875988\n",
      "Here are the actual results: [0.93353941 0.94126739 0.92890263 0.91962906 0.93353941 0.90866873\n",
      " 0.93498452 0.92879257 0.92260062 0.92724458]\n",
      "\n",
      "\n",
      "\n",
      "Training on non-upsampled data: \n",
      "******************LogisticRegression******************\n",
      "Here are our optimal parameters  {'C': 1.0, 'penalty': 'l2'}\n",
      "Here is our score: 0.9699925449097503\n",
      "Here is the mean from the cross validation score from the LogisticRegression classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.009201246290565897\n",
      "Here are the actual results: [0.97058824 0.97058824 0.98663102 0.97050938 0.97050938 0.96782842\n",
      " 0.98123324 0.95978552 0.97050938 0.95174263]\n",
      "\n",
      "\n",
      "\n",
      "Training on upsampled data:\n",
      "******************LogisticRegression******************\n",
      "Here are our optimal parameters  {'C': 1.0, 'penalty': 'l2'}\n",
      "Here is our score: 0.9563788095568496\n",
      "Here is the mean from the cross validation score from the LogisticRegression classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.006782550882876691\n",
      "Here are the actual results: [0.95981453 0.96136012 0.94899536 0.95672334 0.96599691 0.94117647\n",
      " 0.95820433 0.95975232 0.95975232 0.95201238]\n",
      "\n",
      "\n",
      "\n",
      "Training on non-upsampled data: \n",
      "******************DecisionTreeClassifier******************\n",
      "Here are our optimal parameters  {'max_depth': 20}\n",
      "Here is our score: 0.9531117833436079\n",
      "Here is the mean from the cross validation score from the DecisionTreeClassifier classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.009934230426656584\n",
      "Here are the actual results: [0.95989305 0.95989305 0.97326203 0.94906166 0.95442359 0.94906166\n",
      " 0.93565684 0.94101877 0.95442359 0.95442359]\n",
      "\n",
      "\n",
      "\n",
      "Training on upsampled data:\n",
      "******************DecisionTreeClassifier******************\n",
      "Here are our optimal parameters  {'max_depth': 250}\n",
      "Here is our score: 0.9754039840942477\n",
      "Here is the mean from the cross validation score from the DecisionTreeClassifier classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.006134471382862882\n",
      "Here are the actual results: [0.97836167 0.98454405 0.9675425  0.98299845 0.97681607 0.96904025\n",
      " 0.97058824 0.9752322  0.98142415 0.96749226]\n",
      "\n",
      "\n",
      "\n",
      "Training on non-upsampled data: \n",
      "******************RandomForestClassifier******************\n",
      "Here are our optimal parameters  {}\n",
      "Here is our score: 0.9681173029777351\n",
      "Here is the mean from the cross validation score from the RandomForestClassifier classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.00594611138450381\n",
      "Here are the actual results: [0.97058824 0.97058824 0.98128342 0.97319035 0.96246649 0.96782842\n",
      " 0.96514745 0.95978552 0.96782842 0.96246649]\n",
      "\n",
      "\n",
      "\n",
      "Training on upsampled data:\n",
      "******************RandomForestClassifier******************\n",
      "Here are our optimal parameters  {}\n",
      "Here is our score: 0.9853063197132753\n",
      "Here is the mean from the cross validation score from the RandomForestClassifier classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.004331123720749319\n",
      "Here are the actual results: [0.98299845 0.98763524 0.97681607 0.98763524 0.98608964 0.97832817\n",
      " 0.98916409 0.98606811 0.99071207 0.9876161 ]\n",
      "\n",
      "\n",
      "\n",
      "Training on non-upsampled data: \n",
      "******************SVC******************\n",
      "Here are our optimal parameters  {}\n",
      "Here is our score: 0.9201602844403665\n",
      "Here is the mean from the cross validation score from the SVC classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.015656196880628477\n",
      "Here are the actual results: [0.9197861  0.95721925 0.92513369 0.91957105 0.92225201 0.91689008\n",
      " 0.92225201 0.90348525 0.89276139 0.92225201]\n",
      "\n",
      "\n",
      "\n",
      "Training on upsampled data:\n",
      "******************SVC******************\n",
      "Here are our optimal parameters  {}\n",
      "Here is our score: 0.8657342055019356\n",
      "Here is the mean from the cross validation score from the SVC classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.014571399176159593\n",
      "Here are the actual results: [0.88871716 0.89644513 0.85935085 0.84698609 0.86553323 0.86068111\n",
      " 0.85294118 0.85758514 0.86532508 0.86377709]\n",
      "\n",
      "\n",
      "\n",
      "Training on non-upsampled data: \n",
      "******************LinearDiscriminantAnalysis******************\n",
      "Here are our optimal parameters  {}\n",
      "Here is our score: 0.9673108629266963\n",
      "Here is the mean from the cross validation score from the LinearDiscriminantAnalysis classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.01010277121088635\n",
      "Here are the actual results: [0.96791444 0.98128342 0.98128342 0.96782842 0.96246649 0.96246649\n",
      " 0.97587131 0.94906166 0.97050938 0.95442359]\n",
      "\n",
      "\n",
      "\n",
      "Training on upsampled data:\n",
      "******************LinearDiscriminantAnalysis******************\n",
      "Here are our optimal parameters  {}\n",
      "Here is our score: 0.9477196491547077\n",
      "Here is the mean from the cross validation score from the LinearDiscriminantAnalysis classifier: {results.mean()}\n",
      "Here is the standard deviation from the cross validation results: 0.0067655643691776485\n",
      "Here are the actual results: [0.94590417 0.94590417 0.94899536 0.94899536 0.94126739 0.93498452\n",
      " 0.9504644  0.9504644  0.9628483  0.94736842]\n",
      "\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#For loop that goes through each classifier and trains each one\n",
    "\n",
    "bestClassifiers = []\n",
    "bestClassifiersUpsampled = []\n",
    "for c in classifiers: #we can use indexing here to select out a specific model such as [classifiers[-1]]\n",
    "    print(\"Training on non-upsampled data: \")\n",
    "    bestModel = trainModel(c[\"classifier\"], c[\"params\"],X_train,y_train)\n",
    "    bestClassifiers.append(bestModel)\n",
    "    print(\"Training on upsampled data:\")\n",
    "    bestModelUpsampled = trainModel(c[\"classifier\"], c[\"params\"],X_train_Upsample,y_train_Upsample)\n",
    "    bestClassifiersUpsampled.append(bestModelUpsampled)\n",
