PPRL_Multiclass_Block_FirstORlastname_2022.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "40245fc6",
   "metadata": {},
   "source": [
    "## Record-pair Linkage Predictive Modeling Program\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eec8156e",
   "metadata": {},
   "source": [
    "### Author: CDC's Data Science Upskilling (DSU) Team named, \"MYABC 4 PPRL\"\n",
    "### Membership: DSU 2021-2022 (Cohort 3)\n",
    "#### Team Project: Privacy Preserving Record Linkage (PPRL)\n",
    "\n",
    "\n",
    "Python Program Description: \n",
    "This NoteBook contains a record-pair matching predictive modeling Python program. It analyzes a Bloom Filter (field level)-encoded comparison output where the blocking variable was \"firstORlastname\". Separate NoteBooks on Blocking and other PPRL previous steps are included in the project folder.Multiclass classification, a supervised machine learning (SML), was carried out with three algorithms for predicting whether each record-pair matches, doesn't match, may/maynot match. The algorithms are logistic regression, OneVersusRest (Estimator= Random Forest), and Multinomial Naive Bayes).\n",
    "\n",
    "Dataset: Synthetic\n",
    "PPRL Data Source: Febrl (Freely Extensible Biomedical Record Linkage)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70cd6780",
   "metadata": {},
   "source": [
    "## CAUTION!\n",
    "Please run the file named, 'PPRL_with_Field-level_bloom_filters_using_double_hashing' before running this program as the output of that file is the input data to be loaded in this file"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d29ba34d",
   "metadata": {},
   "source": [
    "##### In the next cell is a graphical illustration of the modeling pipeline"
   ]
  },
  {
   "attachments": {
    "DSU_PPRL_ML_Pipeline_2022.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "f61e3deb",
   "metadata": {},
   "source": [
    "![DSU_PPRL_ML_Pipeline_2022.png](attachment:DSU_PPRL_ML_Pipeline_2022.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "df91cba6",
   "metadata": {},
   "source": [
    "### Import the necessary packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e8e0ae37",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import csv\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt  # works with seaborn etc to display plots created by seaborn etc\n",
    "from matplotlib import style # to make the output of matplotlib pretier (all charts then turned red)\n",
    "style.use(\"ggplot\")# used because I prefer blue color for my charts\n",
    "plt.rcParams[\"font.family\"] = \"Times New Roman\"\n",
    "\n",
    "import seaborn as sns\n",
    "\n",
    "import plotly.express as plotExp\n",
    "import plotly.figure_factory as factory_fig\n",
    "import plotly.graph_objs as graphObj\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler # would handle outliers (data standardization)\n",
    "from sklearn.preprocessing import LabelEncoder # for transforming Categorical data to Numerical data\n",
    "\n",
    "#import time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e4c9d366",
   "metadata": {},
   "outputs": [],
   "source": [
    "## So we could display our plots in a cell\n",
    "\n",
    "%matplotlib inline\n",
    "from pylab import rcParams\n",
    "rcParams ['figure.figsize']=10,8"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1e856d1d",
   "metadata": {},
   "source": [
    "***** To be able to print multiple output from a cell, we'll let all output come through from Ipython and we'll utilize Jupyter's interactive shell"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "75173759",
   "metadata": {},
   "outputs": [],
   "source": [
    "from IPython.core.interactiveshell import InteractiveShell\n",
    "InteractiveShell.ast_node_interactivity=\"all\""
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9bff85d6",
   "metadata": {},
   "source": [
    "#### Modify some of the Pandas printing capabilities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c70b6694",
   "metadata": {},
   "outputs": [],
   "source": [
    "# By default, Pandas will display only 10 rows/cols\n",
    "\n",
    "# Globally set Max rows displayed in output to 20\n",
    "pd.set_option('display.max_rows', 30)\n",
    "\n",
    "# Globally set display for Pandas to always display all the columns\n",
    "#pd.set_option('display.line_width', 5000)\n",
    "pd.set_option('display.max_columns', None)# Settig to none makes display unlimited, better than putting a number\n",
    "\n",
    "# Globally set Max rows displayed in output to 20\n",
    "pd.set_option('display.max_rows', None)  # Great! No deprecation, many rows printed as wanted\n",
    "\n",
    "# Globally set decimal places Displayed in output to any number of choice, change it prn\n",
    "pd.set_option('display.precision', 2)\n",
    "\n",
    "# Globally set floatto avoid getting such values in\n",
    "#exponent notation, in correlation especially (Worked)\n",
    "pd.options.display.float_format = '{:20,.4f}'.format \n",
    "\n",
    "# Globally set max_colwidth display\n",
    "pd.options.display.max_colwidth = None # This removes any limit to num of col to be shown\n",
    "\n",
    "# Globally set display expand_frame_repr\n",
    "pd.set_option('display.expand_frame_repr', False)\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49c4c3fb",
   "metadata": {},
   "source": [
    "#### Read the pickle file containing the comparison output to memory "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e40c4620",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2=pd.read_pickle('comparison_output_fln.pkl')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "366f64a9",
   "metadata": {},
   "source": [
    "### Exploratory Data Analysis "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ec95ea11",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.info()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9ccbd531",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.head(5)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "46dcb84a",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb81b5c2",
   "metadata": {},
   "source": [
