notebooks/property_maintenance_fines.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "# Property Maintenance Fines \n",
        "\n",
        "**Predicting the probability that a set of blight tickets will be paid on time**\n",
        "\n",
        "**Supervised Learning. Classification**\n",
        "\n",
        "Source: [Applied Machine Learning in Python | Coursera](https://www.coursera.org/learn/python-machine-learning). Solved with classical machine learning classifiers [here](https://github.com/angelmtenor/applied-ML-coursera/blob/master/4_property_maintenance_fines.ipynb)\n",
        "\n",
        "Data provided by Michigan Data Science Team ([MDST](http://midas.umich.edu/mdst/)), the Michigan Student Symposium for Interdisciplinary Statistical Sciences ([MSSISS](https://sites.lsa.umich.edu/mssiss/)) and the City of Detroit [Detroit Open Data Portal](https://data.detroitmi.gov/).\n",
        " \n",
        "\n",
        "Each row of the dataset   corresponds to a single blight ticket, and includes information about when, why, and to whom each ticket was issued. The target variable is compliance, which is True if the ticket was paid early, on time, or within one month of the hearing data, False if the ticket was paid after the hearing date or not at all, and Null if the violator was found not responsible. \n",
        "\n",
        "**Features**\n",
        "    \n",
        "    ticket_id - unique identifier for tickets\n",
        "    agency_name - Agency that issued the ticket\n",
        "    inspector_name - Name of inspector that issued the ticket\n",
        "    violator_name - Name of the person/organization that the ticket was issued to\n",
        "    violation_street_number, violation_street_name, violation_zip_code - Address where the violation occurred\n",
        "    mailing_address_str_number, mailing_address_str_name, city, state, zip_code, non_us_str_code, country - Mailing address of the violator\n",
        "    ticket_issued_date - Date and time the ticket was issued\n",
        "    hearing_date - Date and time the violator's hearing was scheduled\n",
        "    violation_code, violation_description - Type of violation\n",
        "    disposition - Judgment and judgement type\n",
        "    fine_amount - Violation fine amount, excluding fees\n",
        "    admin_fee - $20 fee assigned to responsible judgments\n",
        "state_fee - $10 fee assigned to responsible judgments\n",
        "    late_fee - 10% fee assigned to responsible judgments\n",
        "    discount_amount - discount applied, if any\n",
        "    clean_up_cost - DPW clean-up or graffiti removal cost\n",
        "    judgment_amount - Sum of all fines and fees\n",
        "    grafitti_status - Flag for graffiti violations\n",
        "    \n",
        "** Labels **\n",
        "\n",
        "    payment_amount - Amount paid, if any\n",
        "    payment_date - Date payment was made, if it was received\n",
        "    payment_status - Current payment status as of Feb 1 2017\n",
        "    balance_due - Fines and fees still owed\n",
        "    collection_status - Flag for payments in collections\n",
        "    compliance [target variable for prediction]\n",
        "     Null = Not responsible\n",
        "     0 = Responsible, non-compliant\n",
        "     1 = Responsible, compliant\n",
        "    compliance_detail - More information on why each ticket was marked compliant or non-compliant"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:18:42 - DEBUG \t Path changed to /home/angel/DRIVE/0_repos/data-science-keras\n",
            "23:18:42 - DEBUG \t HARDWARE:\n",
            "23:18:43 - DEBUG \t CPU:\tIntel(R) Core(TM) i5-8265U CPU @ 1.60GHz\n",
            "23:18:43 - DEBUG \t RAM:\t15 GB\n",
            "23:18:43 - DEBUG \t -- No GPU  --\n",
            "23:18:43 - DEBUG \t OS:\tLinux-5.15.0-57-generic-x86_64-with-glibc2.35\n",
            "23:18:43 - DEBUG \t ENV:\t/home/angel/miniconda3/envs/ds-keras\n",
            "23:18:43 - DEBUG \t PYTHON:\t3.10.4 (main, Mar 31 2022, 08:41:55) [GCC 7.5.0]\n",
            "23:18:43 - DEBUG \t tensorflow                   2.11.0\n",
            "23:18:43 - DEBUG \t pandas                        1.5.2\n",
            "23:18:43 - DEBUG \t numpy                        1.24.1\n",
            "23:18:43 - DEBUG \t EXECUTION PATH: /home/angel/DRIVE/0_repos/data-science-keras\n",
            "23:18:43 - DEBUG \t EXECUTION DATE: Mon Jan 16 23:18:43 2023\n"
          ]
        }
      ],
      "source": [
        "import os, sys\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "\n",
        "import ds_boost\n",
        "from tensorflow import keras\n",
        "\n",
        "log = ds_boost.logger.init(level=\"DEBUG\", save_log=False)\n",
        "\n",
        "ds_boost.set_parent_execution_path()\n",
        "ds_boost.info_system()\n",
        "# sns.set_palette(\"GnBu_d\")\n",
        "# ds_boost.reproducible\n",
        "\n",
        "N_JOBS = 10\n",
        "\n",
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "##  1. Data Processing"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "250306 rows \n",
            "34 columns \n",
            "target: ['compliance']\n"
          ]
        }
      ],
      "source": [
        "data_path = \"data/property_maintenance_fines_data.parquet\"\n",
        "target = [\"compliance\"]\n",
        "\n",
        "# df_original = pd.read_csv(data_path, encoding=\"iso-8859-1\", dtype=\"unicode\")\n",
        "df_original = pd.read_parquet(data_path)  # , encoding=\"iso-8859-1\", dtype=\"unicode\")\n",
        "\n",
        "print(\"{} rows \\n{} columns \\ntarget: {}\".format(*df_original.shape, target))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {},
      "outputs": [],
      "source": [
        "# df_original.to_parquet(\"data/property_maintenance_fines_data.parquet\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Explore and Clean the target"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0.0     148283\n",
            "None     90426\n",
            "1.0      11597\n",
            "Name: compliance, dtype: int64\n"
          ]
        }
      ],
      "source": [
        "print(df_original[target].squeeze().value_counts(dropna=False))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "0.0    148283\n",
            "1.0     11597\n",
            "Name: compliance, dtype: int64\n",
            "(159880, 34)\n"
          ]
        }
      ],
      "source": [
        "# Remove rows with NULL targets\n",
        "\n",
        "df_original = df_original.dropna(subset=target)\n",
        "\n",
        "print(df_original[target].squeeze().value_counts())\n",
        "print(df_original.shape)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "**Imbalanced target: the evaluation metric used in this problem is the Area Under the ROC Curve**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "###  Split original data into training and validation test set"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {},