    "    \n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now that we are ready to proceed with testing and found the best versions of each model, we can inspect their performance on unseen data and measure how well they do based on the confusion matrix and looking at their classification report. \n",
    "\n",
    "For logistic regression and LDA, because they have thresholds for their classification, we are also able to plot their ROC curve which shows us the relationship between precision and recall based on certain threshold values.\n",
    "\n",
    "In addition to our classification report for each model, we also have our confusion matrix. The confusion matrix is a useful visualization to see exact numbers of what got classified as what.\n",
    "\n",
    "For example, below is support vector machines - and we can see that based on the labels in 272 cases where it was classified as spam, but in reality it’s ham. 272 is kind of high number when you look at some of the other outputs. LDA for example is 47.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************KNeighborsClassifier******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       0.96      0.97      0.96      1587\n",
      "        spam       0.79      0.72      0.75       252\n",
      "\n",
      "    accuracy                           0.93      1839\n",
      "   macro avg       0.87      0.84      0.86      1839\n",
      "weighted avg       0.93      0.93      0.93      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************LogisticRegression******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       0.98      0.99      0.98      1587\n",
      "        spam       0.93      0.84      0.88       252\n",
      "\n",
      "    accuracy                           0.97      1839\n",
      "   macro avg       0.95      0.92      0.93      1839\n",
      "weighted avg       0.97      0.97      0.97      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "****************** LogisticRegression ROC Curve ******************\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************DecisionTreeClassifier******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       0.98      0.98      0.98      1587\n",
      "        spam       0.87      0.85      0.86       252\n",
      "\n",
      "    accuracy                           0.96      1839\n",
      "   macro avg       0.92      0.92      0.92      1839\n",
      "weighted avg       0.96      0.96      0.96      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlYAAAHlCAYAAAA3E7JZAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAAA0/0lEQVR4nO3de1xVdb7/8fdmIygSIiqwMVLTVLQcm7xkmZZmNB7U7ExDB0pPN6sRf9rNa6mZmtB9vEQzTXXGGOdE2gW0g2X3pjRnzKEY72ipW0jQUDSBvdfvD3IXKYj0ZW+Wvp7zWI/Y+7vWd382j5w+fr6f9V0Oy7IsAQAA4BcLCnQAAAAAZwoSKwAAAENIrAAAAAwhsQIAADCExAoAAMAQEisAAPCLWd6DgQ6hSXCw3YLkLblR8u4LdBjAGWlMvx6BDgE4o7VtH6WnP54b6DAkSd6S/zL739OgWAW1WWZuPj8IDnQATYJ3n+TZE+gogDNS0a52gQ4BgL/w31MSKwAAYIZXXkleozParWeJxAoAABjhsbySZTCxsry2S1TslggCAAA0WXZLBAEAQBNl/fA/UxwG5/IXKlYAAACGULECAABGeGXJMti8bseKFYkVAAAwwmNZMrk9psOGW22yFAgAAGAIFSsAAGBE9UKguSpTkA2XAqlYAQAAGELFCgAAGOExXLEyuXWDv1CxAgAAMISKFQAAMMJruGIlG1asSKwAAIARXsuSx+QWCWy3AAAAcPaiYgUAAIzw/nCY4jA4l79QsQIAADCEihUAADDCI0semtcBAAB+Oa8leQzmQg775VUsBQIAAJhCxQoAABhhunnd5Fz+QsUKAADAECpWAADACI8c8hjcJMFhww0XSKwAAIARXqv6MDmf3bAUCAAAYAgVKwAAYITX8FJgkA2XAqlYAQAAGELFCgAAGGG6ed2OFSsSKwAAYITXcshrmUuGTM7lLywFAgAAGELFCgAAGGG6ed1pw6VAKlYAAACGULECAABGeBQkj8Gajcm5/MV+EQMAADRRVKwAAIAR1Y+0MXlXoLGp/IbECgAAGGG6ed1L8zoAAMDZi4oVAAAwwqMgeSya1wEAAGwvPT1dQ4YMUbdu3bRly5YTxhctWnTCWGFhoZKTk5WYmKjk5GTt3LmzXmO1IbECAABGeBVk/DgdQ4cOVVZWltq3b3/C2FdffaUvvvhCcXFxNd6fNWuWUlJSlJeXp5SUFM2cObNeY7UhsQIAAEYcb143dZxu83qfPn3kcrlOeL+iokJz5szRrFmz5HD8OGdJSYkKCgqUlJQkSUpKSlJBQYFKS0vrHKsLPVYAAKBJc7vd8ng8Nd6LiIhQREREva5/5plnNHLkSMXHx58wb0xMjJxOpyTJ6XQqOjpabrdblmXVOhYVFVXrZ5FYAQAAIzyWw2zz+g97YqWmpmrPnj01xtLS0jRhwoRTzrFhwwbl5+fr/vvvNxZXXUisAABAk5aVlXXSilV9fP7559qxY4eGDh0qSdq3b59uu+02Pfroo0pISFBRUZE8Ho+cTqc8Ho+Ki4vlcrlkWVatY3UhsQIAAEZ4G9AXdar5JJ0ymanLuHHjNG7cON/rIUOGKDMzU127dpUkJSQkKDc3V6NGjVJubq4SEhJ8S311jdWGxAoAABjhNfwQ5tO9K3Du3LlavXq19u/fr1tuuUWRkZFauXJlndfMnj1bU6dO1ZIlSxQREaH09PR6jdXGYVmWDZ/EY5b32yslz55Tngfg9CXG9Q50CMAZLaZDO71cuCTQYUiS3vvmGh2t2mtsvhbBcboqfrWx+fyBihUAADDCYxneed3gXP5iv4gBAACaKCpWAADAiIbsln6q+ezGfhEDAAA0UVSsAACAEV7rx009Tc1nNyRWAADACI/h7RZMzuUv9osYAACgiaJiBQAAjPBaQfIa3CLB5Fz+Yr+IAQAAmigqVgAAwIhAP9KmKSCxAgAARngsh9G7Ak3O5S/2SwUBAACaKCpWAADACK8chndep2IFAABw1qJiBQAAjPBaQfKc5dstkFgBAAAjqpcCDT7ShqVAAACAsxcVKwAAYITH8FKgybn8xX4RAwAANFFUrAAAgBFeOQzvvE6PFQAAwFmLihUAADDCK4e8Bh9DY8eKFYkVAAAwwmP4Icwm5/IX+0UMAADQRFGxAgAARlhWkNHd0i22WwAAADh7UbECAABGeOSQx2DDucm5/IXECgAAGOG1HEaXAk3eYegvLAUCAAAYQsUKAAAY4TW8FGjHfayoWAEAABhCxQoAABjhNbzdgsm5/IXECgAAGOGxguQxmAyZnMtf7BcxAABAE0XFCgAAGGHJYbTh3KJ5HQAA4OxFxQoAABjhsRyGe6yoWAEAAJy1qFgBAAAjqh9pY3CDUBtWrEisAACAEV4FyWNwMcxrw4U1+0UMAADQRFGxAgAARrAUSMUKAADAGCpWAADACK+CjPZF2bHHisQKAAAY4ZHD6N5THnZeBwAAOHtRsQIAAEZYltmGc8syNpXfULECAAAwhIoVAAAwwmsFyWvwWYEm5/IX+0UMAACaJI8cxo/TkZ6eriFDhqhbt27asmWLJOnAgQO64447lJiYqBEjRigtLU2lpaW+awoLC5WcnKzExEQlJydr586d9RqrDYkVAAA4IwwdOlRZWVlq37697z2Hw6Hbb79deXl5ysnJUXx8vB5//HHf+KxZs5SSkqK8vDylpKRo5syZ9RqrDYkVAAAwwvph53VTh/VDI7zb7dbu3btrHGVlZSd8fp8+feRyuWq8FxkZqf79+/te9+7dW3v37pUklZSUqKCgQElJSZKkpKQkFRQUqLS0tM6xutBjBQAAmrTU1FTt2bOnxntpaWmaMGHCac3j9Xq1bNkyDRkyRFJ1whYTEyOn0ylJcjqdio6OltvtlmVZtY5FRUXV+hkkVgAAwIjGal7PysqSx+OpMRYREXHa8z3yyCMKCwvTTTfdZCS+kyGxQsC88UJbvf1KlHZuaq4rrzuo+5/+WpK075sQje3fQ83DfvxD9LvxxUq9p0iStPTxWC37Q4yahXh945lrNsvVoUKSNKZfDx3YH6ygoOoNUHr0Kdejf9vhr68F2MbkhbvUe+BhNQ/z6kBxsLKfjdb//bWNuv+6XGMn79MFFx2Vxyv96+/hevah9iotbhbokHGW+vnyXkOkp6dr165dyszMVFBQkG/eoqIieTweOZ1OeTweFRcXy+VyybKsWsfqQmKFgGkTW6mUiUVa/8E5qvj+xL/hrNiUL2ct/4YOHnlAUxZ9XevcD7+0Q78edNhUqMAZ6X8Xxuip++JVWRGk+C7fK+PV7dqW30LhrTxa9XIb/eP9c+TxODR+3m7d99Q3mpF6fqBDRhPnlUNeg4+hMTXXU089pS+//FJ//OMfFRIS4nu/TZs2SkhIUG5urkaNGqXc3FwlJCT4lvrqGqtNQBOrbt266Z///KdatmwZyDAQIAOHfydJ2vKvFtrvDjnF2QBM27Wlue9ny6o+4jpW6MOcyBrnvfliWz22fLufo4MdeSyZfVbgae68PnfuXK1evVr79+/XLbfcosjISD399NPKzMxUx44ddeONN0qSzj33XC1evFiSNHv2bE2dOlVLlixRRESE0tPTffPVNVYbKlZosm7u10MOh3TxFYd0x0N71arNj0uDn73dSv/Z40JFRVdq5C37NWJsSY1r09M6yPI61PnCI7r9ob3q3PN7f4cP2ELa/N0allyq5i0sbc1voXVrzjnhnIv6l2vX5uYnuRpoWh588EE9+OCDJ7y/efPmWq/p3LmzsrOzT3usNgFPrJYuXaq3335bBw8e1OTJk5WYmChJuu+++1RYWKjKykqdd955mj9/vlq1aqW1a9dq3rx56tWrlzZu3Kjg4GBlZGRo0aJF2rp1q1wulxYuXKiwsLAAfzM0VKuoKi18a7M69zyqsgPBWjT9XKWnddD8ZdV9UoNGHtDwm/Yrsl2VNv0zTHPv6KTwCI+uGn1QkjRl0S51ueiILEt6/c/tNCOls57/cJPCW3nq+FTg7LRo+rla8mB7JfQpV68B5aqsqLks3ynhqFLvKdLsWzoGJkDYimW4ed1i5/XTFx4eruXLlysjI0Nz5871vT9jxgytWLFCOTk56tKli/70pz/5xrZv367U1FTl5OSod+/euu222zRt2jStWrVKQUFBWrlyZSC+Cgxp0dKrrr86Kmew1LpdlcbP261/fBCh8kPV/7p26HpMbWKr5HRKPfse0XW3fauPVkb6ru/Zr1yhLSw1D7N044RitYzw6Mu1LDcDtfF6HfpqXbjauSqVNHa/7/24jsc09+UdenZmnL5cFx7ACAH7CHjFavjw4ZKqN+wqLi7WsWPHFBoaqjfeeEM5OTmqrKzUkSNH1LFjR981nTp1UkJCgiSpR48e2rt3r2JjYyVJPXv21K5du/z+PdB4HMeX62tba3fUMfbD9XZ8Qjrgb0HBlu/u2uj2FXr0f7frr0/HaM3yupt1geOOb+xpcj67CXjFKjQ0VJJ8G3BVVVVp/fr1WrZsmZ5//nnl5ORo0qRJqqio8F3z045+p9Ppm+P465/vdYGmyVMlVXzvkNfjkNdT/bOnStr0zzB9sy1UXq9UVurUkgfbq9dlh9Qyonp7hb//X4QOHXTKsqRNG8L0xgttNSCxuhG+eHczfbWupSorHKr43qHsJe1UVhqsnn3LA/lVgSanVZtKDR51QM3DPAoKsnTJ4DJddd1BbfwkXG1iK5WevV05L7XVyqVtAx0qbOT4XYEmD7sJeMXqZMrKyhQeHq7IyEhVVFRo+fLlgQ4JjeCvT8fq5Sdjfa/XLI/STffu07mdv9eLC1w6uD9YLc/x6uJBhzRtyY9VyA/eaK0n7ztPlcccauuq1O9+X6xhvzsgSTpS7tTCaedq784QhTS31LnnUc3N2q6IKJJtoAbLoaQxJfp/C3bLESQV7w5R5sw4fZrXSqn37lNcxwrddG+Rbrq3yHfJdRdcFMCAAXtokonVoEGD9Oabb+o3v/mNYmJidOGFFyo/Pz/QYcGwm+/fp5vv33fSseON6Ccz7dnal3o7dvtemWtqv/sDQLXvSoP1wH92OelY1pOxyvrJX3qA+vLK8FKgDStWDsui+8T77ZWSZ88pzwNw+hLjegc6BOCMFtOhnV4uXBLoMCRJEzZM0/5jJac+sZ7ahrbRwosfNTafPzTJihUAALAftlsgsQIAAIZ4LbN38nltuKZmv1QQAACgiaJiBQAAjGiqD2H2JypWAAAAhlCxAgAARrDzOhUrAAAAY6hYAQAAIyzDFSvLhhUrEisAAGAES4EsBQIAABhDxQoAABhBxYqKFQAAgDFUrAAAgBFsEEpiBQAADOGuQJYCAQAAjKFiBQAAjPDKbMO519hM/kPFCgAAwBAqVgAAwAi2WyCxAgAAhpBYsRQIAABgDBUrAABghuUwu0UCFSsAAICzFxUrAABgBDuvU7ECAAAwhooVAAAwgrsCSawAAIAhluHmdZ4VCAAAcBajYgUAAIxgKZCKFQAAgDFUrAAAgBmW4b4oy9xU/kJiBQAAjPDK8FIg+1gBAACcvahYAQAAIyyr+jA5n91QsQIAADCEihUAADCCZwWSWAEAAEPYeZ2lQAAAAGOoWAEAACMswzuvU7ECAAA4i1GxAgAARrDdAhUrAAAAY0isAACAEcfvCjR5nI709HQNGTJE3bp105YtW3zvFxYWKjk5WYmJiUpOTtbOnTt/8VhtSKwAAIARgU6shg4dqqysLLVv377G+7NmzVJKSory8vKUkpKimTNn/uKx2pBYAQCAJs3tdmv37t01jrKyshPO69Onj1wuV433SkpKVFBQoKSkJElSUlKSCgoKVFpa2uCxutC8DgAAjPAa3m7h+Fypqanas2dPjbG0tDRNmDDhlHO43W7FxMTI6XRKkpxOp6Kjo+V2u2VZVoPGoqKiav08EisAANCkZWVlyePx1HgvIiIiQNHUjcQKAAAYYcnwdgs//PPny3unw+VyqaioSB6PR06nUx6PR8XFxXK5XLIsq0FjdaHHCgAAGFG9j5XJ5vVfHlObNm2UkJCg3NxcSVJubq4SEhIUFRXV4LG6OCzLjttvmeX99krJs+eU5wE4fYlxvQMdAnBGi+nQTi8XLgl0GJKkoXlPa++R74zNFxfWSmsSJ9X7/Llz52r16tXav3+/WrdurcjISK1cuVLbt2/X1KlTVVZWpoiICKWnp+v888+XpAaP1YbESiRWQGMisQIaV5NKrP7vGe0xmFi1D2ulNddONDafP7AUCAAAYAjN6wAAwAhLPzacm5rPbkisAACAEQ3ZLf1U89kNS4EAAACGULECAABmsBZIxQoAAMAUKlYAAMAIeqxIrAAAgCHVO6+bnc9uWAoEAAAwhIoVAAAwgqVAKlYAAADGULECAACGOCSjVSYqVgAAAGctKlYAAMAI7goksQIAAKaw8zpLgQAAAKZQsQIAAEaw3QIVKwAAAGOoWAEAADPosSKxAgAAhhheCjS7J5Z/sBQIAABgCBUrAABgBkuBVKwAAABMoWIFAAAMsl9flEkkVgAAwAyWAlkKBAAAMKXWitUDDzwgh+PU5byMjAyjAQEAABuzYZXJpFoTqw4dOvgzDgAAANurNbFKS0vzZxwAAMDuLIfZTT1tuEFovZvXP/nkE61cuVKlpaXKzMxUfn6+Dh8+rAEDBjRmfAAAALZRr+b1pUuXavbs2erYsaM+//xzSVLz5s31zDPPNGpwAADAPizL/GE39Uqs/ud//kcvvviixo0bp6Cg6kvOP/98FRYWNmpwAADARqxGOGymXolVeXm5XC6XJPnuFKyqqlKzZs0aLzIAAACbqVdi1bdvX/3xj3+s8d5f/vIX9e/fv1GCAgAANnS8ed3kYTP1al5/8MEHdddddyk7O1vl5eVKTExUeHi4MjMzGzs+AAAA26hXYhUdHa3ly5crPz9fe/bskcvlUq9evXz9VgAAAA6r+jA5n93Ue7sFr9eryspKSZLH45Flx1Z9AADQeHhWYP0Sq02bNmn8+PGqqKhQTEyM9u3bp9DQUC1evFjdu3dv7BgBAABsoV6J1fTp05WamqpbbrlFDodDlmXppZde0vTp07VixYrGjhEAANgBO6/X767AnTt3auzYsb6tFhwOh8aMGaOdO3c2ZmwAAAC2Uq/EavDgwXr33XdrvPfee+/pyiuvbIyYAACAXZ3Fm4NKdSwFPvDAA74Klcfj0T333KMLL7xQsbGx2rdvn7788ksNHTrUb4ECAIAmjub12hOrDh061HjdtWtX389dunTRwIEDGy8qAAAAG6o1sUpLS/NnHAAAwO6oWNV/H6uKigoVFhbqwIEDNfawGjBgQKMEBgAAYDf1SqzWr1+vSZMmqaKiQocPH1Z4eLjKy8sVGxurNWvWNHaMAADADthuoX53BT766KO6/fbbtW7dOrVs2VLr1q3T3XffrZSUlMaODwAAwDbqvY/VmDFjarw3btw4vfTSS40REwAAsCPrx+cFmjjs2GNVr8TqnHPO0eHDhyVJ7dq107Zt21RWVqYjR440anAAAMBGTO5h1YDE6r333tN1112nUaNGacSIEVq9erUkqbCwUMnJyUpMTFRycnKNDc7rGmuIeiVWw4YN0wcffCBJ+u1vf6sxY8bo+uuv17XXXvuLPhwAAMAEy7I0efJkZWRk6I033tBjjz2mKVOmyOv1atasWUpJSVFeXp5SUlI0c+ZM33V1jTVEvZrXZ8yY4fv51ltvVa9evVReXq4rrrjiF304AADAqbjdbnk8nhrvRUREKCIiosZ7QUFBOnTokCTp0KFDio6O1oEDB1RQUKAXX3xRkpSUlKRHHnlEpaWlsiyr1rGoqKgGxVrv7RZ+qk+fPg36MAAAgNOVmpqqPXv21HgvLS1NEyZM8L12OBx6+umn9fvf/15hYWEqLy/Xc889J7fbrZiYGDmdTkmS0+lUdHS03G63LMuqdcx4YpWSkuJ7pE1dsrKyGvTBTcmYfj1UtKtdoMMAzkjOhAsCHQJwRnPGRQY6BB9f07nB+aTqXONkFaufqqqq0nPPPaclS5bokksu0T/+8Q/dc889ysjIMBdQPdSaWN1www3+jAMAANhdI+1j5XK5Tnnqv//9bxUXF+uSSy6RJF1yySVq0aKFQkNDVVRUJI/HI6fTKY/Ho+LiYrlcLlmWVetYQ9WaWI0ePbrBkwIAAPhTbGys9u3bpx07duj888/X9u3btX//fnXo0EEJCQnKzc3VqFGjlJubq4SEBN9SX11jDdGgHisAAICTCtDeU+3atdPs2bM1ceJEXyvTo48+qsjISM2ePVtTp07VkiVLFBERofT0dN91dY01BIkVAAA4I4wcOVIjR4484f3OnTsrOzv7pNfUNdYQJFYAAMAM07ul23DndRIrAABgRGPdFWgn9dp5vaKiQk899ZSGDh3q67b/+OOP9fLLLzdqcAAAAHZSr8Rq/vz52rJlix5//HFfQ9gFF1ygZcuWNWpwAADARgL8rMCmoF5Lge+8845Wr16tsLAwBQVV52IxMTEqKipq1OAAAADspF6JVbNmzU7Y8bS0tFSRkZGNERMAALAjmtfrtxR47bXXasqUKfrmm28kScXFxZozZ47+4z/+o1GDAwAAsJN6JVb33HOP2rdvr5EjR6qsrEyJiYmKjo7W+PHjGzs+AABgE8fvCjR52E29lgJDQkI0Y8YMzZgxQ6WlpWrdunW9HtAMAADOIo30rEA7qVdidXwJ8Ljy8nLfz/Hx8WYjAgAAsKl6JVbDhg2Tw+GQZf1Ykztesfr3v//dOJEBAAB7oXm9fonVpk2barz+9ttvtWjRIvXp06dRggIAALCjejWv/1y7du00Y8YMPfnkk6bjAQAANuWQ4eb1QH+hBmjwswJ37Niho0ePmowFAADYGUuB9UusUlJSatwFePToUW3bto3tFgAAAH6iXonVDTfcUON1ixYt1L17d3Xs2LExYgIAADZkeu+pM3IfK4/Ho88++0yPPPKIQkJC/BETAACALZ0ysXI6nfrkk0/YEBQAANSNHqv63RU4duxYLVy4UJWVlY0dDwAAsCurEQ6bqbNilZubq6SkJL388svav3+/XnzxRUVFRdWoXr3//vuNHSMAAIAt1JlYzZw5U0lJSXrsscf8FQ8AALApmtdPkVgdf4RNv379/BIMAACAndWZWHm9Xn322Wc1nhH4cwMGDDAeFAAAgB3VmVhVVFRoxowZtSZWDodDa9asaZTAAAAA7KbOxKpFixYkTgAAoH7YbqHhzwoEAAD4KZrXT7GPVV29VQAAAKipzorVhg0b/BUHAAA4E5zlNZl67bwOAACAU6PHCgAAmEHzOokVAAAwxHDzuh0TK5YCAQAADKFiBQAAzGApkIoVAACAKVSsAACAEWwQSmIFAABMYSmQpUAAAABTqFgBAAAzqFhRsQIAADCFihUAADCC5nUqVgAAAMZQsQIAAObYsMpkEokVAAAwg+Z1lgIBAABMoWIFAACMoHmdihUAAIAxVKwAAIAZ9FiRWAEAADNYCmQpEAAAnCGOHTumWbNm6ZprrtGIESP00EMPSZIKCwuVnJysxMREJScna+fOnb5r6hprCBIrAABghtUIx2l47LHHFBoaqry8POXk5GjixImSpFmzZiklJUV5eXlKSUnRzJkzfdfUNdYQJFYAAKBJc7vd2r17d42jrKysxjnl5eV6/fXXNXHiRDkcDklS27ZtVVJSooKCAiUlJUmSkpKSVFBQoNLS0jrHGooeKwAAYEYjNa+npqZqz549NYbS0tI0YcIE3+tvvvlGkZGRWrRokdauXauWLVtq4sSJat68uWJiYuR0OiVJTqdT0dHRcrvdsiyr1rGoqKgGhUxiBQAAjHD8cJicT5KysrLk8XhqjEVERNR4XVVVpW+++UY9evTQlClTtHHjRt1111165plnDEZ0aiRWAACgSXO5XKc8Jy4uTsHBwb5lvV/96ldq3bq1mjdvrqKiInk8HjmdTnk8HhUXF8vlcsmyrFrHGooeKwAAYEYAm9ejoqLUv39/ffLJJ5Kq7/YrKSlRx44dlZCQoNzcXElSbm6uEhISFBUVpTZt2tQ61lBUrAAAwBnh4Ycf1vTp05Wenq7g4GBlZGQoIiJCs2fP1tSpU7VkyRJFREQoPT3dd01dYw1BYgUAAIxwyPAGoad5fnx8vJYuXXrC+507d1Z2dvZJr6lrrCFYCgQAADCEihUAADCDZwWSWAEAAENIrFgKBAAAMIWKFQAAMMJhGW5ep2IFAABw9qJiBQAAzKDHisQKAAAYYngp0I6JFUuBAAAAhlCxAgAAZrAUSMUKAADAFCpWAADACLZbILECAAAm2TAZMomlQAAAAEOoWAEAADNoXqdiBQAAYAoVKwAAYATN61SsAAAAjKFiBQAAzKDHisQKAACY4bAsOSxz2ZDJufyFpUAAAABDqFgBAAAzWAqkYgUAAGAKFSsAAGAE2y2QWAEAAFNYCmQpEAAAwBQqVgAAwAiHDC8FmpvKb6hYAQAAGELFCgAAmEGPFYkVAAAwg7sCWQoEAAAwhooVAAAwg6VAKlYAAACmULECAABG0GNFYgUAAEyxrOrD5Hw2w1IgAACAIVSsAACAGYaXAmleBwAAOItRsQIAAGaw3QIVKwAAAFOoWAEAACMcluTwmp3PbkisAACAGSwFshQIAABgChUrNFmTF+5S74GH1TzMqwPFwcp+Nlr/99c2umr0AU3M2O07zxFkqXkLS+MTL9C2/LAARgw0bcHNPBo/cYMu/nWxws+pkHtPuP7nhZ5av86l4GCvJs9Yqwu6HlBM7BFNuXeQ8jdG+65NHfOVklM3qbLyx7+Pj79jmPa5wwPxVdBEsfM6iRWasP9dGKOn7otXZUWQ4rt8r4xXt2tbfgu991prvfdaa995w35XqpRJRdqW3yKA0QJNn9NpaX9xC02+Z7C+LQ5T3/5uTX1orX5/+zCVlrTQV/lt9fryCzR95mcnvf7D9+P1+KP9/Bw1YC8kVmiydm1p7vv5+FMS4jpWnFCVGnZDqd55tbUkh58jBOzl2PfByvpLT9/rdZ/FqWhfS13Q9YA++ail3lhxgSTJ6+XPEhqIR9r4J7E6evSopkyZom3btik4OFidOnVSSkqK5s2bp549e2rTpk1yOp1asGCBunTpom+//Vb33nuvysvLdezYMQ0ePFiTJ0+WJC1cuFA7duzQ4cOHtXPnTvXs2VPjxo3TggULtHfvXg0bNkxTpkzxx9eCH6TN361hyaVq3sLS1vwWWrfmnBrj0e0rdOGl5Xri3vgARQjYV2Tr79X+3EPatTOiXuf3v3Sv/ve1N1Ra2kI5r3fWqpzOjRwh7IalQD8lVh9//LHKysq0atUqSdJ3332nTZs2afPmzXrwwQfVr18/vfbaa5o8ebJWrFihiIgIZWZmqmXLlqqsrNRtt92mDz/8UIMGDZIkffXVV1q+fLnCwsI0evRoPfHEE3r++edVVVWloUOHKjk5WR07dvTHV0MjWzT9XC15sL0S+pSr14ByVVbUvN/i6hsO6Mu1LVX0TWiAIgTsyen06oFp6/TO6g7a/c2pE6sPP4jXWyvP18EDzdWte4lmzP5M5Yeb6YP3zvNDtIB9+OWuwO7du2vHjh16+OGH9dZbbykkJESS1KFDB/XrV71eP2rUKG3ZskWHDx+Wx+NRRkaGRo4cqeuvv15bt27Vpk2bfPMNHDhQ55xzjpxOp7p166bLLrtMISEhCgsLU6dOnfT111/742vBT7xeh75aF652rkoljd1fY+zq35bqneyoAEUG2JPDYen+aetUVRWkZ/9wcb2u+WZXhEpLWsjrdejfBW31xoouGjhoTyNHCtuxGuFogEWLFqlbt27asmWLJKmwsFDJyclKTExUcnKydu7c6Tu3rrGG8EtiFR8fr1WrVunyyy/Xp59+qlGjRunYsWO1nv/iiy+qrKxM2dnZysnJ0dVXX13j/NDQH6sTTqfzhNcej6dxvggCKijYkqtDhe91j77lahNbpY9yWwUwKsBuLE26f70iWx/TvNkD5PE07D8DlmTPdRqc8b766it98cUXiouL8703a9YspaSkKC8vTykpKZo5c2a9xhrCL4nVvn375HQ6dfXVV2vatGkqLS3Vd999p127dmn9+vWSpJycHHXt2lXh4eE6dOiQ2rVrp9DQUBUVFWnNmjX+CBNNSKs2lRo86oCah3kUFGTpksFluuq6g9r4yY+3dg+7oVQfr2ylo+XOAEYK2EvapA2KP++QHp5xuSoqav7ZCW7mUbNm1X8xDQ72/vBzdfJ06WV7FR5eIclS126lGjl6mz77e5yAnzreY2XykCS3263du3fXOMrKyk74/IqKCs2ZM0ezZs2Sw1F9E0ZJSYkKCgqUlJQkSUpKSlJBQYFKS0vrHGsov/RYbd68WU888YQkyev1aty4cYqOjlZCQoJyc3M1f/58BQUFKSMjQ5J08803a+LEibruuusUGxurAQMG+CNMNCWWQ0ljSvT/FuyWI0gq3h2izJlx+jSvujrVLNSrQSMO6pE7OgY2TsBGoqPLNXzEDlVUBCnr1Rzf+wufukTvrzlPf3opTzGxRyRJ8zI+liT9d8pvVFzUUoOu+kaTHlivZs082v9tmF79WzetWd0xEF8DTZrhuwJ/SOxTU1O1Z0/Npee0tDRNmDChxnvPPPOMRo4cqfj4H29ocrvdiomJkdNZ/RcJp9Op6Ohoud1uWZZV61hUVMPaTPySWA0ePFiDBw+u8d7atWsVFBSk2bNnn3B++/bt9eqrr550rp//EhcsWFDj9dKlS39ZsGgSvisN1gP/2aXW8cpjQfrPhIv8GBFgf8XFLTV86G9rHb8ldXitYxnz+jdGSEC9ZGVlndDmExFR86aLDRs2KD8/X/fff78/QzsB+1gBAAAjGmu7BZfLdcpzP//8c+3YsUNDhw6VVN2GdNttt2natGkqKiqSx+Px9WEXFxfL5XLJsqxaxxoqYM8K7N+/v1asWBGojwcAAGeQcePG6eOPP9a7776rd999V7Gxsfrzn/+s4cOH+1qPJCk3N1cJCQmKiopSmzZtah1rKCpWAADAjF+wRUKt8xkwe/ZsTZ06VUuWLFFERITS09PrNdYQJFYAAOCM8+677/p+7ty5s7Kzs096Xl1jDUFiBQAAjOCRNiRWAADAFK9VfZicz2YC1rwOAABwpqFiBQAAzGiizev+RMUKAADAECpWAADACJrXSawAAIAxjfOsQDthKRAAAMAQKlYAAMAMw0uBNixYUbECAAAwhYoVAAAwg+0WSKwAAIAZ1XcFmsuG7HhXIEuBAAAAhlCxAgAAZnh/OEzOZzNUrAAAAAyhYgUAAIxwWJbhHiv7NVlRsQIAADCEihUAADCD7RZIrAAAgCk8K5ClQAAAAEOoWAEAACMchp8VyAahAAAAZzEqVgAAwAzLcI+VDbdbILECAABGOLzVh8n57IalQAAAAEOoWAEAADNYCqRiBQAAYAoVKwAAYAY7r5NYAQAAM3gIM0uBAAAAxlCxAgAAhvCsQCpWAAAAhlCxAgAAZnh/OEzOZzNUrAAAAAyhYgUAAIzgrkASKwAAYIolwzuvm5vKX1gKBAAAMISKFQAAMINnBVKxAgAAMIWKFQAAMIPtFkisAACAGQ4ZvivQht3rLAUCAAAYQsUKAACYQfM6FSsAAABTqFgBAAAzqFiRWAEAAENIrFgKBAAAMIWKFQAAMIN9rKhYAQAA+ztw4IDuuOMOJSYmasSIEUpLS1NpaakkqbCwUMnJyUpMTFRycrJ27tzpu66usYYgsQIAAGZY1RuEmjpOp8fK4XDo9ttvV15ennJychQfH6/HH39ckjRr1iylpKQoLy9PKSkpmjlzpu+6usYagsQKAAA0aW63W7t3765xlJWV1TgnMjJS/fv3973u3bu39u7dq5KSEhUUFCgpKUmSlJSUpIKCApWWltY51lD0WAEAADMa6a7A1NRU7dmzp8ZQWlqaJkyYcNLLvF6vli1bpiFDhsjtdismJkZOp1OS5HQ6FR0dLbfbLcuyah2LiopqUMgkVgAAwAxLktdkYlX9j6ysLHk8nhpDERERtV72yCOPKCwsTDfddJMKCgrMxVMPJFYAAKBJc7lc9T43PT1du3btUmZmpoKCguRyuVRUVCSPxyOn0ymPx6Pi4mK5XC5ZllXrWEPRYwUAAMw4vhRo8jgNTz31lL788kstXrxYISEhkqQ2bdooISFBubm5kqTc3FwlJCQoKiqqzrGGomIFAABsb+vWrcrMzFTHjh114403SpLOPfdcLV68WLNnz9bUqVO1ZMkSRUREKD093XddXWMNQWIFAADMCOAjbS644AJt3rz5pGOdO3dWdnb2aY81BIkVAAAwg2cF0mMFAABgChUrAABghtcyu92Cybn8hIoVAACAIVSsAACAGZa3+jA5n82QWAEAAEMMN6+LpUAAAICzFhUrAABghleGm9fNTeUvVKwAAAAMoWIFAADMYINQKlYAAACmULECAABmULEisQIAAIaQWLEUCAAAYAoVKwAAYIbXW32YnM9mqFgBAAAYQsUKAAAYwiNtSKwAAIAZNK+zFAgAAGAKFSsAAGAGzwqkYgUAAGAKFSsAAGCG5ZVlGSwzmZzLT0isAACAGV7L8FIgzesAAABnLSpWAADADLZboGIFAABgChUrAABghmX4WYE2bF6nYgUAAGAIFSsAAGAGPVYkVgAAwAzLa8kyuBRosd0CAADA2YuKFQAAMIOlQCpWAAAAplCxAgAAZvBIGxIrAABgitfw3lPsYwUAAHDWomIFAACMqN5uwdzyHdstAAAAnMWoWAEAADMsy2yPlQ23WyCxAgAARrAUyFIgAACAMVSsJLVtHxXoEIAzljMuMtAhAGe0tjERgQ7Bp2371kaXAtu2b21sLn9xWJYNFzABAACaIJYCAQAADCGxAgAAMITECgAAwBASKwAAAENIrAAAAAwhsQIAADCExAoAAMAQEisAAABDSKwAAAAMIbECAAAwhMQKAdetWzeVl5cHOgwAAH4xEisAAABDggMdACBJS5cu1dtvv62DBw9q8uTJSkxMlCTdd999KiwsVGVlpc477zzNnz9frVq10tq1azVv3jz16tVLGzduVHBwsDIyMrRo0SJt3bpVLpdLCxcuVFhYWIC/GRA4R48e1ZQpU7Rt2zYFBwerU6dOSklJ0bx589SzZ09t2rRJTqdTCxYsUJcuXfTtt9/q3nvvVXl5uY4dO6bBgwdr8uTJkqSFCxdqx44dOnz4sHbu3KmePXtq3LhxWrBggfbu3athw4ZpypQpAf7GQBNgAQHWtWtXa+nSpZZlWdb69eutgQMH+sZKSkp8Pz/55JPWY489ZlmWZX322WdWjx49rIKCAsuyLGv27NnWFVdcYbndbsuyLOv222+3XnnlFX99BaBJWr16tTV27Fjf64MHD1qfffaZ1bVrV2vt2rWWZVnWihUrrNGjR1uWZVnff/+9dfjwYcuyLKuiosK6+eabrQ8++MCyLMv6wx/+YA0bNswqKyuzqqqqrBEjRli33nqrdezYMau8vNy69NJLrcLCQr9+P6ApomKFJmH48OGSpN69e6u4uFjHjh1TaGio3njjDeXk5KiyslJHjhxRx44dfdd06tRJCQkJkqQePXpo7969io2NlST17NlTu3bt8vv3AJqS7t27a8eOHXr44YfVr18/XXnllZKkDh06qF+/fpKkUaNG6aGHHtLhw4cVFBSkjIwMbdiwQZZlaf/+/dq0aZMGDRokSRo4cKDOOeccSdW9kd27d1dISIhCQkLUqVMnff311zX+jAJnI3qs0CSEhoZKkpxOpySpqqpK69ev17Jly/T8888rJydHkyZNUkVFhe+akJAQ389Op9M3x/HXHo/HT9EDTVN8fLxWrVqlyy+/XJ9++qlGjRqlY8eO1Xr+iy++qLKyMmVnZysnJ0dXX311jfN//meMP3PAiUis0GSVlZUpPDxckZGRqqio0PLlywMdEmAr+/btk9Pp1NVXX61p06aptLRU3333nXbt2qX169dLknJyctS1a1eFh4fr0KFDateunUJDQ1VUVKQ1a9YE+BsA9sNSIJqsQYMG6c0339RvfvMbxcTE6MILL1R+fn6gwwJsY/PmzXriiSckSV6vV+PGjVN0dLQSEhKUm5ur+fPn+5b/JOnmm2/WxIkTdd111yk2NlYDBgwIZPiALTksy7ICHQQAwD/Wrl2r9PR0rVixItChAGcklgIBAAAMoWIFAABgCBUrAAAAQ0isAAAADCGxAgAAMITECkANU6dO1VNPPSVJWr9+ve+5jY2tW7dute6Wf/PNNys7O7te8wwZMkR///vfGxTDL7kWACQSK8CWhgwZol69euniiy/WZZddpmnTpqm8vNz45/Tp00d5eXmnPG/FihX6r//6L+OfDwB2Q2IF2FRmZqY2bNig1157Tfn5+Xr22WdPOKeqqioAkQHA2YvECrC5mJgYXXHFFdq6dauk6iW1rKwsXXPNNbrmmmskSe+9955GjRqlPn366MYbb9SmTZt81xcUFGj06NG6+OKLNWnSpBrPhlu7dq3vAbyS5Ha7lZaWpksvvVT9+/fXnDlztH37ds2aNUtffPGFLr74YvXp00eSVFFRofT0dF155ZW67LLLNHPmTH3//fe+uZ5//nkNHDhQAwcO1Kuvvlrv7/v1119rzJgx6t+/v/r376/77rtPZWVlNc7Jz8/X8OHD1bdvX02bNq3Gd6rrdwEAvxSJFWBzbrdbH374oRISEnzvvfPOO3rllVe0atUqffXVV5o+fbrmzJmjtWvXKjk5Wb///e9VUVGhiooKjR8/XqNGjdK6det07bXXavXq1Sf9HI/HozvvvFNxcXF699139eGHH2r48OHq3LmzHn74YfXu3VsbNmzwPYPuscceU2FhoV5//XWtXr1axcXFWrx4sSTpww8/1AsvvKAXXnhBq1ev1qefflrv72tZlu6880599NFHeuutt7Rv3z4tXLiwxjk5OTn685//rLfffluFhYVasmSJJNX5uwAAE0isAJsaP368+vTpo5SUFPXt21d33XWXb2zcuHGKjIxU8+bN9corryg5OVm/+tWv5HQ6NXr0aDVr1kxffPGFNm7cqMrKSo0dO1bNmjXTtddeq4suuuikn/evf/1LxcXFmjx5ssLCwhQaGuqrTv2cZVnKzs7W9OnTFRkZqfDwcN15551auXKlJOmtt97S9ddfr65duyosLExpaWn1/t4dOnTQ5ZdfrpCQEEVFRemWW27R559/XuOc1NRUuVwuRUZG6u677/Z9bl2/CwAwgYcwAza1ePFiXXbZZScdc7lcvp/37t2r119/XS+//LLvvcrKShUXF8vhcCgmJkYOh8M3FhcXd9I53W634uLiFBx86v/bKC0t1dGjR3X99df73rMsS16vV5JUXFysCy+80DfWvn37U855XElJiebOnav169ervLxclmUpIiKixjk//f5xcXEqLi6WVPfvAgBMILECzkA/TZRcLpfuuusu3X333Sect27dOhUVFcmyLN81e/fuVXx8/Annulwuud1uVVVVnZBc/fTzJKl169Zq3ry5Vq5cqZiYmBPmio6Oltvt9r3eu3dvvb/bE088IYfDoTfffFOtW7fWO++8ozlz5tQ45+dzR0dH+75Dbb8LADCBpUDgDHfDDTfob3/7mzZu3CjLsnTkyBG9//77Onz4sHr37q3g4GD95S9/UVVVlVavXq38/PyTztOrVy+1a9dOTzzxhI4cOaJjx47pH//4hySpTZs2Kioq8vUqBQUF6YYbbtD8+fNVUlIiSSoqKtJHH30kSbr22mv12muvadu2bTp69KgWLVpU7+9TXl6usLAwRUREqKioSM8///wJ5/z1r3/Vvn37dPDgQT333HMaPnz4KX8XAGACiRVwhrvooov0yCOPaM6cOerbt6+uueYarVixQpIUEhKihQsX6rXXXlPfvn21atUqDRs27KTzOJ1OZWZmateuXbrqqqs0aNAgvfXWW5KkSy+9VF26dNHAgQPVv39/SdIDDzygDh066He/+51+/etf67//+79VWFgoSRo8eLDGjh2rsWPHatiwYbr00kvr/X3S0tJUUFCgPn36aNy4cb47H38qKSlJt956q66++mrFx8f7KlR1/S4AwASHZVlWoIMAAAA4E1CxAgAAMITECgAAwBASKwAAAENIrAAAAAwhsQIAADCExAoAAMAQEisAAABDSKwAAAAM+f8JPAIJtjgXgQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************RandomForestClassifier******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       0.98      0.99      0.99      1587\n",