    "#### Rename columns as needed, using Dictionary data structure"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "db3a6f0b",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.rename(columns={'dob': 'dateOfBirth', 'firstname':'firstName','lasttname':'lastName'}, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "16b23217",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9c654c07",
   "metadata": {},
   "source": [
    "#### Plot data distribution to see if data follows Gaussian distribution (many ML algorithms will not do well if data is not normally distributed)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d2de421e",
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.figure(1, figsize=(15,6))\n",
    "plt.subplots_adjust(hspace=0.5, wspace=0.5)\n",
    "sns.distplot(pprl_DF2, bins=10)\n",
    "plt.title('PPRL Data Distribution')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "013f45e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "#### Distribution plot shows the data is not normally distributed"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "62103033",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "43c32f14",
   "metadata": {},
   "outputs": [],
   "source": [
    "####Describe the data and Transpose result if you want\n",
    "pprl_DF2.describe().T"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "08d29d19",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DataDescrp= pprl_DF2.describe().T # Remember describe only works on numerical data\n",
    "\n",
    "# You may also save the output for reporting purposes\n",
    "#pprl_DataDescrp.to_csv('Report: PPRL_DataDescription.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e025fed",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.head()\n",
    "pprl_DF2.isna().sum()\n",
    "pprl_DF2.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e66d5b0f",
   "metadata": {},
   "source": [
    "##### Install Pandas profiling for auto-descriptive statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "708628f6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#!pip install  https://github.com/pandas-profiling/pandas-profiling/archive/master.zip \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7f840d95",
   "metadata": {},
   "outputs": [],
   "source": [
    "from pandas_profiling import ProfileReport\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f4897a78",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Obtain descriptive statistics of the data using Pandas Profiling package from Git"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0588b111",
   "metadata": {},
   "outputs": [],
   "source": [
    "summaryReport= ProfileReport(pprl_DF2,title='PPRL_UnlabelledData_DescriptiveReport', html={'style':{'full_width':'True'}})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c7698db7",
   "metadata": {},
   "outputs": [],
   "source": [
    "summaryReport"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "400bf1b9",
   "metadata": {},
   "source": [
    "### Multiclass Classification Target Engineering"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "42bdb588",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Remove the binary target. \n",
    "# We won't remove the blocking variable here, since their values are not monotonous \n",
    "#(based on the logic OR used)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9371b09f",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.drop(\"true_links\",axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30438cad",
   "metadata": {},
   "source": [
    "### Compute Df rowwise mean, and round to 1 decimal place to permit float comparison down the road"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1a1d8b18",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2['row_Means'] =pprl_DF2.mean(axis=1).round(1) # Round to permit comparison, based on exp with working wt mock data\n",
    "pprl_DF2['row_Means'].describe()\n",
    "pprl_DF2['row_Means'].head()\n",
    "print('============================')\n",
    "pprl_DF2.info()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30f5b42e",
   "metadata": {},
   "source": [
    "#### Bin the row means and categorize them as matching classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cd6591d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "# #### View distinct values in matched DF and their frequencies, and sort them by frequency\n",
    "\n",
    "pprl_DF2['row_Means'].value_counts().sort_values()\n",
    "\n",
    "pprl_DF2['matched']=pprl_DF2['row_Means']\n",
    "\n",
    "\n",
    "pprl_DF2['matched']=pd.cut(pprl_DF2.matched, bins=[\n",
    "    -0.001, 0.45, 0.75,2.0], labels=['no','mayBe','yes'])# small, medium, large\n",
    "\n",
    "# ##### Check class distribution of the target\n",
    "pprl_DF2['matched'].value_counts()\n",
    "\n",
    "##### Plot a bar chart to display the class distribution\n",
    "pprl_DF2['matched'].value_counts().plot(kind='bar', color='b', \n",
    "                                        title=\"Matching Distribution Plot by Classes\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5491029a",
   "metadata": {},
   "source": [
    "****Both the textual and vizualized distribution results show there is class imbalance in the dataset. We'll address it as we move forward"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a9ccf534",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b6ef4269",
   "metadata": {},
   "source": [
    "#### Get some barplots with sns and add  values to the bars"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "76bf78e3",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "ax = sns.barplot(x='matched', y='dateOfBirth',\n",
    "                 data=pprl_DF2,\n",
    "                 errwidth=0)\n",
    " \n",
    "# now assign the bar values to each bar by passing containers method to bar_label function\n",
    "ax.bar_label(ax.containers[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ba88373f",
   "metadata": {},
   "outputs": [],
   "source": [
    "ax = sns.barplot(x='matched', y='ssn',\n",
    "                 data=pprl_DF2,\n",
    "                 errwidth=0)\n",
    " \n",
    "# now assign the bar values to each bar by passing containers method to bar_label function\n",