      "outputs": [],
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "df, df_test = train_test_split(df_original, test_size=0.2, stratify=df_original[target], random_state=0)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "**To avoid data leakage, only the training dataframe, df, will be explored and processed here** "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Show the training data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ticket_id</th>\n",
              "      <th>agency_name</th>\n",
              "      <th>inspector_name</th>\n",
              "      <th>violator_name</th>\n",
              "      <th>violation_street_number</th>\n",
              "      <th>violation_street_name</th>\n",
              "      <th>violation_zip_code</th>\n",
              "      <th>mailing_address_str_number</th>\n",
              "      <th>mailing_address_str_name</th>\n",
              "      <th>city</th>\n",
              "      <th>...</th>\n",
              "      <th>clean_up_cost</th>\n",
              "      <th>judgment_amount</th>\n",
              "      <th>payment_amount</th>\n",
              "      <th>balance_due</th>\n",
              "      <th>payment_date</th>\n",
              "      <th>payment_status</th>\n",
              "      <th>collection_status</th>\n",
              "      <th>grafitti_status</th>\n",
              "      <th>compliance_detail</th>\n",
              "      <th>compliance</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>131030</th>\n",
              "      <td>159232</td>\n",
              "      <td>Department of Public Works</td>\n",
              "      <td>Montgomery-Coit, Kimberlye</td>\n",
              "      <td>JOHNSON-GREENE, MARGUSIE F.</td>\n",
              "      <td>11645.0</td>\n",
              "      <td>LAKEPOINTE</td>\n",
              "      <td>None</td>\n",
              "      <td>11645</td>\n",
              "      <td>LAKEPOINTE</td>\n",
              "      <td>DETROIT</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>85.0</td>\n",
              "      <td>85.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2008-09-12 00:00:00</td>\n",
              "      <td>PAID IN FULL</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>non-compliant by late payment more than 1 month</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29573</th>\n",
              "      <td>49039</td>\n",
              "      <td>Buildings, Safety Engineering &amp; Env Department</td>\n",
              "      <td>Watson, Jerry</td>\n",
              "      <td>BAPT CHURCH, HOLY  TABERNACLE</td>\n",
              "      <td>3184.0</td>\n",
              "      <td>CANFIELD</td>\n",
              "      <td>None</td>\n",
              "      <td>3184</td>\n",
              "      <td>E. CANFIELD</td>\n",
              "      <td>DETROIT</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>305.0</td>\n",
              "      <td>305.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>2007-06-26 00:00:00</td>\n",
              "      <td>PAID IN FULL</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>non-compliant by late payment more than 1 month</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2 rows \u00d7 34 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "       ticket_id                                     agency_name  \\\n",
              "131030    159232                      Department of Public Works   \n",
              "29573      49039  Buildings, Safety Engineering & Env Department   \n",
              "\n",
              "                    inspector_name                  violator_name  \\\n",
              "131030  Montgomery-Coit, Kimberlye    JOHNSON-GREENE, MARGUSIE F.   \n",
              "29573                Watson, Jerry  BAPT CHURCH, HOLY  TABERNACLE   \n",
              "\n",
              "       violation_street_number violation_street_name violation_zip_code  \\\n",
              "131030                 11645.0            LAKEPOINTE               None   \n",
              "29573                   3184.0              CANFIELD               None   \n",
              "\n",
              "       mailing_address_str_number mailing_address_str_name     city  ...  \\\n",
              "131030                      11645               LAKEPOINTE  DETROIT  ...   \n",
              "29573                        3184              E. CANFIELD  DETROIT  ...   \n",
              "\n",
              "       clean_up_cost judgment_amount payment_amount balance_due  \\\n",
              "131030           0.0            85.0           85.0         0.0   \n",
              "29573            0.0           305.0          305.0         0.0   \n",
              "\n",
              "               payment_date payment_status collection_status grafitti_status  \\\n",
              "131030  2008-09-12 00:00:00   PAID IN FULL              None            None   \n",
              "29573   2007-06-26 00:00:00   PAID IN FULL              None            None   \n",
              "\n",
              "                                      compliance_detail compliance  \n",
              "131030  non-compliant by late payment more than 1 month        0.0  \n",
              "29573   non-compliant by late payment more than 1 month        0.0  \n",
              "\n",
              "[2 rows x 34 columns]"
            ]
          },
          "execution_count": 7,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df.head(2)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Missing values"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 800x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "ds_boost.missing(df)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Transform Data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Remove irrelevant features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "(127904, 12)\n"
          ]
        }
      ],
      "source": [
        "def remove_features(df):\n",
        "    relevant_col = [\n",
        "        \"agency_name\",\n",
        "        \"violation_street_name\",\n",
        "        \"city\",\n",
        "        \"state\",\n",
        "        \"violator_name\",\n",
        "        \"violation_code\",\n",
        "        \"late_fee\",\n",
        "        \"discount_amount\",\n",
        "        \"judgment_amount\",\n",
        "        \"disposition\",\n",
        "        \"fine_amount\",\n",
        "        \"compliance\",\n",
        "    ]\n",
        "    df = df[relevant_col]\n",
        "\n",
        "    return df\n",
        "\n",
        "\n",
        "df = remove_features(df)\n",
        "\n",
        "print(df.shape)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Classify variables"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:18:46 - DEBUG \t 4 numerical features: \t ['late_fee', 'discount_amount', 'judgment_amount', 'fine_amount']\n",
