Confusion Matrix Visualization.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<font size=\"12\">Confusion Matrix Visualization</font>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In machine learning a confusion matrix is a kind of a table that is used to understand how well our classification model predictions perform, typically a supervised learning. It helps us a lot in understanding the model behavior and interpreting the results.  \n",
    "\n",
    "In this notebook we will not focus on interpreting a confusion matrix, as that could be find [here](https://en.wikipedia.org/wiki/Confusion_matrix).  Instead, our main objective is to present multiple techniques for visualizing a confusion matrix.\n",
    "\n",
    "#### Axes convention\n",
    "In the literature, we can find two variants for representing the samples in a confusion matrix:\n",
    "1. each row of the matrix represents samples in an *actual* class, and each column represents samples in a *predicted* class\n",
    "2. in the other variant, this arrangement is reversed, with each row representing samples in the *predicted* class and each column representing samples in the *actual* class.\n",
    "\n",
    "In this notebook we will use the first variant, where *actual* labels are on the horizontal axes and *predicted* labels on the vertical axes. Let us consider a binary classification problem with two classes: `0` (*Negative*) and `1` (*Positive*). The confusion matrix would be:\n",
    "\n",
    "<img src=\"images/cm_convention_table.png\">\n",
    "\n",
    "#### Topics\n",
    "Some of the topics covered in this notebook:  \n",
    "\n",
    "- normalized vs. unnormalized matrix\n",
    "- confusion matrix arrays \n",
    "- confusion matrix dataframes, side by side\n",
    "- plotting:\n",
    "  - single confusion matrix heatmap\n",
    "  - unnormalized and normalized heatmaps, side by side\n",
    "  - multiple (two or more) confusion matrices, in two columns\n",
    "  - using Sankey diagram\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:02.256747Z",
     "iopub.status.busy": "2023-07-31T15:07:02.256345Z",
     "iopub.status.idle": "2023-07-31T15:07:03.082291Z",
     "shell.execute_reply": "2023-07-31T15:07:03.081660Z",
     "shell.execute_reply.started": "2023-07-31T15:07:02.256709Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import os\n",
    "\n",
    "# Classification metrics\n",
    "from sklearn.metrics import confusion_matrix "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Plotly dependencies**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:03.083651Z",
     "iopub.status.busy": "2023-07-31T15:07:03.083392Z",
     "iopub.status.idle": "2023-07-31T15:07:03.246093Z",
     "shell.execute_reply": "2023-07-31T15:07:03.245217Z",
     "shell.execute_reply.started": "2023-07-31T15:07:03.083634Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from plotly import graph_objects as go\n",
    "\n",
    "# Set the appropriate renderer in Jupyter Lab to allow Plotly displays figure correctly\n",
    "# Set the default renderer explicitly as iframe \n",
    "import plotly.io as pio\n",
    "pio.renderers.default = 'iframe' \n",
    "\n",
    "# If multiple notebooks are using 'iframe', set different 'html_directory' for each notebook\n",
    "iframe_renderer = pio.renderers['iframe']\n",
    "iframe_renderer.html_directory='iframe_figures_n2'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Helper Functions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To help us with visualizations, we will import the module `metrics_utilities.py`. It is a collection of several helper functions:\n",
    "\n",
    "- `cm_cr()` - displays unnormalized and normalized confusion matrix dataframes side by side, and classification report \n",
    "- `plot_cm()` - plots one unnormalized or normalized confusion matrix heatmap\n",
    "- `plot_cm_unnorm_and_norm()` - plots one model's confusion matrix heatmaps without and with normalization, side by side\n",
    "- `plot_conf_matrices()` - plots heatmaps for normalized (default) or unnormalized confusion matrices for multiple models\n",
    "- `plot_cm_sankey()` - creates an interactive confusion matrix using Sankey diagram\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:03.247139Z",
     "iopub.status.busy": "2023-07-31T15:07:03.246869Z",
     "iopub.status.idle": "2023-07-31T15:07:03.944034Z",
     "shell.execute_reply": "2023-07-31T15:07:03.943098Z",
     "shell.execute_reply.started": "2023-07-31T15:07:03.247120Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Import the script from different folder\n",
    "import sys  \n",
    "sys.path.append('./scripts')\n",
    "\n",
    "import metrics_utilities as mu"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Preparing Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data used in this notebook is a result from one of my previous projects - [Bank-Churn-Prediction](https://github.com/zunicd/Bank-Churn-Prediction). <br>\n",
    "I saved the true labels and predictions in `.npy` format and in the next two cells we will load them."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Test True (Actual) Labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:03.946037Z",