    "ax.bar_label(ax.containers[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0be3023d",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "ax = sns.barplot(x='matched', y='lastName',\n",
    "                 data=pprl_DF2,\n",
    "                 errwidth=0)\n",
    " \n",
    "# now assign the bar values to each bar by passing containers method to bar_label function\n",
    "\n",
    "ax.bar_label(ax.containers[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5778b7cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "ax = sns.barplot(x='matched', y='firstName',\n",
    "                 data=pprl_DF2,\n",
    "                 errwidth=0)\n",
    " \n",
    "# now assign the bar values to each bar by passing containers method to bar_label function\n",
    "\n",
    "ax.bar_label(ax.containers[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e0b9cf9b",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.info()\n",
    "pprl_DF2.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e6095451",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.columns"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c343ac6e",
   "metadata": {},
   "source": [
    "### Exploratory Data Analysis completed "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb353505",
   "metadata": {},
   "source": [
    "#### Start modeling-related operations\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d25385ee",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.drop('row_Means', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "56595a8b",
   "metadata": {},
   "source": [
    "##### Create feature vectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "81ec6e40",
   "metadata": {},
   "outputs": [],
   "source": [
    "x =pprl_DF2.drop(\"matched\", axis=1)\n",
    "y =pprl_DF2.matched"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6342c38c",
   "metadata": {},
   "source": [
    "###### Transform categorical data into numerical data\n",
    "only the target is categorical in this case"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "395497f5",
   "metadata": {},
   "outputs": [],
   "source": [
    "label_encod=LabelEncoder()\n",
    "label_encod.fit(y)\n",
    "y=label_encod.transform(y)\n",
    "classes=label_encod.classes_\n",
    "classes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "51119cd9",
   "metadata": {},
   "outputs": [],
   "source": [
    "x.shape\n",
    "y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0e9b56d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "##### Import the modelling-related classes\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7ef1e42d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#!pip install imblearn # Required before importing the imblearn pipeline and SMOTE \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5d18c419",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "import sklearn\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.pipeline import Pipeline\n",
    "from numpy.random import RandomState\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.naive_bayes import MultinomialNB\n",
    "from sklearn import svm\n",
    "\n",
    "from sklearn.multiclass import OneVsRestClassifier\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import confusion_matrix, classification_report # useful for evaluating our models, \n",
    "\n",
    "from sklearn.metrics import average_precision_score\n",
    "from sklearn import metrics\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score, f1_score\n",
    "from sklearn.model_selection import cross_validate, GridSearchCV, cross_val_score\n",
    "\n",
    "\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from imblearn.pipeline import Pipeline as imbaPipe # The SMOTE class is in the imblearn package.\n",
    "from imblearn.pipeline import make_pipeline # For gridsearching SMOTE inside Grid\n",
    "\n",
    "\n",
    "from imblearn.over_sampling import SMOTE # SMOTE= synthetic minority oversampling technique\n",
    "from imblearn.metrics import classification_report_imbalanced\n",
    "\n",
    "from sklearn.metrics import recall_score, f1_score, precision_score,roc_curve\n",
    "from sklearn.metrics import roc_auc_score, precision_score, recall_score\n",
    "from sklearn.metrics import precision_recall_fscore_support # For summary table\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab58f883",
   "metadata": {},
   "source": [
    "### Split the dataset (Create train-test data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "240008e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Split Data into training and test (you can split as desired, but 70:30% was used here) \n",
    "\n",
    "x_train, x_test, y_train, y_test = train_test_split(x, y,test_size=0.3,random_state=42)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9a54abd3",
   "metadata": {},
   "outputs": [],
   "source": [
    "##### Check data shape after spliting, before scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "33c373a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Confirm the Output Dimensions\n",
    "# We can confirm the dimensions of the data are the same within test and train\n",
    "# The proportion should also be close to the test_size argument.\n",
    "\n",
    "\n",
    "x_train.shape\n",
    "x_test.shape\n",
    "y_train.shape\n",
    "y_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "391f89f9",
   "metadata": {},
   "source": [
    "#### Count the values in each data split- use the np.unique method\n",
    "**** Remember, these are no longer in DF format but ndarray, so, value_counts() can't work "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "76381aa1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import collections\n",
    "collections.Counter(y_test)\n",
    "\n",
    "## We could obtain the same info using numpy\n",
    "\n",
    "print('y_test value frequency=========')\n",
    "\n",
    "unique, counts = np.unique(y_test, return_counts=True)\n",
    "dict(zip(unique, counts)) \n",
    "\n",
    "print('x_test value frequency=========')\n",
    "\n",
    "unique, counts = np.unique(x_test, return_counts=True)\n",
    "dict(zip(unique, counts))\n",
    "\n",
    "\n",
    "print('y_train value frequency=========')\n",
    "unique, counts = np.unique(y_train, return_counts=True)\n",