            "23:18:46 - DEBUG \t 7 categorical features: \t ['disposition', 'agency_name', 'city', 'violator_name', 'violation_street_name', 'state', 'violation_code']\n",
            "23:18:46 - DEBUG \t Target: \t\tcompliance (object)\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>late_fee</th>\n",
              "      <th>discount_amount</th>\n",
              "      <th>judgment_amount</th>\n",
              "      <th>fine_amount</th>\n",
              "      <th>disposition</th>\n",
              "      <th>agency_name</th>\n",
              "      <th>city</th>\n",
              "      <th>violator_name</th>\n",
              "      <th>violation_street_name</th>\n",
              "      <th>state</th>\n",
              "      <th>violation_code</th>\n",
              "      <th>compliance</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Type</th>\n",
              "      <td>float32</td>\n",
              "      <td>float32</td>\n",
              "      <td>float32</td>\n",
              "      <td>float32</td>\n",
              "      <td>category</td>\n",
              "      <td>category</td>\n",
              "      <td>category</td>\n",
              "      <td>category</td>\n",
              "      <td>category</td>\n",
              "      <td>category</td>\n",
              "      <td>category</td>\n",
              "      <td>object</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "     late_fee discount_amount judgment_amount fine_amount disposition  \\\n",
              "Type  float32         float32         float32     float32    category   \n",
              "\n",
              "     agency_name      city violator_name violation_street_name     state  \\\n",
              "Type    category  category      category              category  category   \n",
              "\n",
              "     violation_code compliance  \n",
              "Type       category     object  "
            ]
          },
          "execution_count": 10,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "num = [\"late_fee\", \"discount_amount\", \"judgment_amount\", \"fine_amount\"]\n",
        "\n",
        "df = ds_boost.sort_columns_by_type(df, target, numerical=num)\n",
        "\n",
        "pd.DataFrame(dict(df.dtypes), index=[\"Type\"])[df.columns].head()  # show data types"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Remove low-frequency categorical values"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:18:46 - DEBUG \t Variables changed to 'category': {columns_to_convert}\n",
            "23:18:46 - DEBUG \t Removing 3595 categories from feature city\n",
            "23:18:46 - DEBUG \t Removing 72749 categories from feature violator_name\n",
            "23:18:46 - DEBUG \t Removing 1377 categories from feature violation_street_name\n",
            "23:18:46 - DEBUG \t Removing 39 categories from feature state\n",
            "23:18:46 - DEBUG \t Removing 151 categories from feature violation_code\n"
          ]
        }
      ],
      "source": [
        "df, dict_categories = ds_boost.remove_categories(df, target=target, ratio=0.001, show=False)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "####  Fill missing values\n",
        "\n",
        "Missing categorical values filled by 'Other'\n",
        "There are no numerical missing values"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:18:46 - DEBUG \t Missing categorical filled with label: \"Other\"\n"
          ]
        }
      ],
      "source": [
        "df = ds_boost.fill_simple(df, target, missing_categorical=\"Other\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:18:46 - DEBUG \t No missing values found\n"
          ]
        }
      ],
      "source": [
        "ds_boost.missing(df);"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Visualize the data"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Categorical features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1700x200 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1700x200 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "for i in [\"state\", \"disposition\"]:\n",
        "    ds_boost.show_categorical(df[[i]])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Target vs Categorical features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {},
      "outputs": [],
      "source": [
        "# for i in [\"state\", \"disposition\"]:\n",
        "#     ds_boost.show_target_vs_categorical(df[[i, target[0]]], target) # TODO Fix"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Numerical features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1700x200 with 4 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "ds_boost.show_numerical(df, kde=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Target vs Numerical features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {},
      "outputs": [],
      "source": [
        "# ds_boost.show_target_vs_numerical(df, target, point_size=10, jitter=0.3, fit_reg=True)\n",
        "# plt.ylim(ymin=-0.2, ymax=1.2)  # TODO: Optimize (high computation)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Correlation between numerical features and target"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {},
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/home/angel/DRIVE/0_repos/data-science-keras/src/ds_boost/ds_boost.py:778: FutureWarning: The default value of numeric_only in DataFrame.corr is deprecated. In a future version, it will default to False. Select only valid columns or specify the value of numeric_only to silence this warning.\n",
            "  corr = copy_df.corr().loc[numerical_f, target].fillna(0).sort_values(target, ascending=False).round(2)\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "[]"
            ]
          },
          "execution_count": 18,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 600x300 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "ds_boost.correlation(df, target, figsize=(6, 3))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "## 2. Neural Network Model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Select the features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>late_fee</th>\n",
              "      <th>discount_amount</th>\n",
              "      <th>judgment_amount</th>\n",
              "      <th>fine_amount</th>\n",
              "      <th>disposition</th>\n",
              "      <th>agency_name</th>\n",
              "      <th>city</th>\n",
              "      <th>violator_name</th>\n",
              "      <th>violation_street_name</th>\n",
              "      <th>state</th>\n",
              "      <th>violation_code</th>\n",
              "      <th>compliance</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>131030</th>\n",
              "      <td>5.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>85.0</td>\n",