     "iopub.status.busy": "2023-07-31T15:07:03.945722Z",
     "iopub.status.idle": "2023-07-31T15:07:03.964149Z",
     "shell.execute_reply": "2023-07-31T15:07:03.963188Z",
     "shell.execute_reply.started": "2023-07-31T15:07:03.946019Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# True labels\n",
    "y_test = np.load('./data/y_test.npy')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:03.965165Z",
     "iopub.status.busy": "2023-07-31T15:07:03.964815Z",
     "iopub.status.idle": "2023-07-31T15:07:04.057241Z",
     "shell.execute_reply": "2023-07-31T15:07:04.056428Z",
     "shell.execute_reply.started": "2023-07-31T15:07:03.965147Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Prepare predictions for our models\n",
    "pred_dt = np.load('./data/pred_dt.npy')\n",
    "pred_dl = np.load('./data/pred_dl.npy')\n",
    "pred_knn = np.load('./data/pred_knn.npy')\n",
    "pred_lr = np.load('./data/pred_lr.npy')\n",
    "pred_rf = np.load('./data/pred_rf.npy')\n",
    "pred_svm = np.load('./data/pred_svm.npy')\n",
    "pred_xgb = np.load('./data/pred_xgb.npy')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Create *models - predictions* Dictionaries"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are going to create the dictionaries to map names of models to their predictions. <br> \n",
    "We will be using them later for plotting and displaying results of the models."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "All models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.058378Z",
     "iopub.status.busy": "2023-07-31T15:07:04.058168Z",
     "iopub.status.idle": "2023-07-31T15:07:04.063431Z",
     "shell.execute_reply": "2023-07-31T15:07:04.062036Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.058361Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Create models_pred dictionary, contains all 7 models\n",
    "models_pred = {\n",
    "    'Decision Tree' : pred_dt,\n",
    "    'Deep Learning' : pred_dl,\n",
    "    'K-nearest Neighbors' : pred_knn,\n",
    "    'Logistic Regression' : pred_lr,\n",
    "    'Random Forest' : pred_rf,\n",
    "    'SVM' : pred_svm,\n",
    "    'XGBoost' : pred_xgb\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Two models only"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.065683Z",
     "iopub.status.busy": "2023-07-31T15:07:04.065149Z",
     "iopub.status.idle": "2023-07-31T15:07:04.072005Z",
     "shell.execute_reply": "2023-07-31T15:07:04.070820Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.065654Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Create models_pred_2 dictionary, contains 2 models\n",
    "models_pred_2 = {\n",
    "    'Decision Tree' : pred_dt,\n",
    "    'Random Forest' : pred_rf\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Names of Classes"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The `target_names` variable holds names of our classes. It will be used later for displaying evaluation results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.073394Z",
     "iopub.status.busy": "2023-07-31T15:07:04.073136Z",
     "iopub.status.idle": "2023-07-31T15:07:04.079026Z",
     "shell.execute_reply": "2023-07-31T15:07:04.078078Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.073375Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Names of our classes\n",
    "target_names = ['Stays', 'Exits']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Confusion Matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Why Normalized Confusion Matrix?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As real-life data is frequently imbalanced, utilizing a confusion matrix without normalization can potentially lead to misleading or incorrect conclusions.  \n",
    "In the subsequent sections of this notebook, we will present practical illustrations that demonstrate the significance of normalization in confusion matrices.\n",
    "\n",
    "The simplest way to display a confusion matrix is as raw numbers in an array."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Unnormalized Confusion Matrix**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.080097Z",
     "iopub.status.busy": "2023-07-31T15:07:04.079869Z",
     "iopub.status.idle": "2023-07-31T15:07:04.092683Z",
     "shell.execute_reply": "2023-07-31T15:07:04.091467Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.080080Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1979  410]\n",
      " [ 198  413]]\n"
     ]
    }
   ],
   "source": [
    "# Confusion matrix\n",
    "cm = confusion_matrix(y_test, pred_dt)\n",
    "print(cm)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Normalized Confusion Matrix**\n",
    "\n",
    "Few points to know:\n",
    "1. to calculate normalized version, divide each row element by the sum of the entire row\n",
    "2. each row represents the total number of true (actual) values for each class label \n",
    "3. the normalized matrix shows the percentage of predictions made by the model for each class with respect to the corresponding *actual* (*true*) label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.095973Z",