    "dict(zip(unique, counts))\n",
    "\n",
    "print('x_train value frequency=========')\n",
    "\n",
    "unique, counts = np.unique(x_train, return_counts=True)\n",
    "dict(zip(unique, counts))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "381ba06d",
   "metadata": {},
   "source": [
    "#### Now scale the training and testing features separately to avoid data leakage"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d93585c6",
   "metadata": {},
   "outputs": [],
   "source": [
    "scaler = MinMaxScaler() #)# to normalize the values so they could be between 0 and 1.\n",
    "\n",
    "# TO AVOID DATA LEAKAGE, SCALE EACH PART OF THE TRAINING DATA SEPARATELY\n",
    "x_train=scaler.fit_transform(x_train)\n",
    "x_test=scaler.transform(x_test)\n",
    "\n",
    "##### It is expected that scaling put all the trainign data in a range of 0 and 1 by now-- yes, achieved!\n",
    "\n",
    "x_train[:5]# Show me 5 records in the scaled Xtrain\n",
    "x_test[:5] # Show me5 records in the scaled Xtest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f6113a42",
   "metadata": {},
   "outputs": [],
   "source": [
    "## Import the necessary modeling packages for multiclass classification"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c8518a2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn import pipeline# import the pipeline class\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import preprocessing\n",
    "\n",
    "from imblearn.pipeline import Pipeline as imbaPipe\n",
    "\n",
    "from imblearn.pipeline import make_pipeline, Pipeline\n",
    "from imblearn.over_sampling import SMOTE\n",
    "from sklearn.preprocessing import label_binarize\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.metrics import recall_score, f1_score, accuracy_score\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.multiclass import OneVsRestClassifier # OVR\n",
    "from sklearn.multiclass import OneVsOneClassifier # OVO\n",
    "from sklearn import svm # Required with OVR\n",
    "from sklearn.linear_model import LogisticRegression# its multinomial parameter is ovr by default\n",
    "from sklearn.naive_bayes import MultinomialNB# its multinomial parameter is ovr by default\n",
    "\n",
    "from sklearn import svm\n",
    "\n",
    "\n",
    "# Import the Pipeline API\n",
    "\n",
    "from sklearn.pipeline import Pipeline\n",
    "# Import a data standaizer of your choice\n",
    "\n",
    "from sklearn.preprocessing import MinMaxScaler  # To handle outliers etc\n",
    "from sklearn.preprocessing import LabelEncoder,     OneHotEncoder, MinMaxScaler, StandardScaler # aka ONEHOT: for transforming Cat features (X) to Numerical\n",
    "\n",
    "# Import the LogReg\n",
    "\n",
    "# Import the Cross validation class\n",
    "import sklearn\n",
    "\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.model_selection import cross_validate, GridSearchCV, cross_val_score\n",
    "\n",
    "\n",
    "from sklearn.metrics import precision_score\n",
    "from sklearn.metrics import accuracy_score, f1_score, classification_report\n",
    "\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score, f1_score\n",
    "\n",
    "from sklearn.metrics import roc_auc_score\n",
    "# import the needed module from scikit\n",
    "from sklearn.metrics import roc_curve, auc\n",
    "from sklearn.metrics import precision_recall_fscore_support # For summary table\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c42fdfa",
   "metadata": {},
   "source": [
    "#### Perform ML Experiments "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "80a7248e",
   "metadata": {},
   "source": [
    "#### Build multiple models\n",
    "***** Steps:\n",
    "\n",
    "1. Build baseline models\n",
    "3. Evaluate the models\n",
    "4. Embark on model performance improvement activities\n",
    "4. Build optimized models\n",
    "5. Evaluate the models\n",
    "6. Identify the most generalizable model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "838208ab",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "eb64dcc8",
   "metadata": {},
   "source": [
    "#### Instantiate the intending classifiers and their SK\n",
    "##### Create SKLearn pipeline per classifier- for Baseline Models (Model Set 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f17af629",
   "metadata": {},
   "outputs": [],
   "source": [
    "nbayesPipe_SK = Pipeline([(\"dataScaler\", scaler), (\"naive\", MultinomialNB())])\n",
    "\n",
    "logregPipe_SK = Pipeline([(\"dataScaler\", scaler), (\"LogReg\", LogisticRegression())]) \n",
    "\n",
    "\n",
    "# Any classifier can be used as an estimator, not only random forest\n",
    "ovr_ForestPipe_SK=Pipeline([(\"dataScaler\", scaler),\n",
    "                            ('ovrForest', OneVsRestClassifier(RandomForestClassifier()))])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c8d4d5f1",
   "metadata": {},
   "source": [
    "### Create Baseline Models \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6008206b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#### Create a result dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "93746866",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "metrics_DF_pipeSK = pd.DataFrame(index =['Accuracy','Precision','Recall','F1'], \n",
    "                          columns=['NBayes', 'OVR','LReg'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a407e90",
   "metadata": {},
   "source": [
    "#### Naive Bayes Baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "99b9a653",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "pipe1 = nbayesPipe_SK.steps[1][1]\n",
    "NBayesmodel= pipe1.fit(x_train, y_train)\n",
    "\n",
    "\n",
    "# #-========================.  =============.   ===========\n",
    "y_pred_testBayes= NBayesmodel.predict(x_test)\n",
    "Bayes_pos_probs =NBayesmodel.predict_proba(x_test)[:,2] # index 2 contains the predicted probabilities for \"yes match\"\n",
    "Bayes_pos_probs[10] # Show me the 11th of the yes probabilities\n",
    "\n",
    "metrics_DF_pipeSK.loc['Accuracy', 'NBayes']= accuracy_score(y_pred = y_pred_testBayes,\n",
    "                                                            y_true=y_test)\n",