      "        spam       0.96      0.88      0.92       252\n",
      "\n",
      "    accuracy                           0.98      1839\n",
      "   macro avg       0.97      0.94      0.95      1839\n",
      "weighted avg       0.98      0.98      0.98      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************SVC******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       0.94      0.97      0.95      1587\n",
      "        spam       0.76      0.62      0.68       252\n",
      "\n",
      "    accuracy                           0.92      1839\n",
      "   macro avg       0.85      0.80      0.82      1839\n",
      "weighted avg       0.92      0.92      0.92      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlYAAAHlCAYAAAA3E7JZAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/Il7ecAAAACXBIWXMAAAsTAAALEwEAmpwYAAA0+klEQVR4nO3de1xVZb7H8e9mIygaIiawNQZM09Aym7xkWTZeonFIs5liDpROl7Ga8GhTeZ1RM7Ow+3g5NqdypmKcV052Ae1oZtNtyrIxsxjvqAkbUJAQRIG11/mD3EUKIj2wWfp5z2u9Xuz1rP3s3/Y1xs/f71nPctm2bQsAAAA/WlCgAwAAADhdkFgBAAAYQmIFAABgCIkVAACAISRWAAAAhpBYAQCAH832lQQ6hBbBxXYLkq/o15IvP9BhAKelsQN6BToE4LR2dpdIPfXB3ECHIUnyFf2X2d+nQTEK6rjM3HzNIDjQAbQIvnzJyg10FMBpqWBPp0CHAKC58PuUxAoAAJjhk0+Sz+iMTluzRGIFAACMsGyfZBtMrGyf4xIVpyWCAAAALZbTEkEAANBC2d/+zxSXwbmaCxUrAAAAQ6hYAQAAI3yyZRtcvO7EihWJFQAAMMKybZncHtPlwK02aQUCAAAYQsUKAAAYUdMINFdlCnJgK5CKFQAAgCFUrAAAgBGW4YqVya0bmgsVKwAAAEOoWAEAACN8hitWcmDFisQKAAAY4bNtWSa3SGC7BQAAgDMXFSsAAGCE79vDFJfBuZoLFSsAAABDqFgBAAAjLNmyWLwOAADw4/lsyTKYC7mcl1fRCgQAADCFihUAADDC9OJ1k3M1FypWAAAAhlCxAgAARlhyyTK4SYLLgRsukFgBAAAjfHbNYXI+p6EVCAAAYAgVKwAAYITPcCswyIGtQCpWAAAAhlCxAgAARphevO7EihWJFQAAMMJnu+SzzSVDJudqLrQCAQAADKFiBQAAjDC9eN3twFYgFSsAAABDqFgBAAAjLAXJMlizMTlXc3FexAAAAC0UiRUAADCi5pE2LoPHqX1+enq6hg4dqp49e2rbtm3HjS9cuPC4sZycHCUnJysxMVHJycnavXt3g8bqQmIFAACMOLZ43dThO8XF68OGDVNGRoa6dOly3NhXX32lzz//XJ07d651ftasWUpJSdHq1auVkpKimTNnNmisLiRWAACgRfN6vdq3b1+to7S09Ljr+vXrJ4/Hc9z5yspKzZkzR7NmzZLL9V2yVlRUpOzsbCUlJUmSkpKSlJ2dreLi4nrH6sPidQAAYISlIFm2+cXrqampys3NrTWWlpamCRMmNGiep59+WqNGjVJsbGyt816vV9HR0XK73ZIkt9utqKgoeb1e2bZd51hkZGSdn0ViBQAAWrSMjAxZllXrXHh4eIPeu3HjRm3evFn33XdfU4R2HBIrAABghE9B8hlcZXRsrhO19xrq008/1a5duzRs2DBJUn5+vm677TY9/PDDSkhIUEFBgSzLktvtlmVZKiwslMfjkW3bdY7VhzVWAADAiEAvXj+R8ePH64MPPtC6deu0bt06xcTE6LnnntPgwYPVsWNHJSQkKCsrS5KUlZWlhIQERUZG1jtWHypWAADgtDB37lytWbNGBw4c0C233KKIiAitXLmy3vfMnj1bU6dO1eLFixUeHq709PQGjdXFZdv2Ke4Scfrx7b9KsnJPeh2AU5fYuW+gQwBOa9FxnfRSzuJAhyFJev/rYTpSnWdsvtbBnXVF7NvG5msOtAIBAAAMoRUIAACM8BlaF/X9+ZyGxAoAABjhM/wQZpN3GDYX50UMAADQQlGxAgAARli24Z3XDc7VXJwXMQAAQAtFxQoAABjRVDuvO4nzIgYAAGihqFgBAAAjfLZk2Qa3W3DgFuYkVgAAwAjL8HYLJudqLs6LGAAAoIWiYgUAAIzw2UHyGdwiweRczcV5EQMAALRQVKwAAIARPNKGxAoAABhi2S6jdwWanKu5OC8VBAAAaKGoWAEAACN8chneeZ2KFQAAwBmLihUAADDCZwfJOsO3WyCxAgAARtS0Ag0+0oZWIAAAwJmLihUAADDCMtwKNDlXc3FexAAAAC0UFSsAAGCETy7DO6+zxgoAAOCMRcUKAAAY4ZNLPoOPoXFixYrECgAAGGEZfgizybmai/MiBgAAaKGoWAEAACNsO8jobuk22y0AAACcuahYAQAAIyy5ZBlccG5yruZCYgUAAIzw2S6jrUCTdxg2F1qBAAAAhlCxAgAARvgMtwKduI8VFSsAAABDqFgBAAAjfIa3WzA5V3MhsQIAAEZYdpAsg8mQybmai/MiBgAAaKGoWAEAACNsuYwuOLdZvA4AAHDmomIFAACMsGyX4TVWVKwAAADOWFSsAACAETWPtDG4QagDK1YkVgAAwAifgmQZbIb5HNhYc17EAAAALRQVKwAAYAStQCpWAAAAxlCxAgAARvgUZHRdlBPXWJFYAQAAIyy5jO49ZbHzOgAAwJmLxAoAABhh298tYDdx2PapfX56erqGDh2qnj17atu2bZKkgwcP6re//a0SExN17bXXKi0tTcXFxf735OTkKDk5WYmJiUpOTtbu3bsbNFYXEisAAHBaGDZsmDIyMtSlSxf/OZfLpdtvv12rV69WZmamYmNj9dhjj/nHZ82apZSUFK1evVopKSmaOXNmg8bqQmIFAACM8NlBxg9J8nq92rdvX62jtLT0uM/v16+fPB5PrXMREREaOHCg/3Xfvn2Vl5cnSSoqKlJ2draSkpIkSUlJScrOzlZxcXG9Y/Vh8ToAADDCksvogvNjc6Wmpio3N7fWWFpamiZMmHBK8/l8Pi1btkxDhw6VVJOwRUdHy+12S5LcbreioqLk9Xpl23adY5GRkXV+BokVAABo0TIyMmRZVq1z4eHhpzzPgw8+qLCwMN10002mQjsOiRUAADDCNrzzuv3tXD9s7zVGenq69uzZoyVLligoKMg/b0FBgSzLktvtlmVZKiwslMfjkW3bdY7VhzVWAADgtPbkk0/qyy+/1KJFixQSEuI/37FjRyUkJCgrK0uSlJWVpYSEBEVGRtY7Vh+XbZ/qzYynH9/+qyQr96TXATh1iZ37BjoE4LQWHddJL+UsDnQYkqRpX9yrosoDxubrGHK2Hu7zeIOvnzt3rtasWaMDBw6oQ4cOioiI0FNPPaWkpCTFx8erdevWkqRzzjlHixYtkiTt3LlTU6dOVWlpqcLDw5Wenq5zzz33pGN1IbESiVWgvP782Xrr5Ujt3tJaV11Xovue2itJyv86ROMG9lLrsO/66TfeXajUewokSSv+t5Nef+5slRYHq3Vbn4aMOqjf/jFP7mCpcF8r/faq82t9zpHDbv12Zq5+def+5vty8COxarnm/2OHEn56WJZV0245kN9Kt19R8/en7+BDSpuXq05dKrX132F6bNJPVJgbUt90CBASq5aFNVYImI4xVUqZWKAN756lyiPHd6VXbNks9wn+H3rpiG909Y3FatfeUulBt+aOj9drz3XSL+/Yr6hzqvT6js3+a/P3huiWyxI0+BffNOVXARxr0R+66P/+1rHWufDIas18dreevC9WH78VrnGT8zV9yR5Nuva8AEUJp/DJJZ/BuwJNztVcArrGqmfPniovLw9kCAigwSO/0WU//0bhHapP6X2d4yvVrv231SxbcrmkvN2hJ7x27fIOuuDSMsXEVv7YcIEzxuU//0Z7trXW+1kRqjoapBcfi9a5vSoU2/1IoENDC2fZkmW7DB6B/kanjsXraLFuHtBLqZf00mOTYvVNkbvW2LoVERrT40LdcMGFysluo5E3nbj0vPYfkRpxw8HmCBdwpFumefXyl1/qide3q8+gMklSXM8j2vVVG/81Ryvc8u4JVVwPEivgZALeCnzxxRf11ltvqaSkRJMnT1ZiYqIk6d5771VOTo6qqqr0k5/8RPPmzVP79u21fv16PfTQQ+rTp482bdqk4OBgzZ8/XwsXLtT27dvl8Xi0YMEChYWFBfibobHaR1ZrwZtb1a13hUoPBmvh9HOUnhanect2+a8Zen2Jhl5fotxdIVr7j0h16HR81Wvz+rY6uD9YVySVNGP0gHM895BHe7e1VnWVS0NGl+iBv+bodyN6qE1b33H/mCk/FKQ27XwBihROYX9vt3RT8zlNwCNu166dXnnlFc2fP19z5871n58xY4ZWrFihzMxMde/eXf/7v//rH9u5c6dSU1OVmZmpvn376rbbbtO0adO0atUqBQUFaeXKlYH4KjCkTVufelxUIXew1KFTte5+aJ8+ezdc5YeO/79rl3MrFdfjiBZOO+e4sbUvR2rwL75Rm7b8MgBOZOvGtqood6uqMkhrl0cq+9O26j+sVBXlQQr7QRIV1s6nirKA/8oAWryAV6xGjhwpqebZPYWFhTp69KhCQ0P1+uuvKzMzU1VVVTp8+LDi4+P97+natasSEhIkSb169VJeXp5iYmIkSb1799aePXua/Xug6biOrV2so9duWS7l7am9xupohUvvZUVo1nM5TRsccBqxv12zuGdra4248bvnoYW2seSJP6o921oHMDo4gc/wBqEm52ouAf/nR2hozS/EY8/iqa6u1oYNG7Rs2TI9++yzyszM1KRJk1RZ+d3i4+9v7uV2u/1zHHv9w23v0TJZ1VLlEZd8lks+q+Znq1ra8u8wfb0jVD6fVFrs1uI/dFGfyw6pbXjNv6DfzIhUyYGafxPs2Raqvy+I0sWDD9Wa+8M326tduKWLLi9r9u8FOEHbcEuXDClVq1Cfgty2fjbmoC68tFyf/fMs/evN9orveUSDR5aoVahPN/2+QDnZbfT1DhIr1O/YXYEmD6cJeMXqREpLS9WuXTtFRESosrJSr7zySqBDQhP421MxeumJGP/rt1+J1E2/z9c53Y5o6SMelRwIVtuzfLr4ykOatvi7KuRXn7bVX9I9qigPUkRHS1cklWjcZG+tudcuj9TwG4q/q3YBqCU42Na4KfmK7X5UPkv6ekdrPXBrvPbtrEmeHvxtvO5+KFeTF+zVlo1heviuuABHDDhDi0ysrrzySr3xxhv6+c9/rujoaF1wwQXavHnzyd8IR7n5vnzdfF/+Ccd+Nqakzvfd99TXJ537+wvdARzvm+Jg/ffIHnWOb3z/LN1+5fl1jgMn4pPhVqADK1bsvC52XgeaEjuvA02rJe28PmHjNB04WmRsvrNDO2rBxQ8bm685tMiKFQAAcB62WyCxAgAAhvhss3fy+RzYU3NeKggAANBCUbECAABG8BBmKlYAAADGULECAABGsPM6FSsAAABjqFgBAAAjbMMVK9uBFSsSKwAAYAStQFqBAAAAxlCxAgAARlCxomIFAABgDBUrAABgBBuEklgBAABDuCuQViAAAIAxVKwAAIARPpldcO4zNlPzoWIFAABgCBUrAABgBNstkFgBAABDSKxoBQIAABhDxQoAAJhhu8xukUDFCgAA4MxFxQoAABjBzutUrAAAAIyhYgUAAIzgrkASKwAAYIhtePE6zwoEAAA4g1GxAgAARtAKpGIFAABgDBUrAABghm14XZRtbqrmQmIFAACM8MlwK5B9rAAAAM5cVKwAAIARtl1zmJzPaahYAQAAGELFCgAAGMGzAkmsAACAIey8TisQAADAGCpWAADACNvwzutUrAAAAM5gJFYAAMCIY9stmDxORXp6uoYOHaqePXtq27Zt/vM5OTlKTk5WYmKikpOTtXv37h89VhcSKwAAcFoYNmyYMjIy1KVLl1rnZ82apZSUFK1evVopKSmaOXPmjx6rC4kVAAAw4thdgSYPSfJ6vdq3b1+to7S09LjP79evnzweT61zRUVFys7OVlJSkiQpKSlJ2dnZKi4ubvRYfVi8DgAAjGiq7RZSU1OVm5tbaywtLU0TJkw46Rxer1fR0dFyu92SJLfbraioKHm9Xtm23aixyMjIOj+PxAoAALRoGRkZsiyr1rnw8PAARVM/EisAAGCEz/B2C8fm+mF771R4PB4VFBTIsiy53W5ZlqXCwkJ5PB7Ztt2osfqwxgoAAJy2OnbsqISEBGVlZUmSsrKylJCQoMjIyEaP1cdl2058drRZvv1XSVbuSa8DcOoSO/cNdAjAaS06rpNeylkc6DAkScPXPKW8wyXG5uscFqG1V09q8PVz587VmjVrdODAAXXo0EERERFauXKldu7cqalTp6q0tFTh4eFKT0/XueeeK0mNHqsLiZVIrICmRGIFNK2WlFgNW/2U8g5/Y2y+zmHt9XbiJGPzNQdagQAAAIaweB0AAJhheLsF8axAAACAMxcVKwAAYIT97WFyPqchsQIAAEY01c7rTkIrEAAAwBAqVgAAwAx6gVSsAAAATKFiBQAAjGCNFYkVAAAwxLZrDpPzOQ2tQAAAAEOoWAEAACNoBVKxAgAAMIaKFQAAMMRl+Pl+VKwAAADOWFSsAACAEdwVSGIFAABMYed1WoEAAACmULECAABGsN0CFSsAAABjqFgBAAAzWGNFYgUAAAwx3Ao0uydW86AVCAAAYAgVKwAAYAatQCpWAAAAplCxAgAABjlvXZRJJFYAAMAMWoG0AgEAAEyps2J1//33y+U6eTlv/vz5RgMCAAAO5sAqk0l1JlZxcXHNGQcAAIDj1ZlYpaWlNWccAADA6WyX2U09HbhBaIMXr3/44YdauXKliouLtWTJEm3evFllZWUaNGhQU8YHAADgGA1avP7iiy9q9uzZio+P16effipJat26tZ5++ukmDQ4AADiHbZs/nKZBidVf//pXLV26VOPHj1dQUM1bzj33XOXk5DRpcAAAwEHsJjgcpkGJVXl5uTwejyT57xSsrq5Wq1atmi4yAAAAh2lQYtW/f3/9+c9/rnXuhRde0MCBA5skKAAA4EDHFq+bPBymQYvX//CHP+jOO+/U8uXLVV5ersTERLVr105Llixp6vgAAAAco0GJVVRUlF555RVt3rxZubm58ng86tOnj3+9FQAAgMuuOUzO5zQN3m7B5/OpqqpKkmRZlmwnLtUHAABNh2cFNiyx2rJli+6++25VVlYqOjpa+fn5Cg0N1aJFi3T++ec3dYwAAACO0KDEavr06UpNTdUtt9wil8sl27b1l7/8RdOnT9eKFSuaOkYAAOAE7LzesLsCd+/erXHjxvm3WnC5XBo7dqx2797dlLEBAAA4SoMSqyFDhmjdunW1zr3zzju66qqrmiImAADgVGfw5qBSPa3A+++/31+hsixL99xzjy644ALFxMQoPz9fX375pYYNG9ZsgQIAgBaOxet1J1ZxcXG1Xvfo0cP/c/fu3TV48OCmiwoAAMCB6kys0tLSmjMOAADgdFSsGr6PVWVlpXJycnTw4MFae1gNGjSoSQIDAABwmgYlVhs2bNCkSZNUWVmpsrIytWvXTuXl5YqJidHbb7/d1DECAAAnYLuFht0V+PDDD+v222/XJ598orZt2+qTTz7RXXfdpZSUlKaODwAAwDEavI/V2LFja50bP368/vKXvzRFTAAAwIns754XaOJw4hqrBiVWZ511lsrKyiRJnTp10o4dO1RaWqrDhw83aXAAAMBBTO5h1YjE6p133tF1112n0aNH69prr9WaNWskSTk5OUpOTlZiYqKSk5NrbXBe31hjNCixGjFihN59911J0q9+9SuNHTtW119/va655pof9eEAAAAm2LatyZMna/78+Xr99df16KOPasqUKfL5fJo1a5ZSUlK0evVqpaSkaObMmf731TfWGA1avD5jxgz/z7feeqv69Omj8vJyXXHFFT/qwwEAAE7G6/XKsqxa58LDwxUeHl7rXFBQkA4dOiRJOnTokKKionTw4EFlZ2dr6dKlkqSkpCQ9+OCDKi4ulm3bdY5FRkY2KtYGb7fwff369WvUhwEAAJyq1NRU5ebm1jqXlpamCRMm+F+7XC499dRT+t3vfqewsDCVl5frmWeekdfrVXR0tNxutyTJ7XYrKipKXq9Xtm3XOWY8sUpJSfE/0qY+GRkZjfrgluQ3Qy5Rwd64k18I4JQFe8JPfhGARnNHNS4BaAr+RecG55Nqco0TVay+r7q6Ws8884wWL16sSy65RJ999pnuuecezZ8/31xADVBnYnXDDTc0ZxwAAMDpmmgfK4/Hc9JL//Of/6iwsFCXXHKJJOmSSy5RmzZtFBoaqoKCAlmWJbfbLcuyVFhYKI/HI9u26xxrrDoTqzFjxjR6UgAAgOYUExOj/Px87dq1S+eee6527typAwcOKC4uTgkJCcrKytLo0aOVlZWlhIQEf6uvvrHGaNQaKwAAgBMK0N5TnTp10uzZszVx4kT/UqaHH35YERERmj17tqZOnarFixcrPDxc6enp/vfVN9YYJFYAAOC0MGrUKI0aNeq48926ddPy5ctP+J76xhqDxAoAAJhherd0B+68TmIFAACMaKq7Ap2kQTuvV1ZW6sknn9SwYcP8q+0/+OADvfTSS00aHAAAgJM0KLGaN2+etm3bpscee8y/IOy8887TsmXLmjQ4AADgIAF+VmBL0KBW4Nq1a7VmzRqFhYUpKKgmF4uOjlZBQUGTBgcAAOAkDUqsWrVqddyOp8XFxYqIiGiKmAAAgBOxeL1hrcBrrrlGU6ZM0ddffy1JKiws1Jw5c/SLX/yiSYMDAABwkgYlVvfcc4+6dOmiUaNGqbS0VImJiYqKitLdd9/d1PEBAACHOHZXoMnDaRrUCgwJCdGMGTM0Y8YMFRcXq0OHDg16QDMAADiDNNGzAp2kQYnVsRbgMeXl5f6fY2NjzUYEAADgUA1KrEaMGCGXyyXb/q4md6xi9Z///KdpIgMAAM7C4vWGJVZbtmyp9Xr//v1auHCh+vXr1yRBAQAAOFGDFq//UKdOnTRjxgw98cQTpuMBAAAO5ZLhxeuB/kKN0OhnBe7atUsVFRUmYwEAAE5GK7BhiVVKSkqtuwArKiq0Y8cOtlsAAAD4ngYlVjfccEOt123atNH555+v+Pj4pogJAAA4kOm9p07Lfawsy9LHH3+sBx98UCEhIc0REwAAgCOdNLFyu9368MMP2RAUAADUjzVWDbsrcNy4cVqwYIGqqqqaOh4AAOBUdhMcDlNvxSorK0tJSUl66aWXdODAAS1dulSRkZG1qlf//Oc/mzpGAAAAR6g3sZo5c6aSkpL06KOPNlc8AADAoVi8fpLE6tgjbAYMGNAswQAAADhZvYmVz+fTxx9/XOsZgT80aNAg40EBAAA4Ub2JVWVlpWbMmFFnYuVyufT22283SWAAAABOU29i1aZNGxInAADQMGy30PhnBQIAAHwfi9dPso9VfWurAAAAUFu9FauNGzc2VxwAAOB0cIbXZBq08zoAAABOjjVWAADADBavk1gBAABDDC9ed2JiRSsQAADAECpWAADADFqBVKwAAABMoWIFAACMYINQEisAAGAKrUBagQAAAKZQsQIAAGZQsaJiBQAAYAoVKwAAYASL16lYAQAAGEPFCgAAmOPAKpNJJFYAAMAMFq/TCgQAADCFihUAADCCxetUrAAAAIyhYgUAAMxgjRWJFQAAMINWIK1AAAAAY0isAACAGXYTHKfg6NGjmjVrlq6++mpde+21+uMf/yhJysnJUXJyshITE5WcnKzdu3f731PfWGOQWAEAgNPCo48+qtDQUK1evVqZmZmaOHGiJGnWrFlKSUnR6tWrlZKSopkzZ/rfU99YY5BYAQAAM5qoYuX1erVv375aR2lpaa2PLi8v12uvvaaJEyfK5XJJks4++2wVFRUpOztbSUlJkqSkpCRlZ2eruLi43rHGYvE6AAAwwvXtYXI+SUpNTVVubm6tsbS0NE2YMMH/+uuvv1ZERIQWLlyo9evXq23btpo4caJat26t6Ohoud1uSZLb7VZUVJS8Xq9s265zLDIyslExk1gBAIAWLSMjQ5Zl1ToXHh5e63V1dbW+/vpr9erVS1OmTNGmTZt055136umnn27OUEmsAACAIU20j5XH4znppZ07d1ZwcLC/rXfRRRepQ4cOat26tQoKCmRZltxutyzLUmFhoTwej2zbrnOssVhjBQAAHC8yMlIDBw7Uhx9+KKnmbr+ioiLFx8crISFBWVlZkqSsrCwlJCQoMjJSHTt2rHOssahYAQAAI1wyvEHoKV7/wAMPaPr06UpPT1dwcLDmz5+v8PBwzZ49W1OnTtXixYsVHh6u9PR0/3vqG2sMEisAAHBaiI2N1Ysvvnjc+W7dumn58uUnfE99Y41BYgUAAMzgWYEkVgAAwBASKxavAwAAmELFCgAAGOGyDS9ep2IFAABw5qJiBQAAzGCNFYkVAAAwxHAr0ImJFa1AAAAAQ6hYAQAAM2gFUrECAAAwhYoVAAAwgu0WSKwAAIBJDkyGTKIVCAAAYAgVKwAAYAaL16lYAQAAmELFCgAAGMHidSpWAAAAxlCxAgAAZrDGisQKAACY4bJtuWxz2ZDJuZoLrUAAAABDqFgBAAAzaAVSsQIAADCFihUAADCC7RZIrAAAgCm0AmkFAgAAmELFCgAAGOGS4VaguamaDRUrAAAAQ6hYAQAAM1hjRWIFAADM4K5AWoEAAADGULECAABm0AqkYgUAAGAKFSsAAGAEa6xIrAAAgCm2XXOYnM9haAUCAAAYQsUKAACYYbgVyOJ1AACAMxgVKwAAYAbbLVCxAgAAMIWKFQAAMMJlSy6f2fmchsQKAACYQSuQViAAAIApVKzQYsV2O6zfzc7ReReU65viYD33SJz+9VZHRXU5or++u1EV5d/9u2D5n7to2aJzAhgt0PIl3bhHw6/NU3z3Q3p3tUdPzr5QkhTlqdDSrPdUcdjtv/Yff+2qvz/bTZL0wJ8+U++LD/rHglv5lLunre5Ovrx5vwBaPHZeJ7FCCxXktjVzyVatWhatGeN66cIBpZr95y1KGxWmqiqXJOlXPx0gn+UKcKSAcxTvb62/P3eufjqoSKGh1nHjN141VD7r+EbGrP++pNbrh5/5RF9siGyyOAEnoxWIFin23Ap1jKrUq8975PO5tOnj9sr+91kaet3+QIcGONa/3onWx/+M1qGSVo2eI8pTod4XH9S6lZ0NRobTxrFH2pg8HKZZKlYVFRWaMmWKduzYoeDgYHXt2lUpKSl66KGH1Lt3b23ZskVut1uPPPKIunfvrv379+v3v/+9ysvLdfToUQ0ZMkSTJ0+WJC1YsEC7du1SWVmZdu/erd69e2v8+PF65JFHlJeXpxEjRmjKlCnN8bXQhFx11H/jexz2//zXdz+T5NK/P2iv59LjVHqw8b8sAEhLs96TbGnj+o56/umeKi0JOe6aYUm5+mpjBxXkhQUgQrR0tAKbqWL1wQcfqLS0VKtWrdIbb7yhOXPmSJK2bt2qMWPG6NVXX1Vqaqo/eQoPD9eSJUu0YsUKvfbaa/ryyy/13nvv+ef76quv9MQTT+j//u//tGvXLj3++ON69tln9cYbb+i1117T7t27m+NroQl9vauNSopa6Ve/zZM72KefDi7RhQNKFdrGp9KDrfTf112ocUMu0YTRFyqsnaXJT2wPdMiAY5WWtNLEmy7VLUlXauJNg9QmzNJ9c7844bVDf5Gnt7O6NHOEgHM0S2J1/vnna9euXXrggQf05ptvKiSk5l9BcXFxGjBggCRp9OjR2rZtm8rKymRZlubPn69Ro0bp+uuv1/bt27Vlyxb/fIMHD9ZZZ50lt9utnj176rLLLlNISIjCwsLUtWtX7d27tzm+FpqQVR2kOXf11ICfHdTfPvpM19+Wp/ff7KgD+SE6ctit7V+2k89yqaQoRItnd9UlV3yjsHbVgQ4bcKQjFcHa8Z/28llBKikO1ZL5CbpkUJHatK39d6pX34Pq0LFSH6yNDlCkaPHsJjgcplkSq9jYWK1atUqXX365PvroI40ePVpHjx6t8/qlS5eqtLRUy5cvV2ZmpoYPH17r+tDQUP/Pbrf7uNeWdfyiTDjP7q1tNTnlAiX3768/3NJLMbFHtHXTWcdd58AWPNCiHfs79cOW/LCkXP1rXbSOVHDfE1CXZkms8vPz5Xa7NXz4cE2bNk3FxcX65ptvtGfPHm3YsEGSlJmZqR49eqhdu3Y6dOiQOnXqpNDQUBUUFOjtt99ujjDRwsT3LFerEJ9CW1v65W15iuxUpbUrOqnnRYfUpWuFXC5bZ0VU6c6Zu7Xp43AdLuM/9kB9gtw+tQqxFOS2FRRkf/uzTz0vKFGXuPKav1PtK3XH/Vv0xYYOOlz23brFkFBLg4cXaG0Wi9ZRt2NrrEweTtMsv4m2bt2qxx9/XJLk8/k0fvx4RUVFKSEhQVlZWZo3b56CgoI0f/58SdLNN9+siRMn6rrrrlNMTIwGDRrUHGGihRl23X4l3lio4GBbX24I1/TfJKiqMkgxsUf1m3u3K6JjlQ6XufXvD9srfdJ5gQ4XaPF+fdsupd6x0/966C+8ynimm3L3tNXYu79QRGSlDpe7tfHjszV/+kW13jvoqkIdLgvWF5+yzQLqY/pOvsbNtXDhQi1YsMBftMnJydHUqVNVUlKiiIgIpaenKz4+XpLqHWsMl20HppGyfv16paena8WKFYH4+FrG9rpXBXsPBDoM4LQU1D480CEAp7WocyL11/UPBDoMSdIN459RfmGpsfliosK1/M93nNJ7vvrqKz355JPauXOnnnnmGfXo0UNjx47VL3/5S40ePVqvv/66XnnlFb3wwguSVO9YY7CPFQAAMKKpWoFer1f79u2rdZSWHp/AVVZWas6cOZo1a5ZcrpoNpIuKipSdna2kpCRJUlJSkrKzs1VcXFzvWGMFbFHKwIEDW0S1CgAAtGypqanKzc2tdS4tLU0TJkyode7pp5/WqFGjFBsb6z/n9XoVHR0tt7vmkU1ut1tRUVHyer2ybbvOscjIxrW9We0LAADMML1FwrdzZWRkHHfHf3h47WUGGzdu1ObNm3XfffcZDODUkVgBAIAWzePxnPSaTz/9VLt27dKwYcMk1exIcNttt2natGkqKCiQZVn+LZkKCwvl8Xhk23adY43FGisAAGBEILdbGD9+vD744AOtW7dO69atU0xMjJ577jmNHDnSvwuBJGVlZSkhIUGRkZHq2LFjnWONRcUKAACY4bNrDpPzGTB79mxNnTpVixcvVnh4uNLT0xs01hgkVgAA4LSzbt06/8/dunXT8uXLT3hdfWONQWIFAADMaKLF607CGisAAABDqFgBAAAjTD/fj2cFAgCAM1jLeFZgINEKBAAAMISKFQAAMMNwK9CBBSsqVgAAAKZQsQIAAGaw3QKJFQAAMKPmrkBz2ZAT7wqkFQgAAGAIFSsAAGCG79vD5HwOQ8UKAADAECpWAADACJdtG15j5bxFVlSsAAAADKFiBQAAzGC7BRIrAABgCs8KpBUIAABgCBUrAABghMvwswLZIBQAAOAMRsUKAACYYRteY+XA7RZIrAAAgBEuX81hcj6noRUIAABgCBUrAABgBq1AKlYAAACmULECAABmsPM6iRUAADCDhzDTCgQAADCGihUAADCEZwVSsQIAADCEihUAADDD9+1hcj6HoWIFAABgCBUrAABgBHcFklgBAABTbBneed3cVM2FViAAAIAhVKwAAIAZPCuQihUAAIApVKwAAIAZbLdAYgUAAMxwyfBdgQ5cvU4rEAAAwBAqVgAAwAwWr1OxAgAAMIWKFQAAMIOKFYkVAAAwhMSKViAAAIApVKwAAIAZ7GNFxQoAAMAUKlYAAMAM2+wGoayxAgAAOINRsQIAAGZwVyCJFQAAMMSW5DOZWDX80oMHD2ry5Mnau3evQkJCFBcXpzlz5igyMlI5OTmaOnWqSkpKFBERofT0dMXHx0tSvWONQSsQAAA4nsvl0u23367Vq1crMzNTsbGxeuyxxyRJs2bNUkpKilavXq2UlBTNnDnT/776xhqDxAoAAJhxrBVo8pDk9Xq1b9++WkdpaWmtj46IiNDAgQP9r/v27au8vDwVFRUpOztbSUlJkqSkpCRlZ2eruLi43rHGohUIAABatNTUVOXm5tY6l5aWpgkTJpzwep/Pp2XLlmno0KHyer2Kjo6W2+2WJLndbkVFRcnr9cq27TrHIiMjGxUriRUAADCjiRavZ2RkyLKsWkPh4eF1vu3BBx9UWFiYbrrpJmVnZ5uLpwFIrAAAgBlNlFh5PJ4GvyU9PV179uzRkiVLFBQUJI/Ho4KCAlmWJbfbLcuyVFhYKI/HI9u26xxrLNZYAQCA08KTTz6pL7/8UosWLVJISIgkqWPHjkpISFBWVpYkKSsrSwkJCYqMjKx3rLGoWAEAADN8ttntFk5hru3bt2vJkiWKj4/Xr3/9a0nSOeeco0WLFmn27NmaOnWqFi9erPDwcKWnp/vfV99YY5BYAQAAxzvvvPO0devWE45169ZNy5cvP+WxxiCxAgAAZti+msPkfA5DYgUAAAwxvHj9VLZebyFYvA4AAGAIFSsAAGCGT4YXr5ubqrlQsQIAADCEihUAADCjiTYIdRIqVgAAAIZQsQIAAGZQsSKxAgAAhpBY0QoEAAAwhYoVAAAww+erOUzO5zBUrAAAAAyhYgUAAAzhkTYkVgAAwAwWr9MKBAAAMIWKFQAAMINnBVKxAgAAMIWKFQAAMMP2ybYNlplMztVMSKwAAIAZPttwK5DF6wAAAGcsKlYAAMAMtlugYgUAAGAKFSsAAGCGbfhZgQ5cvE7FCgAAwBAqVgAAwAzWWJFYAQAAM2yfLdtgK9BmuwUAAIAzFxUrAABgBq1AKlYAAACmULECAABm8EgbEisAAGCKz/DeU+xjBQAAcMaiYgUAAIyo2W7BXPuO7RYAAADOYFSsAACAGbZtdo2VA7dbILECAABG0AqkFQgAAGAMFStJZ3fuEOgQgNOW66x2gQ4BOK2d7YkIdAh+Z3fpYLQVeHYX5/1+dtm2AxuYAAAALRCtQAAAAENIrAAAAAwhsQIAADCExAoAAMAQEisAAABDSKwAAAAMIbECAAAwhMQKAADAEBIrAAAAQ0isAAAADCGxQsD17NlT5eXlgQ4DAIAfjcQKAADAkOBABwBI0osvvqi33npLJSUlmjx5shITEyVJ9957r3JyclRVVaWf/OQnmjdvntq3b6/169froYceUp8+fbRp0yYFBwdr/vz5WrhwobZv3y6Px6MFCxYoLCwswN8MCJyKigpNmTJFO3bsUHBwsLp27aqUlBQ99NBD6t27t7Zs2SK3261HHnlE3bt31/79+/X73/9e5eXlOnr0qIYMGaLJkydLkhYsWKBdu3aprKxMu3fvVu/evTV+/Hg98sgjysvL04gRIzRlypQAf2OgBbCBAOvRo4f94osv2rZt2xs2bLAHDx7sHysqKvL//MQTT9iPPvqobdu2/fHHH9u9evWys7Ozbdu27dmzZ9tXXHGF7fV6bdu27dtvv91++eWXm+srAC3SmjVr7HHjxvlfl5SU2B9//LHdo0cPe/369bZt2/aKFSvsMWPG2LZt20eOHLHLysps27btyspK++abb7bfffdd27Zt+09/+pM9YsQIu7S01K6urravvfZa+9Zbb7WPHj1ql5eX25deeqmdk5PTrN8PaImoWKFFGDlypCSpb9++Kiws1NGjRxUaGqrXX39dmZmZqqqq0uHDhxUfH+9/T9euXZWQkCBJ6tWrl/Ly8hQTEyNJ6t27t/bs2dPs3wNoSc4//3zt2rVLDzzwgAYMGKCrrrpKkhQXF6cBAwZIkkaPHq0//vGPKisrU1BQkObPn6+NGzfKtm0dOHBAW7Zs0ZVXXilJGjx4sM466yxJNWsjzz//fIWEhCgkJERdu3bV3r17a/0dBc5ErLFCixAaGipJcrvdkqTq6mpt2LBBy5Yt07PPPqvMzExNmjRJlZWV/veEhIT4f3a73f45jr22LKuZogdaptjYWK1atUqXX365PvroI40ePVpHjx6t8/qlS5eqtLRUy5cvV2ZmpoYPH17r+h/+HePvHHA8Eiu0WKWlpWrXrp0iIiJUWVmpV155JdAhAY6Sn58vt9ut4cOHa9q0aSouLtY333yjPXv2aMOGDZKkzMxM9ejRQ+3atdOhQ4fUqVMnhYaGqqCgQG+//XaAvwHgPLQC0WJdeeWVeuONN/Tzn/9c0dHRuuCCC7R58+ZAhwU4xtatW/X4449Lknw+n8aPH6+oqCglJCQoKytL8+bN87f/JOnmm2/WxIkTdd111ykmJkaDBg0KZPiAI7ls27YDHQQAoHmsX79e6enpWrFiRaBDAU5LtAIBAAAMoWIFAABgCBUrAAAAQ0isAAAADCGxAgAAMITECkAtU6dO1ZNPPilJ2rBhg/+5jU2tZ8+ede6Wf/PNN2v58uUNmmfo0KH617/+1agYfsx7AUAisQIcaejQoerTp48uvvhiXXbZZZo2bZrKy8uNf06/fv20evXqk163YsUK/dd//ZfxzwcApyGxAhxqyZIl2rhxo1599VVt3rxZ//M//3PcNdXV1QGIDADOXCRWgMNFR0friiuu0Pbt2yXVtNQyMjJ09dVX6+qrr5YkvfPOOxo9erT69eunX//619qyZYv//dnZ2RozZowuvvhiTZo0qdaz4davX+9/AK8keb1epaWl6dJLL9XAgQM1Z84c7dy5U7NmzdLnn3+uiy++WP369ZMkVVZWKj09XVdddZUuu+wyzZw5U0eOHPHP9eyzz2rw4MEaPHiw/vGPfzT4++7du1djx47VwIEDNXDgQN17770qLS2tdc3mzZs1cuRI9e/fX9OmTav1ner7swCAH4vECnA4r9er9957TwkJCf5za9eu1csvv6xVq1bpq6++0vTp0zVnzhytX79eycnJ+t3vfqfKykpVVlbq7rvv1ujRo/XJJ5/ommuu0Zo1a074OZZl6Y477lDnzp21bt06vffeexo5cqS6deumBx54QH379tXGjRv9z6B79NFHlZOTo9dee01r1qxRYWGhFi1aJEl677339Pzzz+v555/XmjVr9NFHHzX4+9q2rTvuuEPvv/++3nzzTeXn52vBggW1rsnMzNRzzz2nt956Szk5OVq8eLEk1ftnAQAmkFgBDnX33XerX79+SklJUf/+/XXnnXf6x8aPH6+IiAi1bt1aL7/8spKTk3XRRRfJ7XZrzJgxatWqlT7//HNt2rRJVVVVGjdunFq1aqVrrrlGF1544Qk/74svvlBhYaEmT56ssLAwhYaG+qtTP2TbtpYvX67p06crIiJC7dq10x133KGVK1dKkt58801df/316tGjh8LCwpSWltbg7x0XF6fLL79cISEhioyM1C233KJPP/201jWpqanyeDyKiIjQXXfd5f/c+v4sAMAEHsIMONSiRYt02WWXnXDM4/H4f87Ly9Nrr72ml156yX+uqqpKhYWFcrlcio6Olsvl8o917tz5hHN6vV517txZwcEn/89GcXGxKioqdP311/vP2bYtn88nSSosLNQFF1zgH+vSpctJ5zymqKhIc+fO1YYNG1ReXi7bthUeHl7rmu9//86dO6uwsFBS/X8WAGACiRVwGvp+ouTxeHTnnXfqrrvuOu66Tz75RAUFBbJt2/+evLw8xcbGHnetx+OR1+tVdXX1ccnV9z9Pkjp06KDWrVtr5cqVio6OPm6uqKgoeb1e/+u8vLwGf7fHH39cLpdLb7zxhjp06KC1a9dqzpw5ta754dxRUVH+71DXnwUAmEArEDjN3XDDDfr73/+uTZs2ybZtHT58WP/85z9VVlamvn37Kjg4WC+88IKqq6u1Zs0abd68+YTz9OnTR506ddLjjz+uw4cP6+jRo/rss88kSR07dlRBQYF/rVJQUJBuuOEGzZs3T0VFRZKkgoICvf/++5Kka665Rq+++qp27NihiooKLVy4sMHfp7y8XGFhYQoPD1dBQYGeffbZ467529/+pvz8fJWUlOiZZ57RyJEjT/pnAQAmkFgBp7kLL7xQDz74oObMmaP+/fvr6quv1ooVKyRJISEhWrBggV599VX1799fq1at0ogRI044j9vt1pIlS7Rnzx797Gc/05VXXqk333xTknTppZeqe/fuGjx4sAYOHChJuv/++xUXF6cbb7xRP/3pT/Wb3/xGOTk5kqQhQ4Zo3LhxGjdunEaMGKFLL720wd8nLS1N2dnZ6tevn8aPH++/8/H7kpKSdOutt2r48OGKjY31V6jq+7MAABNctm3bgQ4CAADgdEDFCgAAwBASKwAAAENIrAAAAAwhsQIAADCExAoAAMAQEisAAABDSKwAAAAMIbECAAAw5P8BWah2ITGTjrIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************LinearDiscriminantAnalysis******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       0.97      0.99      0.98      1587\n",
      "        spam       0.95      0.83      0.88       252\n",
      "\n",
      "    accuracy                           0.97      1839\n",
      "   macro avg       0.96      0.91      0.93      1839\n",
      "weighted avg       0.97      0.97      0.97      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "****************** LinearDiscriminantAnalysis ROC Curve ******************\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Final testing. Testing against the test set\n",
    "#plotting our ROC curve\n",
    "\n",
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import plot_confusion_matrix, accuracy_score, roc_curve, auc\n",