     "iopub.status.busy": "2023-07-31T15:07:04.095378Z",
     "iopub.status.idle": "2023-07-31T15:07:04.101052Z",
     "shell.execute_reply": "2023-07-31T15:07:04.100274Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.095945Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0.83 0.17]\n",
      " [0.32 0.68]]\n"
     ]
    }
   ],
   "source": [
    "# Normalized confusion matrix\n",
    "cmn = np.around(cm / cm.sum(axis=1)[:, np.newaxis], 2)\n",
    "print(cmn)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can display both matrices by using 2 `print` commands, or using the separator argument like below:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.102184Z",
     "iopub.status.busy": "2023-07-31T15:07:04.101916Z",
     "iopub.status.idle": "2023-07-31T15:07:04.109264Z",
     "shell.execute_reply": "2023-07-31T15:07:04.108320Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.102167Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1979  410]\n",
      " [ 198  413]]\n",
      "\n",
      "[[0.83 0.17]\n",
      " [0.32 0.68]]\n"
     ]
    }
   ],
   "source": [
    "# Normalized confusion matrix\n",
    "cmn = np.around(cm / cm.sum(axis=1)[:, np.newaxis], 2)\n",
    "print(cm, cmn, sep='\\n\\n')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This does not look very visually appealing, and this is just a binary class example.  \n",
    "For a multi-class classification problem, the confusion matrix becomes larger and more complex, making it harder to extract meaningful insights from the raw numbers alone.\n",
    "\n",
    "\n",
    "I think we could do better."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Display Confusion Matrix Dataframes Side by Side - `cm_cr()`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can utilize a Pandas DataFrame, and display both, unnormalized and normalized, matrices side by side.  \n",
    "For that we will employ `cm_cr()` function.  \n",
    "We can also decide to print a classification report, if desired.\n",
    "\n",
    "\n",
    "**REQUIREMENTS:**\n",
    "- pandas.Styler requires jinja2, use conda or pip to install it\n",
    "- Pandas 1.3.0+ for `precision` in [Styler.format](https://pandas.pydata.org/docs/reference/api/pandas.io.formats.style.Styler.format.html)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To call this function:  \n",
    "`mu.cm_cr(model_name, y_test, y_pred, target_names, cr=True)`\n",
    "\n",
    "<pre>Args:\n",
    "        model_name: name of the model\n",
    "        y_test: test target variable\n",
    "        y_pred: prediction\n",
    "        target_names: list of class names\n",
    "        cr: print classification report if True - default\n",
    "\n",
    "Returns:\n",
    "        Display confusion matrix dataframes side by side\n",
    "        and classification report if selected (default)</pre>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-03-29T15:52:01.153043Z",
     "iopub.status.busy": "2023-03-29T15:52:01.152203Z",
     "iopub.status.idle": "2023-03-29T15:52:01.161156Z",
     "shell.execute_reply": "2023-03-29T15:52:01.159592Z",
     "shell.execute_reply.started": "2023-03-29T15:52:01.152987Z"
    }
   },
   "source": [
    "### One Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrices Only"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To display confusion matrices only, we will use `cr=False`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.110349Z",
     "iopub.status.busy": "2023-07-31T15:07:04.110113Z",
     "iopub.status.idle": "2023-07-31T15:07:04.181045Z",
     "shell.execute_reply": "2023-07-31T15:07:04.180122Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.110331Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                       Random Forest\n",
      "                       =============\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "  <style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_2071f\" style='display:inline'>\n",
       "  <caption>Confusion Matrix</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >PREDICTED</th>\n",
       "      <th id=\"T_2071f_level0_col0\" class=\"col_heading level0 col0\" >Stays</th>\n",
       "      <th id=\"T_2071f_level0_col1\" class=\"col_heading level0 col1\" >Exits</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >ACTUAL</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_2071f_level0_row0\" class=\"row_heading level0 row0\" >Stays</th>\n",
       "      <td id=\"T_2071f_row0_col0\" class=\"data row0 col0\" >2137</td>\n",
       "      <td id=\"T_2071f_row0_col1\" class=\"data row0 col1\" >252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_2071f_level0_row1\" class=\"row_heading level0 row1\" >Exits</th>\n",
       "      <td id=\"T_2071f_row1_col0\" class=\"data row1 col0\" >233</td>\n",
       "      <td id=\"T_2071f_row1_col1\" class=\"data row1 col1\" >378</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "               <style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_3c5ad\" style='display:inline'>\n",
       "  <caption>Normalized Confusion Matrix</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >PREDICTED</th>\n",
       "      <th id=\"T_3c5ad_level0_col0\" class=\"col_heading level0 col0\" >Stays</th>\n",
       "      <th id=\"T_3c5ad_level0_col1\" class=\"col_heading level0 col1\" >Exits</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >ACTUAL</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_3c5ad_level0_row0\" class=\"row_heading level0 row0\" >Stays</th>\n",
       "      <td id=\"T_3c5ad_row0_col0\" class=\"data row0 col0\" >0.89</td>\n",
       "      <td id=\"T_3c5ad_row0_col1\" class=\"data row0 col1\" >0.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_3c5ad_level0_row1\" class=\"row_heading level0 row1\" >Exits</th>\n",
       "      <td id=\"T_3c5ad_row1_col0\" class=\"data row1 col0\" >0.38</td>\n",
       "      <td id=\"T_3c5ad_row1_col1\" class=\"data row1 col1\" >0.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "mu.cm_cr('Random Forest', y_test, pred_rf, target_names, cr=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Confusion Matrix and Classification Report"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If we do not enter the value for `cr` or we enter `cr=True`, classification report will be printed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.183240Z",
     "iopub.status.busy": "2023-07-31T15:07:04.182020Z",
     "iopub.status.idle": "2023-07-31T15:07:04.202779Z",
     "shell.execute_reply": "2023-07-31T15:07:04.201556Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.183200Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                       Random Forest\n",
      "                       =============\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "  <style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_0a26f\" style='display:inline'>\n",
       "  <caption>Confusion Matrix</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >PREDICTED</th>\n",
       "      <th id=\"T_0a26f_level0_col0\" class=\"col_heading level0 col0\" >Stays</th>\n",
       "      <th id=\"T_0a26f_level0_col1\" class=\"col_heading level0 col1\" >Exits</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >ACTUAL</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_0a26f_level0_row0\" class=\"row_heading level0 row0\" >Stays</th>\n",
       "      <td id=\"T_0a26f_row0_col0\" class=\"data row0 col0\" >2137</td>\n",
       "      <td id=\"T_0a26f_row0_col1\" class=\"data row0 col1\" >252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0a26f_level0_row1\" class=\"row_heading level0 row1\" >Exits</th>\n",
       "      <td id=\"T_0a26f_row1_col0\" class=\"data row1 col0\" >233</td>\n",
       "      <td id=\"T_0a26f_row1_col1\" class=\"data row1 col1\" >378</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "               <style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_595f0\" style='display:inline'>\n",
       "  <caption>Normalized Confusion Matrix</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >PREDICTED</th>\n",
       "      <th id=\"T_595f0_level0_col0\" class=\"col_heading level0 col0\" >Stays</th>\n",
       "      <th id=\"T_595f0_level0_col1\" class=\"col_heading level0 col1\" >Exits</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >ACTUAL</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_595f0_level0_row0\" class=\"row_heading level0 row0\" >Stays</th>\n",
       "      <td id=\"T_595f0_row0_col0\" class=\"data row0 col0\" >0.89</td>\n",
       "      <td id=\"T_595f0_row0_col1\" class=\"data row0 col1\" >0.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_595f0_level0_row1\" class=\"row_heading level0 row1\" >Exits</th>\n",
       "      <td id=\"T_595f0_row1_col0\" class=\"data row1 col0\" >0.38</td>\n",
       "      <td id=\"T_595f0_row1_col1\" class=\"data row1 col1\" >0.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "       Stays       0.90      0.89      0.90      2389\n",
      "       Exits       0.60      0.62      0.61       611\n",
      "\n",
      "    accuracy                           0.84      3000\n",
      "   macro avg       0.75      0.76      0.75      3000\n",
      "weighted avg       0.84      0.84      0.84      3000\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "mu.cm_cr('Random Forest', y_test, pred_rf, target_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "execution": {
     "iopub.status.busy": "2023-03-29T15:52:01.162456Z",
     "iopub.status.idle": "2023-03-29T15:52:01.162749Z",
     "shell.execute_reply": "2023-03-29T15:52:01.162635Z",
     "shell.execute_reply.started": "2023-03-29T15:52:01.162621Z"
    }
   },
   "source": [
    "### Multiple Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can create a dictionary to map model's names to their predictions.  \n",
    "Running the function `cm_cr()` in each iteration, will print confusion matrices and classification report (if `cm=True`) for all models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.203790Z",
     "iopub.status.busy": "2023-07-31T15:07:04.203580Z",
     "iopub.status.idle": "2023-07-31T15:07:04.232789Z",
     "shell.execute_reply": "2023-07-31T15:07:04.231775Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.203772Z"
    },
    "scrolled": true,
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                       Decision Tree\n",