    "metrics_DF_pipeSK.loc['Precision', 'NBayes']= precision_score(y_pred = y_pred_testBayes,\n",
    "                                                              y_true=y_test,average='weighted')\n",
    "metrics_DF_pipeSK.loc['Recall', 'NBayes']= recall_score(y_pred = y_pred_testBayes,\n",
    "                                                        y_true=y_test,average='weighted')\n",
    "metrics_DF_pipeSK.loc['F1', 'NBayes']= f1_score(y_pred = y_pred_testBayes,\n",
    "                                                y_true=y_test,average='weighted')\n",
    "# classification_report\n",
    "train_y_predBayes= NBayesmodel.predict(x_train)\n",
    "print('Bayes train :', classification_report(y_train, train_y_predBayes),flush=True)\n",
    "print('Bayes test :', classification_report(y_test, y_pred_testBayes),flush=True)\n",
    "\n",
    "# Get Matrix\n",
    "\n",
    "sns.set(rc = {'figure.figsize': (15, 8)})\n",
    "conf_matrix = confusion_matrix(y_pred_testBayes, y_test)\n",
    "np.set_printoptions(precision = 2)\n",
    "print(\"Print Bayes's crosstab (Conf matrix) :\" '\\n',\"===================\",'\\n',conf_matrix)# Matrix Table, not the graphic one\n",
    "\n",
    "# Now Plot the Conf Matrix\n",
    "\n",
    "#sns.heatmap(conf_matrix, annot=True, fmt='d')\n",
    "\n",
    "group_names = ['T MayB','F MayB','F MayB',\n",
    "               'F No (F-Neg : T2Error)','T No (T-Neg)','F No (F-Neg : T2Error)',\n",
    "               'F Yes (F-Pos : T1Error)','F Yes (F-Pos : T1Error)','T Yes (T-Pos)']\n",
    "\n",
    "\n",
    "group_counts = ['{0:0.0f}'.format(value) for value in conf_matrix.flatten()]\n",
    "group_percentages = ['{0:.2%}'.format(value) for value in \n",
    "                                         conf_matrix.flatten()/np.sum(conf_matrix)]\n",
    "labels = [f'{v1}\\n{v2}\\n{v3}' for v1, v2, v3 in\n",
    "          zip(group_names,group_counts,group_percentages)]\n",
    "\n",
    "labels = np.asarray(labels).reshape(3,3)\n",
    "\n",
    "\n",
    "sns.heatmap(conf_matrix, annot=labels, fmt='', cmap=\"Blues\")# other colors::cmap=\"OrRd\", \"OrRd_r\" colors\n",
    "\n",
    "\n",
    "plt.xlabel('Actual Label')\n",
    "plt.ylabel('Predicted Label')\n",
    "plt.title(\"CMatrix for NBayes Baseline: PPRL, Multiclass [FirstORlastname]\")\n",
    "\n",
    "#plt.savefig('')\n",
    "plt.show() \n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "73f6bd7a",
   "metadata": {},
   "source": [
    "#### OVR- Baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "018f45f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "pipe2 = ovr_ForestPipe_SK.steps[1][1]\n",
    "ovrmodel= pipe2.fit(x_train, y_train)\n",
    "\n",
    "# #-========================.  =============.   ===========\n",
    "Y_pred_testOVR= ovrmodel.predict(x_test)\n",
    "OVR_pos_probs =ovrmodel.predict_proba(x_test)[:,2]# this is  y_score in scikit\n",
    "\n",
    "\n",
    "metrics_DF_pipeSK.loc['Accuracy', 'OVR']= accuracy_score(y_pred = Y_pred_testOVR,\n",
    "                                                            y_true=y_test)\n",
    "metrics_DF_pipeSK.loc['Precision', 'OVR']= precision_score(y_pred = Y_pred_testOVR,\n",
    "                                                              y_true=y_test, average='weighted')\n",
    "metrics_DF_pipeSK.loc['Recall', 'OVR']= recall_score(y_pred = Y_pred_testOVR,\n",
    "                                                        y_true=y_test,average='weighted')\n",
    "metrics_DF_pipeSK.loc['F1', 'OVR']= f1_score(y_pred = Y_pred_testOVR,\n",
    "                                                y_true=y_test,average='weighted')\n",
    "\n",
    "# classification_report\n",
    "train_y_predOVR= ovrmodel.predict(x_train)\n",
    "print('OVR train :', classification_report(y_train, train_y_predOVR),flush=True)\n",
    "print('OVR test :', classification_report(y_test, Y_pred_testOVR),flush=True)\n",
    "\n",
    "# Get confusion matrix\n",
    "\n",
    "sns.set(rc = {'figure.figsize': (15, 8)})\n",
    "conf_matrix = confusion_matrix(Y_pred_testOVR, y_test)\n",
    "np.set_printoptions(precision = 2)\n",
    "print(\"Print OVR's crosstab (Conf matrix) :\" '\\n',\"===================\",'\\n',conf_matrix)# Matrix Table, not the graphic one\n",
    "\n",
    "# Now Plot the Conf Matrix\n",
    "\n",
    "es = ['T MayB','F MayB','F MayB',\n",
    "               'F No (F-Neg : T2Error)','T No (T-Neg)','F No (F-Neg : T2Error)',\n",
    "               'F Yes (F-Pos : T1Error)','F Yes (F-Pos : T1Error)','T Yes (T-Pos)']\n",
    "\n",
    "group_counts = ['{0:0.0f}'.format(value) for value in conf_matrix.flatten()]\n",
    "group_percentages = ['{0:.2%}'.format(value) for value in \n",
    "                                         conf_matrix.flatten()/np.sum(conf_matrix)]\n",
    "labels = [f'{v1}\\n{v2}\\n{v3}' for v1, v2, v3 in\n",
    "          zip(group_names,group_counts,group_percentages)]\n",
    "\n",
    "labels = np.asarray(labels).reshape(3,3)\n",
    "    \n",
    "sns.heatmap(conf_matrix, annot=labels, fmt='', cmap=\"Blues\")# cmap=\"OrRd\", \"OrRd_r\" colors\n",
    "plt.xlabel('Actual Label')\n",
    "plt.ylabel('Predicted Label')\n",
    "plt.title(\"CMatrix for OVR-Forest Baseline: PPRL, Multiclass [FirstORlastname]\")\n",
    "\n",
    "#plt.savefig('')\n",
    "plt.show() \n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b60b9d51",
   "metadata": {},
   "source": [
    "#### LReg Baseline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d4aeb55e",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "pipe4 =  logregPipe_SK.steps[1][1]\n",
    "logregmodel= pipe4.fit(x_train, y_train)\n",
    "\n",
    "# #-========================.  =============.   ===========\n",
    "y_pred_testlogReg= logregmodel.predict(x_test)\n",
    "lreg_pos_probs =logregmodel.predict_proba(x_test)[:,2]# this is  y_score in scikit\n",
    "\n",
    "\n",
    "metrics_DF_pipeSK.loc['Accuracy', 'LReg']= accuracy_score(y_pred = y_pred_testlogReg,\n",
    "                                                            y_true=y_test)\n",
    "metrics_DF_pipeSK.loc['Precision', 'LReg']= precision_score(y_pred = y_pred_testlogReg,\n",
    "                                                              y_true=y_test, average='weighted')\n",