              "      <td>50.0</td>\n",
              "      <td>Responsible by Determination</td>\n",
              "      <td>Department of Public Works</td>\n",
              "      <td>DETROIT</td>\n",
              "      <td>Other</td>\n",
              "      <td>LAKEPOINTE</td>\n",
              "      <td>MI</td>\n",
              "      <td>9-1-103(C)</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29573</th>\n",
              "      <td>25.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>305.0</td>\n",
              "      <td>250.0</td>\n",
              "      <td>Responsible by Default</td>\n",
              "      <td>Buildings, Safety Engineering &amp; Env Department</td>\n",
              "      <td>DETROIT</td>\n",
              "      <td>Other</td>\n",
              "      <td>Other</td>\n",
              "      <td>MI</td>\n",
              "      <td>9-1-36(a)</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "        late_fee  discount_amount  judgment_amount  fine_amount  \\\n",
              "131030       5.0              0.0             85.0         50.0   \n",
              "29573       25.0              0.0            305.0        250.0   \n",
              "\n",
              "                         disposition  \\\n",
              "131030  Responsible by Determination   \n",
              "29573         Responsible by Default   \n",
              "\n",
              "                                           agency_name     city violator_name  \\\n",
              "131030                      Department of Public Works  DETROIT         Other   \n",
              "29573   Buildings, Safety Engineering & Env Department  DETROIT         Other   \n",
              "\n",
              "       violation_street_name state violation_code compliance  \n",
              "131030            LAKEPOINTE    MI     9-1-103(C)        0.0  \n",
              "29573                  Other    MI      9-1-36(a)        0.0  "
            ]
          },
          "execution_count": 19,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "droplist = []  # features to drop\n",
        "\n",
        "# For the model 'data' instead of 'df'\n",
        "data = df.copy()\n",
        "# del(df)\n",
        "data.drop(droplist, axis=\"columns\", inplace=True)\n",
        "data.head(2)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Scale numerical variables\n",
        "Shift and scale numerical variables to a standard normal distribution. The scaling factors are saved to be used for predictions."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {},
      "outputs": [],
      "source": [
        "data, scale_param = ds_boost.scale(data)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Create dummy features\n",
        "Replace categorical features (no target) with dummy features"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>late_fee</th>\n",
              "      <th>discount_amount</th>\n",
              "      <th>judgment_amount</th>\n",
              "      <th>fine_amount</th>\n",
              "      <th>compliance</th>\n",
              "      <th>disposition_Responsible (Fine Waived) by Deter</th>\n",
              "      <th>disposition_Responsible by Admission</th>\n",
              "      <th>disposition_Responsible by Default</th>\n",
              "      <th>disposition_Responsible by Determination</th>\n",
              "      <th>disposition_Other</th>\n",
              "      <th>...</th>\n",
              "      <th>violation_code_9-1-105</th>\n",
              "      <th>violation_code_9-1-110(a)</th>\n",
              "      <th>violation_code_9-1-206</th>\n",
              "      <th>violation_code_9-1-209</th>\n",
              "      <th>violation_code_9-1-36(a)</th>\n",
              "      <th>violation_code_9-1-43(a) - (Dwellin</th>\n",
              "      <th>violation_code_9-1-43(a) - (Structu</th>\n",
              "      <th>violation_code_9-1-81(a)</th>\n",
              "      <th>violation_code_9-1-82(d) - (Dwellin</th>\n",
              "      <th>violation_code_Other</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>131030</th>\n",
              "      <td>-0.422306</td>\n",
              "      <td>-0.049474</td>\n",
              "      <td>-0.451255</td>\n",
              "      <td>-0.453714</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>29573</th>\n",
              "      <td>-0.126339</td>\n",
              "      <td>-0.049474</td>\n",
              "      <td>-0.154289</td>\n",
              "      <td>-0.156975</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2657</th>\n",
              "      <td>-0.126339</td>\n",
              "      <td>-0.049474</td>\n",
              "      <td>-0.154289</td>\n",
              "      <td>-0.156975</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>...</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>1</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>3 rows \u00d7 446 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "        late_fee  discount_amount  judgment_amount  fine_amount compliance  \\\n",
              "131030 -0.422306        -0.049474        -0.451255    -0.453714        0.0   \n",
              "29573  -0.126339        -0.049474        -0.154289    -0.156975        0.0   \n",
              "2657   -0.126339        -0.049474        -0.154289    -0.156975        0.0   \n",
              "\n",
              "       disposition_Responsible (Fine Waived) by Deter  \\\n",
              "131030                                              0   \n",
              "29573                                               0   \n",
              "2657                                                0   \n",
              "\n",
              "       disposition_Responsible by Admission  \\\n",
              "131030                                    0   \n",
              "29573                                     0   \n",
              "2657                                      0   \n",
              "\n",
              "       disposition_Responsible by Default  \\\n",
              "131030                                  0   \n",
              "29573                                   1   \n",
              "2657                                    1   \n",
              "\n",
              "       disposition_Responsible by Determination disposition_Other  ...  \\\n",
              "131030                                        1                 0  ...   \n",
              "29573                                         0                 0  ...   \n",
              "2657                                          0                 0  ...   \n",
              "\n",
              "       violation_code_9-1-105 violation_code_9-1-110(a)  \\\n",
              "131030                      0                         0   \n",
              "29573                       0                         0   \n",