      "                       =============\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "  <style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_93474\" style='display:inline'>\n",
       "  <caption>Confusion Matrix</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >PREDICTED</th>\n",
       "      <th id=\"T_93474_level0_col0\" class=\"col_heading level0 col0\" >Stays</th>\n",
       "      <th id=\"T_93474_level0_col1\" class=\"col_heading level0 col1\" >Exits</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >ACTUAL</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_93474_level0_row0\" class=\"row_heading level0 row0\" >Stays</th>\n",
       "      <td id=\"T_93474_row0_col0\" class=\"data row0 col0\" >1979</td>\n",
       "      <td id=\"T_93474_row0_col1\" class=\"data row0 col1\" >410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_93474_level0_row1\" class=\"row_heading level0 row1\" >Exits</th>\n",
       "      <td id=\"T_93474_row1_col0\" class=\"data row1 col0\" >198</td>\n",
       "      <td id=\"T_93474_row1_col1\" class=\"data row1 col1\" >413</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "               <style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_0a1ab\" style='display:inline'>\n",
       "  <caption>Normalized Confusion Matrix</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >PREDICTED</th>\n",
       "      <th id=\"T_0a1ab_level0_col0\" class=\"col_heading level0 col0\" >Stays</th>\n",
       "      <th id=\"T_0a1ab_level0_col1\" class=\"col_heading level0 col1\" >Exits</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >ACTUAL</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_0a1ab_level0_row0\" class=\"row_heading level0 row0\" >Stays</th>\n",
       "      <td id=\"T_0a1ab_row0_col0\" class=\"data row0 col0\" >0.83</td>\n",
       "      <td id=\"T_0a1ab_row0_col1\" class=\"data row0 col1\" >0.17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0a1ab_level0_row1\" class=\"row_heading level0 row1\" >Exits</th>\n",
       "      <td id=\"T_0a1ab_row1_col0\" class=\"data row1 col0\" >0.32</td>\n",
       "      <td id=\"T_0a1ab_row1_col1\" class=\"data row1 col1\" >0.68</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "       Stays       0.91      0.83      0.87      2389\n",
      "       Exits       0.50      0.68      0.58       611\n",
      "\n",
      "    accuracy                           0.80      3000\n",
      "   macro avg       0.71      0.75      0.72      3000\n",
      "weighted avg       0.83      0.80      0.81      3000\n",
      "\n",
      "\n",
      "                       Random Forest\n",
      "                       =============\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "  <style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_0327e\" style='display:inline'>\n",
       "  <caption>Confusion Matrix</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >PREDICTED</th>\n",
       "      <th id=\"T_0327e_level0_col0\" class=\"col_heading level0 col0\" >Stays</th>\n",
       "      <th id=\"T_0327e_level0_col1\" class=\"col_heading level0 col1\" >Exits</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >ACTUAL</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_0327e_level0_row0\" class=\"row_heading level0 row0\" >Stays</th>\n",
       "      <td id=\"T_0327e_row0_col0\" class=\"data row0 col0\" >2137</td>\n",
       "      <td id=\"T_0327e_row0_col1\" class=\"data row0 col1\" >252</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_0327e_level0_row1\" class=\"row_heading level0 row1\" >Exits</th>\n",
       "      <td id=\"T_0327e_row1_col0\" class=\"data row1 col0\" >233</td>\n",
       "      <td id=\"T_0327e_row1_col1\" class=\"data row1 col1\" >378</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "               <style type=\"text/css\">\n",
       "</style>\n",
       "<table id=\"T_003dc\" style='display:inline'>\n",
       "  <caption>Normalized Confusion Matrix</caption>\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >PREDICTED</th>\n",
       "      <th id=\"T_003dc_level0_col0\" class=\"col_heading level0 col0\" >Stays</th>\n",
       "      <th id=\"T_003dc_level0_col1\" class=\"col_heading level0 col1\" >Exits</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th class=\"index_name level0\" >ACTUAL</th>\n",
       "      <th class=\"blank col0\" >&nbsp;</th>\n",
       "      <th class=\"blank col1\" >&nbsp;</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th id=\"T_003dc_level0_row0\" class=\"row_heading level0 row0\" >Stays</th>\n",
       "      <td id=\"T_003dc_row0_col0\" class=\"data row0 col0\" >0.89</td>\n",
       "      <td id=\"T_003dc_row0_col1\" class=\"data row0 col1\" >0.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th id=\"T_003dc_level0_row1\" class=\"row_heading level0 row1\" >Exits</th>\n",
       "      <td id=\"T_003dc_row1_col0\" class=\"data row1 col0\" >0.38</td>\n",
       "      <td id=\"T_003dc_row1_col1\" class=\"data row1 col1\" >0.62</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "              precision    recall  f1-score   support\n",
      "\n",
      "       Stays       0.90      0.89      0.90      2389\n",