    "metrics_DF_pipeSK.loc['Recall', 'LReg']= recall_score(y_pred = y_pred_testlogReg,\n",
    "                                                        y_true=y_test,average='weighted')\n",
    "metrics_DF_pipeSK.loc['F1', 'LReg']= f1_score(y_pred = y_pred_testlogReg,\n",
    "                                                y_true=y_test,average='weighted')\n",
    "\n",
    "# classification_report\n",
    "train_y_predlreg= logregmodel.predict(x_train)\n",
    "print('Lreg Sklearn train :', classification_report(y_train, train_y_predlreg),flush=True)\n",
    "print('Lreg Sklearn test :', classification_report(y_test, y_pred_testlogReg),flush=True)\n",
    "\n",
    "# GET CMATRIX \n",
    "\n",
    "sns.set(rc = {'figure.figsize': (15, 8)})\n",
    "conf_matrix = confusion_matrix(y_pred_testlogReg, y_test)\n",
    "np.set_printoptions(precision = 2)\n",
    "print(\"Print LReg Sklearn crosstab (Conf matrix) :\" '\\n',\"===================\",'\\n',conf_matrix)# Matrix Table, not the graphic one\n",
    "\n",
    "# Now Plot the Conf Matrix\n",
    "\n",
    "#sns.heatmap(conf_matrix, annot=True, fmt='d')\n",
    "group_names = ['T MayB','F MayB','F MayB',\n",
    "               'F No (F-Neg : T2Error)','T No (T-Neg)','F No (F-Neg : T2Error)',\n",
    "               'F Yes (F-Pos : T1Error)','F Yes (F-Pos : T1Error)','T Yes (T-Pos)']\n",
    "\n",
    "group_counts = ['{0:0.0f}'.format(value) for value in conf_matrix.flatten()]\n",
    "group_percentages = ['{0:.2%}'.format(value) for value in \n",
    "                                         conf_matrix.flatten()/np.sum(conf_matrix)]\n",
    "labels = [f'{v1}\\n{v2}\\n{v3}' for v1, v2, v3 in\n",
    "          zip(group_names,group_counts,group_percentages)]\n",
    "\n",
    "labels = np.asarray(labels).reshape(3,3)\n",
    "\n",
    "sns.heatmap(conf_matrix, annot=labels, fmt='', cmap=\"Blues\")# cmap=\"OrRd\", \"OrRd_r\" colors\n",
    "plt.xlabel('Actual Label')\n",
    "plt.ylabel('Predicted Label')\n",
    "plt.title(\"CMatrix for LReg Baseline: PPRL, Multiclass [FirstORlastname]\")\n",
    "\n",
    "\n",
    "#plt.savefig('')\n",
    "plt.show() "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90b60896",
   "metadata": {},
   "source": [
    "#### Results dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "411ed2ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "print('== Score Table: PPRL Model Set 1 Baseline) ==')\n",
    "\n",
    "100*metrics_DF_pipeSK\n",
    "### STORE THE DATAFRAME OF METRIC RESULT IN A VARIABLE AND SAVE IT AS A CSV FILE\n",
    "\n",
    "firstORlast_MClass_ResultTable=100*metrics_DF_pipeSK\n",
    "firstORlast_MClass_ResultTable.to_csv('PPRL_Multiclass_BaselineModels_FirstORlastname.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "24139ef8",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "print('PLOT RESULT VISUALS')\n",
    "fig, ax = plt.subplots(figsize = (8,5))\n",
    "metrics_DF_pipeSK.plot(kind='barh', ax=ax)\n",
    "ax.grid()\n",
    "\n",
    "plt.title('Subplots of PPRL MultiClass Baseline Models And Scores: FirstORlastname')\n",
    "#plt.savefig('PPRL_BaselineMultiClass_FirstORlast.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2e7f57b6",
   "metadata": {},
   "source": [
    "*****The performance of the baseline models is not satisfactory and it's unreliable due to the imbalance nature of the data. Therefore, we need to carry out some model improvement operations"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "aa2c3db5",
   "metadata": {},
   "source": [
    "# Pause here!!\n",
    "### Go to the file named, 'PPRL_Multiclass_Fselection+ParamTuning_2022' to perform two model performance improvement activities (Feature selection, and Parameter tuning)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7a08901a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "06bb10ee",
   "metadata": {},
   "source": [
    "## Now Use the Optimized Algorithms for Further Modeling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c3291b05",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Information threshold =0.1 (no approximation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1210baf8",
   "metadata": {},
   "outputs": [],
   "source": [
    "pprl_DF2.drop('zipCode', axis=1, inplace=True)# based on FImportance result"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe7120ba",
   "metadata": {},
   "source": [
    "### Use the output of parameter tuning and K search to create full imblearn pipelines"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab9d106e",
   "metadata": {},
   "source": [
    "##### Full Imblearn Pipeline for Naive Bayes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "47d6ff12",
   "metadata": {},
   "outputs": [],
   "source": [
    "smNbayes_Pipe= imbaPipe([('smote', SMOTE(sampling_strategy =\"auto\", k_neighbors=9, random_state=40)),\n",
    "                                           (\"dataScaler\", scaler), \n",
    "                         (\"naive\", MultinomialNB(alpha=1))])\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa41786f",
   "metadata": {},
   "source": [
    "##### Full Imblearn pipeline for OneVersusRest."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d9f55079",
   "metadata": {},
   "outputs": [],
   "source": [
    "smOvrRforest_Pipe = imbaPipe([('smote', SMOTE(sampling_strategy =\"auto\", k_neighbors=5, random_state=40)),\n",
    "                                (\"dataScaler\", scaler), \n",
    "                          (\"oneVRestRF\", OneVsRestClassifier(RandomForestClassifier(\n",
    "                              bootstrap =False,max_depth=25,\n",
    "                            min_samples_leaf= 5,n_estimators = 2,\n",
    "                              random_state= 42)))])\n",
    "              "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1682b575",
   "metadata": {},
   "source": [
    "##### Full Imblearn Pipeline for LReg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3ff46f59",
   "metadata": {},
   "outputs": [],
   "source": [
    "smLReg_Pipe = imbaPipe([('smote', SMOTE(sampling_strategy =\"auto\", k_neighbors=6, random_state=40)),\n",
    "                              (\"dataScaler\", scaler), (\"LReg\", LogisticRegression(C=1000, \n",