              "2657                        0                         0   \n",
              "\n",
              "       violation_code_9-1-206 violation_code_9-1-209 violation_code_9-1-36(a)  \\\n",
              "131030                      0                      0                        0   \n",
              "29573                       0                      0                        1   \n",
              "2657                        0                      0                        1   \n",
              "\n",
              "       violation_code_9-1-43(a) - (Dwellin  \\\n",
              "131030                                   0   \n",
              "29573                                    0   \n",
              "2657                                     0   \n",
              "\n",
              "       violation_code_9-1-43(a) - (Structu violation_code_9-1-81(a)  \\\n",
              "131030                                   0                        0   \n",
              "29573                                    0                        0   \n",
              "2657                                     0                        0   \n",
              "\n",
              "       violation_code_9-1-82(d) - (Dwellin violation_code_Other  \n",
              "131030                                   0                    0  \n",
              "29573                                    0                    0  \n",
              "2657                                     0                    0  \n",
              "\n",
              "[3 rows x 446 columns]"
            ]
          },
          "execution_count": 21,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "data, dict_dummies = ds_boost.replace_by_dummies(data, target)\n",
        "\n",
        "model_features = [f for f in data if f not in target]  # sorted neural network inputs\n",
        "\n",
        "data.head(3)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Split the data into training and validation sets"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {},
      "outputs": [],
      "source": [
        "val_size = 0.2\n",
        "random_state = 0\n",
        "\n",
        "\n",
        "def validation_split(data, val_size=0.25):\n",
        "    train, test = train_test_split(data, test_size=val_size, random_state=random_state, stratify=data[target])\n",
        "\n",
        "    # Separate the data into features and target (x=features, y=target)\n",
        "    x_train, y_train = train.drop(target, axis=1).values, train[target].values\n",
        "    x_val, y_val = test.drop(target, axis=1).values, test[target].values\n",
        "    # _nc: non-categorical yet (needs one-hot encoding)\n",
        "\n",
        "    return x_train, y_train, x_val, y_val\n",
        "\n",
        "\n",
        "x_train, y_train, x_val, y_val = validation_split(data, val_size=val_size)\n",
        "\n",
        "# x_train = x_train.astype(np.float16)\n",
        "y_train = y_train.astype(np.float16)\n",
        "# X_val = x_val.astype(np.float16)\n",
        "y_val = y_val.astype(np.float16)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Encode the output"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {},
      "outputs": [],
      "source": [
        "def one_hot_output(y_train, y_val):\n",
        "    num_classes = len(np.unique(y_train))\n",
        "    y_train = keras.utils.to_categorical(y_train, num_classes)\n",
        "    y_val = keras.utils.to_categorical(y_val, num_classes)\n",
        "    return y_train, y_val\n",
        "\n",
        "\n",
        "y_train, y_val = one_hot_output(y_train, y_val)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "train size \t X:(102323, 445) \t Y:(102323, 2)\n",
            "val size \t X:(25581, 445) \t Y:(25581, 2)\n"
          ]
        }
      ],
      "source": [
        "print(\"train size \\t X:{} \\t Y:{}\".format(x_train.shape, y_train.shape))\n",
        "print(\"val size \\t X:{} \\t Y:{}\".format(x_val.shape, y_val.shape))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Build a dummy classifier"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:19:12 - DEBUG \t Confusion matrix: \n",
            " [[23725     0]\n",
            " [ 1856     0]]\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Loss</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>ROC-AUC</th>\n",
              "      <th>F1-score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>Dummy</th>\n",
              "      <td>1.16</td>\n",
              "      <td>0.93</td>\n",
              "      <td>0.0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       Loss  Accuracy  Precision  Recall  ROC-AUC  F1-score\n",
              "Dummy  1.16      0.93        0.0       0        0         0"
            ]
          },
          "execution_count": 25,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "ds_boost.dummy_clf(x_train, y_train[:, 1], x_val, y_val[:, 1])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Build a random forest classifier (best of grid search)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "CPU times: user 1min 59s, sys: 455 ms, total: 1min 59s\n",
            "Wall time: 18 s\n"
          ]
        }
      ],
      "source": [
        "from sklearn.ensemble import RandomForestClassifier\n",
        "\n",
        "\n",
        "%time clf_random_forest_opt = RandomForestClassifier(n_estimators = 30, max_features=150, \\\n",
        "                                max_depth=13, class_weight='balanced', n_jobs=N_JOBS, \\\n",
        "                                   random_state=0).fit(x_train, np.ravel(y_train[:,1]))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:19:31 - DEBUG \t Confusion matrix: \n",
            " [[21173  2552]\n",
            " [  730  1126]]\n"
          ]
        }
      ],
      "source": [
        "y_pred = clf_random_forest_opt.predict(x_val).reshape([-1, 1])\n",
        "ds_boost.binary_classification_scores(y_val[:, 1], y_pred);"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Build the Neural Network for Binary Classification"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            " Layer (type)                Output Shape              Param #   \n",
            "=================================================================\n",
            " dense (Dense)               (None, 55)                24530     \n",
            "                                                                 \n",
            " dense_1 (Dense)             (None, 2)                 112       \n",
            "                                                                 \n",
            "=================================================================\n",
            "Total params: 24,642\n",
            "Trainable params: 24,642\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ]
        }
      ],
      "source": [