      "       Exits       0.60      0.62      0.61       611\n",
      "\n",
      "    accuracy                           0.84      3000\n",
      "   macro avg       0.75      0.76      0.75      3000\n",
      "weighted avg       0.84      0.84      0.84      3000\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "# Display classification report and confusion matrix for multiple models \n",
    "#  models_pred_2 dictionary\n",
    "for model, pred in models_pred_2.items():\n",
    "    mu.cm_cr(model, y_test, pred, target_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Plot Confusion Matrix Heatmaps for One Model - `plot_cm()`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's use Seaborn `heatmap()` to plot a confusion matrix for one model.  \n",
    "The helper function `plot_cm()` will help us with that."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To call this function:  \n",
    "`mu.plot_cm(model_name, y_test, y_pred, target_names, color, norm=True)`\n",
    "\n",
    "<pre>Args:\n",
    "        model_name: name of the model\n",
    "        y_test: test target variable\n",
    "        y_pred: prediction\n",
    "        target_names: list of class names\n",
    "        color: color palette\n",
    "        norm: plot normalized matrix if True - default \n",
    "              plot unnormalized matrix if False\n",
    "\n",
    "    Returns:\n",
    "        Plot one model's confusion matrix,\n",
    "         normalized (default) or unnormalized</pre>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Unnormalized Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.234143Z",
     "iopub.status.busy": "2023-07-31T15:07:04.233910Z",
     "iopub.status.idle": "2023-07-31T15:07:04.457465Z",
     "shell.execute_reply": "2023-07-31T15:07:04.456662Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.234125Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mu.plot_cm('XGBoost', y_test, pred_xgb, target_names, \"Blues\", norm=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Normalized Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-31T15:07:04.458785Z",
     "iopub.status.busy": "2023-07-31T15:07:04.458560Z",
     "iopub.status.idle": "2023-07-31T15:07:04.659093Z",
     "shell.execute_reply": "2023-07-31T15:07:04.657968Z",
     "shell.execute_reply.started": "2023-07-31T15:07:04.458767Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mu.plot_cm('XGBoost', y_test, pred_xgb, target_names, \"Blues\", norm=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Unnormalized and Normalized Confusion Matrix Heatmaps Side by Side - `plot_cm_unnorm_and_norm()`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can utilize the function `plot_cm_unnorm_and_norm()` to plot those two matrices from previous two cells, side by side."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To run this function:  \n",
    "`mu.plot_cm_unnorm_and_norm(model_name, y_test, y_pred, target_names, color)`\n",
    "\n",
    "<pre>Args:\n",
    "        model_name: name of the model\n",
    "        y_test: test target variable\n",
    "        y_pred: prediction\n",
    "        target_names: list of class names\n",
    "        color: color palette\n",
    "\n",
    "    Returns:\n",
    "        Plot one model's confusion matrix heatmaps side by side\n",
    "        left unnormalized and right normalized</pre>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-21T00:17:39.830131Z",
     "iopub.status.busy": "2023-07-21T00:17:39.829903Z",
     "iopub.status.idle": "2023-07-21T00:17:40.229550Z",
     "shell.execute_reply": "2023-07-21T00:17:40.228680Z",
     "shell.execute_reply.started": "2023-07-21T00:17:39.830112Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x288 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot confusion matrix and its normalized version.\n",
    "mu.plot_cm_unnorm_and_norm('XGBoost', y_test, pred_xgb, target_names, 'Reds')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Confusion Matrix Heatmaps for Multiple Models - `plot_conf_matrices()`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To compare confusion matrices for all our models in one plot, we can utilize the function `plot_conf_matrices()`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To run this function:     \n",
    "`mu.plot_conf_matrices(models_pred, y_test, target_names, color, norm=True)`\n",
    "\n",
    "<pre>Args:\n",
    "    models_pred: dictionary with model names as keys and predictions as values \n",
    "    y_test: test target variable\n",
    "    target_names: list of class names\n",
    "    color: color palette\n",
    "    norm: plot normalized matrices if True - default \n",
    "          plot unnormalized matrices if False\n",
    "\n",
    "Returns:\n",
    "    Plot all model's confusion matrix heatmaps, in 2 columns,\n",
    "    normalized (default) or unnormalized<pre>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Two Models - Normalized Connfusion Matrices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-21T00:17:40.231800Z",
     "iopub.status.busy": "2023-07-21T00:17:40.231541Z",
     "iopub.status.idle": "2023-07-21T00:17:40.465235Z",