    "                                                            penalty='l2',solver='newton-cg'))])\n",
    "\n",
    "## LReg will automatically use multi_class='multinomial', penalty='l2' when it discovers it's a multiclass problem\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4b1fc28f",
   "metadata": {},
   "source": [
    "### Construct optimized models with the imblearn pipelines"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f30fea09",
   "metadata": {},
   "source": [
    "##### Create A Dataframe of Model Eval Metrics, with classifier's name as their columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "22e79af7",
   "metadata": {},
   "outputs": [],
   "source": [
    "metrics_DF_pipeSM = pd.DataFrame(index =['Accuracy','Precision','Recall','F1'], \n",
    "                          columns=['NBayes', 'OVR','LReg'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "624c110e",
   "metadata": {},
   "source": [
    "##### Naive Bayes Optimized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9e5774d0",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "pipeline1 = smNbayes_Pipe\n",
    "x_train.shape\n",
    "y_train.shape\n",
    "\n",
    "# Augment the training data\n",
    "\n",
    "x_train_oversampled, y_train_oversampled = pipeline1.steps[0][1].fit_resample(x_train, y_train)\n",
    "x_train_oversampled.shape\n",
    "y_train_oversampled.shape\n",
    "NBayesmodel= pipeline1.steps[2][1].fit(x_train_oversampled, y_train_oversampled)\n",
    "\n",
    "\n",
    "\n",
    "# #-========================.  =============.   ===========\n",
    "y_pred_testNBayes= NBayesmodel.predict(x_test)\n",
    "NBayes_pos_probs =NBayesmodel.predict_proba(x_test)[:,2]# this is  y_score in scikit\n",
    "\n",
    "\n",
    "metrics_DF_pipeSM.loc['Accuracy', 'NBayes']= accuracy_score(y_pred = y_pred_testNBayes,\n",
    "                                                            y_true=y_test)\n",
    "metrics_DF_pipeSM.loc['Precision', 'NBayes']= precision_score(y_pred = y_pred_testNBayes,\n",
    "                                                              y_true=y_test,average='weighted')\n",
    "metrics_DF_pipeSM.loc['Recall', 'NBayes']= recall_score(y_pred = y_pred_testNBayes,\n",
    "                                                        y_true=y_test,average='weighted')\n",
    "metrics_DF_pipeSM.loc['F1', 'NBayes']= f1_score(y_pred = y_pred_testNBayes,\n",
    "                                                y_true=y_test,average='weighted')\n",
    "# classification_report\n",
    "train_y_predNBayes= NBayesmodel.predict(x_train_oversampled)\n",
    "print('NBayes train :')\n",
    "print(classification_report(y_train_oversampled, train_y_predNBayes))\n",
    "\n",
    "print('NBayes test :')\n",
    "print(classification_report(y_test, y_pred_testNBayes))\n",
    "\n",
    "# Get matrix\n",
    "\n",
    "sns.set(rc = {'figure.figsize': (15, 8)})\n",
    "conf_matrix = confusion_matrix(y_pred_testNBayes, y_test)\n",
    "np.set_printoptions(precision = 2)\n",
    "print(\"Print NBayes's crosstab (Conf matrix) :\" '\\n',\"===================\",'\\n',conf_matrix)# Matrix Table, not the graphic one\n",
    "\n",
    "# Now Plot the Conf Matrix\n",
    "\n",
    "#sns.heatmap(conf_matrix, annot=True, fmt='d')\n",
    "\n",
    "group_names = ['T MayB','F MayB','F MayB',\n",
    "               'F No (F-Neg : T2Error)','T No (T-Neg)','F No (F-Neg : T2Error)',\n",
    "               'F Yes (F-Pos : T1Error)','F Yes (F-Pos : T1Error)','T Yes (T-Pos)']\n",
    "\n",
    "group_counts = ['{0:0.0f}'.format(value) for value in conf_matrix.flatten()]\n",
    "group_percentages = ['{0:.2%}'.format(value) for value in \n",
    "                                         conf_matrix.flatten()/np.sum(conf_matrix)]\n",
    "labels = [f'{v1}\\n{v2}\\n{v3}' for v1, v2, v3 in\n",
    "          zip(group_names,group_counts,group_percentages)]\n",
    "\n",
    "labels = np.asarray(labels).reshape(3,3)\n",
    "    \n",
    "sns.heatmap(conf_matrix, annot=labels, fmt='', cmap=\"Blues\")# other colors: cmap=\"OrRd\", \"OrRd_r\" colors\n",
    "plt.xlabel('Actual Label')\n",
    "plt.ylabel('Predicted Label')\n",
    "#plt.savefig('')\n",
    "plt.title(\"CMatrix for NBayes Model: PPRL, AugmMulticlass [FirstORlastname]\")\n",
    "\n",
    "plt.show()  \n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1fbc4541",
   "metadata": {},
   "source": [
    "##### OVR Optimized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "07ec68d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "pipeline2 = smOvrRforest_Pipe\n",
    "\n",
    "# Augment the training data\n",
    "\n",
    "x_train_oversampled, y_train_oversampled = pipeline2.steps[0][1].fit_resample(x_train, y_train)\n",
    "\n",
    "x_train_oversampled.shape\n",
    "y_train_oversampled.shape\n",
    "\n",
    "\n",
    "ovrModel= pipeline2.steps[2][1].fit(x_train_oversampled, y_train_oversampled)\n",
    "\n",
    "# #-========================.  =============.   ===========\n",
    "y_pred_testOVR= ovrModel.predict(x_test)\n",
    "y_pred_testOVR[5]\n",
    "OVR_pos_probs =ovrModel.predict_proba(x_test)[:,2]# this is  y_score in scikit\n",
    "\n",
    "OVR_pos_probs\n",
    "\n",
    "metrics_DF_pipeSM.loc['Accuracy', 'OVR']= accuracy_score(y_pred = y_pred_testOVR,\n",
    "                                                            y_true=y_test)\n",
    "metrics_DF_pipeSM.loc['Precision', 'OVR']= precision_score(y_pred = y_pred_testOVR,\n",
    "                                                              y_true=y_test, average='weighted')\n",
    "metrics_DF_pipeSM.loc['Recall', 'OVR']= recall_score(y_pred = y_pred_testOVR,\n",
    "                                                        y_true=y_test,average='weighted')\n",
    "metrics_DF_pipeSM.loc['F1', 'OVR']= f1_score(y_pred = y_pred_testOVR,\n",
    "                                                y_true=y_test,average='weighted')\n",
    "\n",
    "# classification_report\n",
    "train_y_predOVR= ovrModel.predict(x_train_oversampled)\n",
    "print('OVR train :', classification_report(y_train_oversampled, train_y_predOVR),flush=True)\n",
    "print('OVR test :', classification_report(y_test, y_pred_testOVR),flush=True)\n",
    "\n",
    "# Get Matrix\n",
    "\n",
    "sns.set(rc = {'figure.figsize': (15, 8)})\n",
    "conf_matrix = confusion_matrix(y_pred_testOVR, y_test)\n",
    "np.set_printoptions(precision = 2)\n",
    "print(\"Print OVR's crosstab (Conf matrix) :\" '\\n',\"===================\",'\\n',conf_matrix)# Matrix Table, not the graphic one\n",
    "    \n",
    "# Now Plot the Conf Matrix\n",
    "\n",
    "sns.heatmap(conf_matrix, annot=True, fmt='d')\n",
    "\n",
    "group_names = ['T MayB','F MayB','F MayB',\n",
    "               'F No (F-Neg : T2Error)','T No (T-Neg)','F No (F-Neg : T2Error)',\n",
    "               'F Yes (F-Pos : T1Error)','F Yes (F-Pos : T1Error)','T Yes (T-Pos)']\n",
    "\n",
    "group_counts = ['{0:0.0f}'.format(value) for value in conf_matrix.flatten()]\n",
    "group_percentages = ['{0:.2%}'.format(value) for value in \n",
    "                                         conf_matrix.flatten()/np.sum(conf_matrix)]\n",
    "labels = [f'{v1}\\n{v2}\\n{v3}' for v1, v2, v3 in\n",
    "          zip(group_names,group_counts,group_percentages)]\n",
    "\n",
    "labels = np.asarray(labels).reshape(3,3)\n",
    "\n",
    "sns.heatmap(conf_matrix, annot=labels, fmt='', cmap=\"Blues\")# cmap=\"OrRd\", \"OrRd_r\" colors\n",
    "plt.xlabel('Actual Label')\n",
    "plt.ylabel('Predicted Label')\n",
    "plt.title(\"CMatrix for OVR-Forest Model: PPRL, AugmMulticlass [FirstORlastname]\")\n",
    "\n",
    "#plt.savefig('')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5860d702",
   "metadata": {},
   "source": [
    "#### LReg Optimized"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c15dadf8",
   "metadata": {},
   "outputs": [],
   "source": [
    "pipeline4 = smLReg_Pipe\n",
    "\n",
    "# Augment the training data\n",
    "\n",
    "x_train_oversampled, y_train_oversampled = pipeline4.steps[0][1].fit_resample(x_train, y_train)\n",
    "x_train_oversampled.shape\n",
    "\n",
    "y_train_oversampled.shape\n",
    "LRegModel= pipeline4.steps[2][1].fit(x_train_oversampled, y_train_oversampled)\n",
    "\n",
    "# #-========================.  =============.   ===========\n",
    "y_pred_testLReg= LRegModel.predict(x_test)\n",
    "LReg_pos_probs =LRegModel.predict_proba(x_test)[:,2]# this is  y_score in scikit\n",
    "\n",
    "\n",
    "metrics_DF_pipeSM.loc['Accuracy', 'LReg']= accuracy_score(y_pred = y_pred_testLReg,\n",
    "                                                            y_true=y_test)\n",
    "metrics_DF_pipeSM.loc['Precision', 'LReg']= precision_score(y_pred = y_pred_testLReg,\n",
    "                                                              y_true=y_test,average='weighted')\n",
    "metrics_DF_pipeSM.loc['Recall', 'LReg']= recall_score(y_pred = y_pred_testLReg,\n",
    "                                                        y_true=y_test,average='weighted')\n",
    "metrics_DF_pipeSM.loc['F1', 'LReg']= f1_score(y_pred = y_pred_testLReg,\n",
    "                                                y_true=y_test,average='weighted')\n",
    "\n",
    "# classification_report\n",
    "train_y_predLReg= LRegModel.predict(x_train_oversampled)\n",
    "print('LReg train :', classification_report(y_train_oversampled, train_y_predLReg),flush=True)\n",
    "print('LReg test :', classification_report(y_test, y_pred_testLReg),flush=True)\n",
    "\n",
    "# Get Confusion Matrix\n",
    "\n",
    "sns.set(rc = {'figure.figsize': (15, 8)})\n",
    "conf_matrix = confusion_matrix(y_pred_testLReg, y_test)\n",
    "np.set_printoptions(precision = 2)\n",
    "print(\"Print LReg's crosstab (Conf matrix) :\" '\\n',\"===================\",'\\n',conf_matrix)# Matrix Table, not the graphic one\n",
    "\n",
    "#Now Plot the Conf Matrix\n",
    "\n",
    "group_names = ['T MayB','F MayB','F MayB',\n",
    "               'F No (F-Neg : T2Error)','T No (T-Neg)','F No (F-Neg : T2Error)',\n",
    "               'F Yes (F-Pos : T1Error)','F Yes (F-Pos : T1Error)','T Yes (T-Pos)']\n",
    "\n",
    "group_counts = ['{0:0.0f}'.format(value) for value in conf_matrix.flatten()]\n",
    "group_percentages = ['{0:.2%}'.format(value) for value in \n",
    "                                         conf_matrix.flatten()/np.sum(conf_matrix)]\n",
    "\n",
    "## Using list comprehension and formating with this label computation\n",
    "labels = [f'{v1}\\n{v2}\\n{v3}' for v1, v2, v3 in\n",
    "          zip(group_names,group_counts,group_percentages)]\n",
    "\n",
    "labels = np.asarray(labels).reshape(3,3)\n",
    "\n",
    "sns.heatmap(conf_matrix, annot=labels, fmt='', cmap=\"Blues\")# other colors: cmap=\"OrRd\", \"OrRd_r\" colors\n",
    "plt.xlabel('Actual Label')\n",
    "plt.ylabel('Predicted Label')\n",
    "plt.title(\"CMatrix for LReg Model: PPRL, AugmMulticlass [FirstORlastname]\")\n",
    "#plt.savefig('')\n",
    "plt.show() \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8d2c2a08",
   "metadata": {},
   "outputs": [],
   "source": [
    "### RESULT DATAFRAME"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0642178a",
   "metadata": {},
   "outputs": [],
   "source": [
    "print('== Score Table: PPRL Model Set 2 (Augmented data) ==')\n",
    "\n",
    "### STORE THE DATAFRAME OF METRIC RESULT IN A VARIABLE AND SAVE IT AS A CSV FILE\n",
    "Aug_firstORlast_MClass_ResultTable=100*metrics_DF_pipeSM\n",
    "#Aug_firstORlast_MClass_ResultTable.to_csv('PPRL_Multiclass_Models_Augm_FirstORlastname.csv')\n",
    "Aug_firstORlast_MClass_ResultTable"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5181b836",
   "metadata": {},
   "outputs": [],
   "source": [
    "### PLOT A BARH CHART OF RESULTS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "22635350",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "print('PLOT RESULT VISUALS')\n",
    "fig, ax = plt.subplots(figsize = (8,5))\n",
    "metrics_DF_pipeSM.plot(kind='barh', ax=ax)\n",
    "ax.grid()\n",
    "\n",
    "plt.title('Subplots of PPRL MultiClass Models And Scores: Augm_FirstORlastname')\n",
    "#plt.savefig('PPRL_MultiClass_Augm_FirstORlast.png')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "47710194",
   "metadata": {},
   "outputs": [],
   "source": [
    "print('FirstORlastname data says, \"The end!\"')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e0d705fb",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}