        "cw = ds_boost.get_class_weight(y_train[:, 1])  # class weight (imbalanced target)\n",
        "\n",
        "import keras\n",
        "from keras.models import Sequential\n",
        "from keras.layers.core import Dense, Dropout\n",
        "\n",
        "\n",
        "def build_nn(input_size, output_size, summary=False):\n",
        "    input_nodes = input_size // 8\n",
        "\n",
        "    model = Sequential()\n",
        "    model.add(Dense(input_nodes, input_dim=input_size, activation=\"relu\"))\n",
        "\n",
        "    model.add(Dense(output_size, activation=\"softmax\"))\n",
        "\n",
        "    if summary:\n",
        "        model.summary()\n",
        "\n",
        "    model.compile(loss=\"binary_crossentropy\", optimizer=\"adam\", metrics=[\"accuracy\"])\n",
        "\n",
        "    return model\n",
        "\n",
        "\n",
        "model = build_nn(x_train.shape[1], y_train.shape[1], summary=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Train the Neural Network"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Training ....\n",
            "Epoch 1/100\n",
            "35/35 [==============================] - 1s 17ms/step - loss: 0.6584 - accuracy: 0.7086 - val_loss: 0.6157 - val_accuracy: 0.7166\n",
            "Epoch 2/100\n",
            "35/35 [==============================] - 0s 10ms/step - loss: 0.5845 - accuracy: 0.7405 - val_loss: 0.5459 - val_accuracy: 0.7867\n",
            "Epoch 2: early stopping\n",
            "time: \t 2.0 s\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1400x400 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:19:33 - DEBUG \t Training loss:  \t0.5845\n",
            "23:19:33 - DEBUG \t Validation loss: \t 0.5459\n",
            "23:19:33 - DEBUG \t Training accuracy: \t0.741\n",
            "23:19:33 - DEBUG \t Validation accuracy:\t0.787\n",
            "3198/3198 [==============================] - 4s 1ms/step\n",
            "\n",
            "\n",
            " ROC_AUC train:\t0.80 \n",
            "\n",
            "800/800 [==============================] - 1s 1ms/step\n",
            "\n",
            "\n",
            " ROC_AUC val:\t0.80\n"
          ]
        }
      ],
      "source": [
        "import os\n",
        "from time import time\n",
        "\n",
        "model_path = os.path.join(\"models\", \"detroit.h5\")\n",
        "\n",
        "\n",
        "def train_nn(model, x_train, y_train, validation_data=None, path=False, show=True):\n",
        "    \"\"\"\n",
        "    Train the neural network model. If no validation_data is provided, a split for validation\n",
        "    will be used\n",
        "    \"\"\"\n",
        "\n",
        "    if show:\n",
        "        print(\"Training ....\")\n",
        "\n",
        "    callbacks = [keras.callbacks.EarlyStopping(monitor=\"val_loss\", patience=0, verbose=1)]\n",
        "    t0 = time()\n",
        "\n",
        "    history = model.fit(\n",
        "        x_train,\n",
        "        y_train,\n",
        "        epochs=100,\n",
        "        batch_size=2048,\n",
        "        class_weight=cw,\n",
        "        verbose=1,\n",
        "        validation_split=0.3,\n",
        "        validation_data=validation_data,\n",
        "        callbacks=callbacks,\n",
        "    )\n",
        "\n",
        "    if show:\n",
        "        print(\"time: \\t {:.1f} s\".format(time() - t0))\n",
        "        ds_boost.show_training(history)\n",
        "\n",
        "    if path:\n",
        "        model.save(path)\n",
        "        print(\"\\nModel saved at\", path)\n",
        "\n",
        "    return history\n",
        "\n",
        "\n",
        "model = None\n",
        "model = build_nn(x_train.shape[1], y_train.shape[1], summary=False)\n",
        "train_nn(model, x_train, y_train, path=None)\n",
        "\n",
        "\n",
        "from sklearn.metrics import roc_auc_score\n",
        "\n",
        "y_pred_train = model.predict(x_train, verbose=1)\n",
        "print(\"\\n\\n ROC_AUC train:\\t{:.2f} \\n\".format(roc_auc_score(y_train, y_pred_train)))\n",
        "y_pred_val = model.predict(x_val, verbose=1)\n",
        "print(\"\\n\\n ROC_AUC val:\\t{:.2f}\".format(roc_auc_score(y_val, y_pred_val)))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Validate the model (validation set)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:19:41 - DEBUG \t Confusion matrix: \n",
            " [[19058  4667]\n",
            " [  616  1240]]\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Loss</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>ROC-AUC</th>\n",
              "      <th>F1-score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <td>0.5</td>\n",
              "      <td>0.79</td>\n",
              "      <td>0.21</td>\n",
              "      <td>0.67</td>\n",
              "      <td>0.8</td>\n",
              "      <td>0.32</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Loss  Accuracy  Precision  Recall  ROC-AUC  F1-score\n",
              "    0.5      0.79       0.21    0.67      0.8      0.32"
            ]
          },
          "execution_count": 30,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "ds_boost.binary_classification_scores(y_val[:, 1], y_pred_val[:, 1], return_dataframe=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Evaluate the final model (test set)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>ticket_id</th>\n",
              "      <th>agency_name</th>\n",
              "      <th>inspector_name</th>\n",
              "      <th>violator_name</th>\n",
              "      <th>violation_street_number</th>\n",
              "      <th>violation_street_name</th>\n",
              "      <th>violation_zip_code</th>\n",
              "      <th>mailing_address_str_number</th>\n",
              "      <th>mailing_address_str_name</th>\n",
              "      <th>city</th>\n",
              "      <th>...</th>\n",
              "      <th>clean_up_cost</th>\n",
              "      <th>judgment_amount</th>\n",
              "      <th>payment_amount</th>\n",
              "      <th>balance_due</th>\n",
              "      <th>payment_date</th>\n",
              "      <th>payment_status</th>\n",
              "      <th>collection_status</th>\n",
              "      <th>grafitti_status</th>\n",
              "      <th>compliance_detail</th>\n",
              "      <th>compliance</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>29156</th>\n",
              "      <td>48696</td>\n",
              "      <td>Department of Public Works</td>\n",
              "      <td>Funchess, Mitchell</td>\n",
              "      <td>CORP, CONTIMORTGAGE</td>\n",
              "      <td>8200.0</td>\n",
              "      <td>HEYDEN</td>\n",
              "      <td>None</td>\n",
              "      <td>3815</td>\n",
              "      <td>WEST TEMPLE</td>\n",