     "shell.execute_reply": "2023-07-21T00:17:40.464364Z",
     "shell.execute_reply.started": "2023-07-21T00:17:40.231782Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot all confusion matrices \n",
    "mu.plot_conf_matrices(models_pred_2, y_test, target_names, 'Reds')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "### Two Models - Unnormalized Confusion Matrices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-21T00:17:40.466322Z",
     "iopub.status.busy": "2023-07-21T00:17:40.466084Z",
     "iopub.status.idle": "2023-07-21T00:17:40.708844Z",
     "shell.execute_reply": "2023-07-21T00:17:40.707787Z",
     "shell.execute_reply.started": "2023-07-21T00:17:40.466303Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x288 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot all confusion matrices \n",
    "mu.plot_conf_matrices(models_pred_2, y_test, target_names, 'Reds', norm=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "### All Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And finally, let's plot confusion matrices for all models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-21T00:17:40.710057Z",
     "iopub.status.busy": "2023-07-21T00:17:40.709807Z",
     "iopub.status.idle": "2023-07-21T00:17:42.263152Z",
     "shell.execute_reply": "2023-07-21T00:17:42.261890Z",
     "shell.execute_reply.started": "2023-07-21T00:17:40.710038Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x1152 with 7 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot all confusion matrices \n",
    "mu.plot_conf_matrices(models_pred, y_test, target_names, 'Reds')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>  \n",
    "\n",
    "**NOTE 1:**  \n",
    "\n",
    "- in matplotlib >3.5.2 `dpi` default value was changed from 72 to 100\n",
    "- deafult `figsize` was also changed: 6.0, 4.0 &rarr; 6.4, 4.8\n",
    "- in our functions we used `dpi=72` and `figsize=(6.0, 4.0)` to make plots smaller \n",
    "\n",
    "**NOTE 2:**  \n",
    "- the examples in this notebook are for binary classification\n",
    "- for multi-class classification, with many classes, some adjusstments might be needed in helper functions\n",
    "- increasing `dpi` above 72 might be sufficient"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sankey Diagram"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To dispiay a confusion matrix on an interactive and more elgant way, we will use **Sankey diagram**. \n",
    "By hovering over the diagram we can easily interpret the number of samples for each class (nodes) and how they were classified (links).  \n",
    "We will also include values from normalized confusion matrix, and by hovering over the links they will be displayed as percentages, for example:  \n",
    "\n",
    "<pre>413 (68%) <b>Exits</b> samples correctly classified as <b>Exits</b></pre>\n",
    "\n",
    "The notebook [Confusion Matrix as Sankey Diagram](Confusion%20Matrix%20as%20Sankey%20Diagram.ipynb) describes step-by-step the process of creating the function `plot_cm_sankey()`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To run this function:     \n",
    "`mu.plot_cm_sankey(model_name, y_test, y_pred, target_names=None)`\n",
    "\n",
    "<pre>Args:\n",
    "    model_name: name of the model\n",
    "    y_test: test target variable\n",
    "    y_pred: prediction\n",
    "    target_names: list of class names\n",
    "\n",
    "Returns:\n",
    "    Plot Sankey diagram of confusion matrix</pre>\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-21T00:17:42.265174Z",
     "iopub.status.busy": "2023-07-21T00:17:42.264758Z",
     "iopub.status.idle": "2023-07-21T00:17:42.340522Z",
     "shell.execute_reply": "2023-07-21T00:17:42.339549Z",
     "shell.execute_reply.started": "2023-07-21T00:17:42.265152Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<iframe\n",
       "    scrolling=\"no\"\n",
       "    width=\"645px\"\n",
       "    height=\"520\"\n",
       "    src=\"iframe_figures_n2/figure_21.html\"\n",
       "    frameborder=\"0\"\n",
       "    allowfullscreen\n",
       "></iframe>\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mu.plot_cm_sankey('Decision Tree', y_test, pred_dt, target_names)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### How to interpret our Sankey confusion matrix?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To interpret our Sankey confusion matrix, with **Stays** representing the negative class and **Exits** representing the positive class, take note of the following key elements:\n",
    "\n",
    "- **Nodes:** The source nodes (\"ACTUAL **Stays**\" and \"ACTUAL **Exits**\" ) are positioned on the left, representing the *actual* class labels. The target nodes (\"PREDICTED **Stays**\" and \"PREDICTED **Exits**\") are on the right, representing the *predicted* class labels.\n",
    "- **Node Size:** The size of each node is proportional to the number of samples belonging to that specific class, offering insights into the class distribution.\n",
    "- **Links:** The links between nodes show the flow of samples during classification. Wider links indicate a higher number of samples classified correctly (shown in green) or incorrectly (depicted in red) between the respective class labels.\n",