    "\n",
    "for classifier in bestClassifiers:\n",
    "    print(\"******************\" + type(classifier).__name__ + \"******************\")\n",
    "    yPred = classifier.predict(X_test) #actual predicted values\n",
    "    print (classification_report(y_test, yPred, target_names = [\"ham\", \"spam\"]))  #comparing what I predicted with what it actually is\n",
    "    plot_confusion_matrix(classifier,X_test,y_test)\n",
    "    plt.grid()\n",
    "    plt.show()\n",
    " \n",
    "\n",
    "    if type(classifier).__name__ in [\"LogisticRegression\", \"LinearDiscriminantAnalysis\"]:\n",
    "        \n",
    "        yProb = [yprob[0] for yprob in classifier.predict_proba(X_test)]\n",
    "        \n",
    "        #print(yProb[:10])\n",
    "        print(\"****************** \" + type(classifier).__name__ + \" ROC Curve ******************\")\n",
    "        fp, tp, _ = roc_curve(y_test, yProb, pos_label = \"ham\")\n",
    "        aucNum = auc(fp,tp)\n",
    "        plt.plot(fp,tp, label= 'Ham = ' + str(aucNum));\n",
    "        plt.title(type(classifier).__name__ + \" ROC Curve\")\n",
    "        plt.xlabel('False Positive Rate')\n",
    "        plt.ylabel('True Positive Rate')\n",
    "        plt.legend();\n",
    "        plt.show();\n",
    "\n",
    "#Precision: How often is it true when we call it true? false-negatives\n",
    "#recall: How many did we detect to be true? Did we actually detect it? Ignores false-positives.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Within the classification report here is what I find most important. I want to maximize the true hams coming through at the expense of potentially letting some spam messages sneak through. This way I ensure that the legitimate messages always get through. Bias to everything being ham unless proven otherwise.  A spam getting through is bad but not the end of the world, but a ham message should always get through."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************KNeighborsClassifier******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       1.00      0.90      0.94      1587\n",
      "        spam       0.60      0.98      0.74       252\n",
      "\n",
      "    accuracy                           0.91      1839\n",
      "   macro avg       0.80      0.94      0.84      1839\n",
      "weighted avg       0.94      0.91      0.92      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************LogisticRegression******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       0.99      0.97      0.98      1587\n",
      "        spam       0.82      0.95      0.88       252\n",
      "\n",
      "    accuracy                           0.97      1839\n",
      "   macro avg       0.91      0.96      0.93      1839\n",
      "weighted avg       0.97      0.97      0.97      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************DecisionTreeClassifier******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       1.00      0.99      0.99      1587\n",
      "        spam       0.92      1.00      0.96       252\n",
      "\n",
      "    accuracy                           0.99      1839\n",
      "   macro avg       0.96      0.99      0.98      1839\n",
      "weighted avg       0.99      0.99      0.99      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************RandomForestClassifier******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       1.00      0.99      1.00      1587\n",
      "        spam       0.95      1.00      0.97       252\n",
      "\n",
      "    accuracy                           0.99      1839\n",
      "   macro avg       0.98      0.99      0.98      1839\n",
      "weighted avg       0.99      0.99      0.99      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************SVC******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       0.98      0.83      0.90      1587\n",
      "        spam       0.46      0.91      0.61       252\n",
      "\n",
      "    accuracy                           0.84      1839\n",
      "   macro avg       0.72      0.87      0.75      1839\n",
      "weighted avg       0.91      0.84      0.86      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************LinearDiscriminantAnalysis******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       0.99      0.97      0.98      1587\n",
      "        spam       0.83      0.93      0.88       252\n",
      "\n",
      "    accuracy                           0.97      1839\n",
      "   macro avg       0.91      0.95      0.93      1839\n",
      "weighted avg       0.97      0.97      0.97      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for classifier in bestClassifiersUpsampled:\n",
    "    print(\"******************\" + type(classifier).__name__ + \"******************\")\n",
    "    yPred = classifier.predict(X_test) #actual predicted values\n",
    "    print (classification_report(y_test, yPred, target_names = [\"ham\", \"spam\"]))  #comparing what I predicted with what it actually is\n",
    "    plot_confusion_matrix(classifier,X_test,y_test)\n",
    "    plt.grid()\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In conclusion, If I were responsible for creating a machine learning model that decides whether a text message is ham or spam, the model I would choose is Random Forest. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************RandomForestClassifier******************\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "         ham       1.00      0.99      1.00      1587\n",
      "        spam       0.95      1.00      0.97       252\n",
      "\n",
      "    accuracy                           0.99      1839\n",
      "   macro avg       0.98      0.99      0.98      1839\n",
      "weighted avg       0.99      0.99      0.99      1839\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#I have shown that for this type of classification, the Random Forest model performs the best. \n",
    "\n",
    "for classifier in bestClassifiersUpsampled:\n",
    "    if type(classifier).__name__ in [\"RandomForestClassifier\"]:\n",
    "\n",
    "        print(\"******************\" + type(classifier).__name__ + \"******************\")\n",
    "        yPred = classifier.predict(X_test) #actual predicted values\n",
    "        print (classification_report(y_test, yPred, target_names = [\"ham\", \"spam\"]))  #comparing what I predicted with what it actually is\n",
    "        plot_confusion_matrix(classifier,X_test,y_test)\n",
    "        plt.grid()\n",
    "        plt.show()\n"
   ]
  },
  {
   "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": "3.8.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}