              "      <td>SALT LAKE CITY</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>140.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>140.0</td>\n",
              "      <td>None</td>\n",
              "      <td>NO PAYMENT APPLIED</td>\n",
              "      <td>IN COLLECTION</td>\n",
              "      <td>None</td>\n",
              "      <td>non-compliant by no payment</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>125262</th>\n",
              "      <td>152329</td>\n",
              "      <td>Buildings, Safety Engineering &amp; Env Department</td>\n",
              "      <td>Doetsch, James</td>\n",
              "      <td>JACKSON, THEO</td>\n",
              "      <td>13821.0</td>\n",
              "      <td>GLENWOOD</td>\n",
              "      <td>None</td>\n",
              "      <td>1464</td>\n",
              "      <td>PO BOX</td>\n",
              "      <td>DETROIT</td>\n",
              "      <td>...</td>\n",
              "      <td>0.0</td>\n",
              "      <td>305.0</td>\n",
              "      <td>0.0</td>\n",
              "      <td>305.0</td>\n",
              "      <td>None</td>\n",
              "      <td>NO PAYMENT APPLIED</td>\n",
              "      <td>None</td>\n",
              "      <td>None</td>\n",
              "      <td>non-compliant by no payment</td>\n",
              "      <td>0.0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>2 rows \u00d7 34 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "       ticket_id                                     agency_name  \\\n",
              "29156      48696                      Department of Public Works   \n",
              "125262    152329  Buildings, Safety Engineering & Env Department   \n",
              "\n",
              "            inspector_name        violator_name violation_street_number  \\\n",
              "29156   Funchess, Mitchell  CORP, CONTIMORTGAGE                  8200.0   \n",
              "125262      Doetsch, James        JACKSON, THEO                 13821.0   \n",
              "\n",
              "       violation_street_name violation_zip_code mailing_address_str_number  \\\n",
              "29156                 HEYDEN               None                       3815   \n",
              "125262              GLENWOOD               None                       1464   \n",
              "\n",
              "       mailing_address_str_name            city  ... clean_up_cost  \\\n",
              "29156               WEST TEMPLE  SALT LAKE CITY  ...           0.0   \n",
              "125262                   PO BOX         DETROIT  ...           0.0   \n",
              "\n",
              "       judgment_amount payment_amount balance_due payment_date  \\\n",
              "29156            140.0            0.0       140.0         None   \n",
              "125262           305.0            0.0       305.0         None   \n",
              "\n",
              "            payment_status collection_status grafitti_status  \\\n",
              "29156   NO PAYMENT APPLIED     IN COLLECTION            None   \n",
              "125262  NO PAYMENT APPLIED              None            None   \n",
              "\n",
              "                  compliance_detail compliance  \n",
              "29156   non-compliant by no payment        0.0  \n",
              "125262  non-compliant by no payment        0.0  \n",
              "\n",
              "[2 rows x 34 columns]"
            ]
          },
          "execution_count": 31,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_test.head(2)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Process test data with training set parameters (no data leakage)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:19:41 - DEBUG \t 4 numerical features: \t ['late_fee', 'discount_amount', 'judgment_amount', 'fine_amount']\n",
            "23:19:41 - DEBUG \t 7 categorical features: \t ['disposition', 'agency_name', 'city', 'violator_name', 'violation_street_name', 'state', 'violation_code']\n",
            "23:19:41 - DEBUG \t Target: \t\tcompliance (object)\n",
            "23:19:41 - DEBUG \t Variables changed to 'category': {columns_to_convert}\n",
            "23:19:41 - DEBUG \t Missing categorical filled with label: \"Other\"\n"
          ]
        }
      ],
      "source": [
        "df_test = remove_features(df_test)\n",
        "df_test = ds_boost.sort_columns_by_type(df_test, target, numerical=num)\n",
        "df_test, _ = ds_boost.remove_categories(df_test, target=target, show=False, dict_categories=dict_categories)\n",
        "df_test = ds_boost.fill_simple(df_test, target, missing_categorical=\"Other\")\n",
        "df_test, _ = ds_boost.scale(df_test, scale_param)\n",
        "df_test, _ = ds_boost.replace_by_dummies(df_test, target, dict_dummies)\n",
        "df_test = df_test[model_features + target]  # sort columns to match training features order"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {},
      "outputs": [],
      "source": [
        "def separate_x_y(data):\n",
        "    \"\"\"Separate the data into features and target (x=features, y=target)\"\"\"\n",
        "\n",
        "    x, y = data.drop(target, axis=1).values, data[target].values\n",
        "    x = x.astype(np.float16)\n",
        "    y = y.astype(np.float16)\n",
        "\n",
        "    return x, y\n",
        "\n",
        "\n",
        "x_test, y_test = separate_x_y(df_test)\n",
        "\n",
        "y_test = keras.utils.to_categorical(y_test, 2)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "####  Random Forest model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:19:43 - DEBUG \t Confusion matrix: \n",
            " [[26456  3201]\n",
            " [  955  1364]]\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Loss</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>ROC-AUC</th>\n",
              "      <th>F1-score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <td>0.46</td>\n",
              "      <td>0.87</td>\n",
              "      <td>0.3</td>\n",
              "      <td>0.59</td>\n",
              "      <td>0.81</td>\n",
              "      <td>0.4</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Loss  Accuracy  Precision  Recall  ROC-AUC  F1-score\n",
              "   0.46      0.87        0.3    0.59     0.81       0.4"
            ]
          },
          "execution_count": 34,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "y_pred = clf_random_forest_opt.predict_proba(x_test)\n",
        "ds_boost.binary_classification_scores(y_test[:, 1], y_pred[:, 1], return_dataframe=True)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Importance</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>late_fee</th>\n",
              "      <td>0.29</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>disposition_Responsible by Default</th>\n",
              "      <td>0.28</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>judgment_amount</th>\n",