    "- **Tooltips:** Hovering over the nodes and links provides additional information, displaying the numerical and textual representation of the confusion matrix. This interactive feature enhances the understanding of classification outcomes and allows for a more detailed examination of the model's performance.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3x3 Confusion Matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's see how this works for an 3x3 confusion matrix.\n",
    "\n",
    "We will use data from my project [T2D-Predictions](https://github.com/zunicd/T2D-Predictions)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-21T00:17:42.341596Z",
     "iopub.status.busy": "2023-07-21T00:17:42.341406Z",
     "iopub.status.idle": "2023-07-21T00:17:42.358057Z",
     "shell.execute_reply": "2023-07-21T00:17:42.357243Z",
     "shell.execute_reply.started": "2023-07-21T00:17:42.341578Z"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Actual (True) labels\n",
    "t2d_y_test = np.load('./data/t2d_y_test.npy')\n",
    "# Prediction from random forest model\n",
    "t2d_pred_rf = np.load('./data/t2d_pred_rf.npy')\n",
    "# Classes\n",
    "t2d_classes = ['no_diabetes', 'pre_diabetes', 'diabetes']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-21T00:17:42.359110Z",
     "iopub.status.busy": "2023-07-21T00:17:42.358924Z",
     "iopub.status.idle": "2023-07-21T00:17:42.426568Z",
     "shell.execute_reply": "2023-07-21T00:17:42.425351Z",
     "shell.execute_reply.started": "2023-07-21T00:17:42.359093Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<iframe\n",
       "    scrolling=\"no\"\n",
       "    width=\"645px\"\n",
       "    height=\"520\"\n",
       "    src=\"iframe_figures_n2/figure_23.html\"\n",
       "    frameborder=\"0\"\n",
       "    allowfullscreen\n",
       "></iframe>\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mu.plot_cm_sankey('Random Forest', t2d_y_test, t2d_pred_rf, t2d_classes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hover over the diagram to get more information about the confusion matrix. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**NOTE 3:**  \n",
    "\n",
    "- Depending on the values in the confusion matrix, especially in the case of multi-class classification, Plotly might display the target nodes (PREDICTED labels) in a different order compared to the source nodes (ACTUAL labels).\n",
    "\n",
    "- However, rest assured that despite the potential rearrangement of nodes, all the connected links in the Sankey Diagram will accurately represent the correct width (values) as per the confusion matrix. The flow volumes will be appropriately maintained, enabling a precise representation of the classification outcomes between the actual and predicted labels."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let's compare it with the heatmap plot."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2023-07-21T00:17:42.427779Z",
     "iopub.status.busy": "2023-07-21T00:17:42.427543Z",
     "iopub.status.idle": "2023-07-21T00:17:42.966428Z",
     "shell.execute_reply": "2023-07-21T00:17:42.965655Z",
     "shell.execute_reply.started": "2023-07-21T00:17:42.427757Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x288 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot confusion matrix and its normalized version.\n",
    "mu.plot_cm_unnorm_and_norm('Random Forest', t2d_y_test, t2d_pred_rf, t2d_classes, 'Reds')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>  \n",
    "\n",
    "**NOTE 4 - for Jupyter Lab users:**  \n",
    "\n",
    "-  Set the appropriate renderer in Jupyter Lab to allow Plotly displays figure correctly\n",
    "- As suggested in the [Plotly documentation](https://plotly.com/python/renderers/#interactive-renderers), you might set the default renderer explicitly as iframe by adding following lines into your codes:\n",
    "\n",
    "```python\n",
    "import plotly.io as pio\n",
    "pio.renderers.default = 'iframe'\n",
    "```\n",
    "\n",
    "- This renderer writes figures out as standalone HTML files and then displays [iframe](https://www.w3schools.com/html/html_iframe.asp) elements that reference these files\n",
    "- They are stored in a subdirectory named *iframe_figures*\n",
    "- The names of the files are given based on the execution number of the notebook cell that produced the figure\n",
    "-  Storing multiple notebooks using an iframe renderer in the same directory, could result in notebooks overwriting each other's figures\n",
    "- To avoid this, add the following lines in each notebook to set different names for the `html_directory`. in this notebook we named it *iframe_figures_n2*\n",
    "\n",
    "\n",
    "```python\n",
    "iframe_renderer = pio.renderers['iframe']\n",
    "iframe_renderer.html_directory='iframe_figures_n2'\n",
    "```\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "omdena_p310",
   "language": "python",
   "name": "conda-env-omdena_p310-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.10"
  },
  "toc-autonumbering": true,
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}