              "      <td>0.09</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>disposition_Responsible by Admission</th>\n",
              "      <td>0.07</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>disposition_Responsible by Determination</th>\n",
              "      <td>0.05</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>fine_amount</th>\n",
              "      <td>0.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>discount_amount</th>\n",
              "      <td>0.02</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>violation_code_9-1-36(a)</th>\n",
              "      <td>0.01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>city_DETROIT</th>\n",
              "      <td>0.01</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>violation_code_22-2-61</th>\n",
              "      <td>0.01</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                          Importance\n",
              "late_fee                                        0.29\n",
              "disposition_Responsible by Default              0.28\n",
              "judgment_amount                                 0.09\n",
              "disposition_Responsible by Admission            0.07\n",
              "disposition_Responsible by Determination        0.05\n",
              "fine_amount                                     0.02\n",
              "discount_amount                                 0.02\n",
              "violation_code_9-1-36(a)                        0.01\n",
              "city_DETROIT                                    0.01\n",
              "violation_code_22-2-61                          0.01"
            ]
          },
          "execution_count": 35,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 800x600 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "ds_boost.feature_importance(model_features, clf_random_forest_opt)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "#### Neural Network model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 36,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "1000/1000 [==============================] - 1s 1ms/step\n",
            "23:19:45 - DEBUG \t Confusion matrix: \n",
            " [[23736  5921]\n",
            " [  786  1533]]\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Loss</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>ROC-AUC</th>\n",
              "      <th>F1-score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th></th>\n",
              "      <td>0.5</td>\n",
              "      <td>0.79</td>\n",
              "      <td>0.21</td>\n",
              "      <td>0.66</td>\n",
              "      <td>0.79</td>\n",
              "      <td>0.31</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "   Loss  Accuracy  Precision  Recall  ROC-AUC  F1-score\n",
              "    0.5      0.79       0.21    0.66     0.79      0.31"
            ]
          },
          "execution_count": 36,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "y_pred = model.predict(x_test, verbose=1)[:, 1]\n",
        "ds_boost.binary_classification_scores(y_test[:, 1], y_pred, return_dataframe=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {},
      "source": [
        "### Compare with other non-neural ML models"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 37,
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "23:19:45 - DEBUG \t Naive Bayes\n",
            "23:19:47 - DEBUG \t Random Forest\n",
            "23:19:55 - DEBUG \t Extremely Randomized Trees\n",
            "23:20:04 - DEBUG \t LGBM\n"
          ]
        },
        {
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Time (s)</th>\n",
              "      <th>Loss</th>\n",
              "      <th>Accuracy</th>\n",
              "      <th>Precision</th>\n",
              "      <th>Recall</th>\n",
              "      <th>ROC-AUC</th>\n",
              "      <th>F1-score</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>LGBM</th>\n",
              "      <td>1.12</td>\n",
              "      <td>0.19</td>\n",
              "      <td>0.94</td>\n",
              "      <td>0.90</td>\n",
              "      <td>0.24</td>\n",
              "      <td>0.81</td>\n",
              "      <td>0.38</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Random Forest</th>\n",
              "      <td>7.29</td>\n",
              "      <td>0.20</td>\n",
              "      <td>0.94</td>\n",
              "      <td>0.93</td>\n",
              "      <td>0.22</td>\n",
              "      <td>0.81</td>\n",
              "      <td>0.35</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Extremely Randomized Trees</th>\n",
              "      <td>8.87</td>\n",
              "      <td>0.21</td>\n",
              "      <td>0.93</td>\n",
              "      <td>0.99</td>\n",
              "      <td>0.03</td>\n",
              "      <td>0.81</td>\n",
              "      <td>0.06</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Naive Bayes</th>\n",
              "      <td>1.25</td>\n",
              "      <td>21.65</td>\n",
              "      <td>0.36</td>\n",
              "      <td>0.09</td>\n",
              "      <td>0.83</td>\n",
              "      <td>0.59</td>\n",
              "      <td>0.16</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                            Time (s)   Loss  Accuracy  Precision  Recall  \\\n",
              "LGBM                            1.12   0.19      0.94       0.90    0.24   \n",
              "Random Forest                   7.29   0.20      0.94       0.93    0.22   \n",
              "Extremely Randomized Trees      8.87   0.21      0.93       0.99    0.03   \n",
              "Naive Bayes                     1.25  21.65      0.36       0.09    0.83   \n",
              "\n",
              "                            ROC-AUC  F1-score  \n",
              "LGBM                           0.81      0.38  \n",
              "Random Forest                  0.81      0.35  \n",
              "Extremely Randomized Trees     0.81      0.06  \n",
              "Naive Bayes                    0.59      0.16  "
            ]
          },
          "execution_count": 37,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "ds_boost.ml_classification(x_train, y_train[:, 1], x_test, y_test[:, 1])"
      ]
    }
  ],
  "metadata": {
    "coursera": {
      "course_slug": "python-machine-learning",
      "graded_item_id": "nNS8l",
      "launcher_item_id": "yWWk7",
      "part_id": "w8BSS"
    },
    "kernelspec": {
      "display_name": "ds-keras",
      "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": "3.10.4 (main, Mar 31 2022, 08:41:55) [GCC 7.5.0]"
    },
    "vscode": {
      "interpreter": {
        "hash": "366a4202e041eebdd8a4edd8a0024926c2e45e658d7ba92f3c9b74c9b0ce97a9"
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 2
}