tabulator.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Board of Estimates Tabulator  \n",
    "The purpose of this software tool is to use the pdf files that store the minutes of Baltimore's Board of Estimates to create a small database with linked tables for entities that could possibly include:\n",
    "\n",
    "- __meetings__\n",
    "    - one entity per BoE meeting\n",
    "    - primary key is the date\n",
    "- __agreements__\n",
    "    - primary key is BAN plus the partner organization's name\n",
    "    - features include: date, dollar amount, BAN, description\n",
    "- __prequalifications__\n",
    "- __contractors__ \n",
    "- __personnel__\n",
    "- __reclassifications__"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setup\n",
    "### Import packages  \n",
    "**We may want to break visualization code into a separate notebook"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from datetime import datetime\n",
    "import pandas as pd\n",
    "import requests\n",
    "from bs4 import BeautifulSoup\n",
    "import re\n",
    "from pathlib import Path\n",
    "import time \n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "from common.utils import del_dir_contents, is_empty\n",
    "from common.parse_utils import parse_pdf, store_pdf_text_to_df\n",
    "from common.scrape_utils import get_boe_pdfs\n",
    "\n",
    "from pandas.plotting import register_matplotlib_converters\n",
    "register_matplotlib_converters() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define directories and urls\n",
    "We'll also create the `pdf_files` directory if it doesn't exist already."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The pdf directory already exists.\n"
     ]
    }
   ],
   "source": [
    "base_url = \"https://comptroller.baltimorecity.gov/\"\n",
    "minutes_url = base_url + \"boe/meetings/minutes\"\n",
    "\n",
    "root = Path.cwd()\n",
    "pdf_dir = root / \"pdf_files\"\n",
    "\n",
    "try:\n",
    "    pdf_dir.mkdir(parents=True, exist_ok=False)\n",
    "except FileExistsError:\n",
    "    print(\"The pdf directory already exists.\")\n",
    "else:\n",
    "    print(\"The pdf directory has been created.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Store PDFs to local directory  \n",
    "The following code downloads all the .pdf files with minutes from the Board of Estimates and saves them in your local version of this repository. \n",
    "\n",
    "The code will skip this time-consuming step if it detects files within the `pdf_files` directory.\n",
    "\n",
    "The tricky part of this is getting a correct date for every file. Some files have a typo somewhere in html, so the functions `store_boe_pdfs()` and  `parse_meeting_date()` are built to handle these errors. We may see new types of errors appear in the future, however."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There was an issue parsing tag <a href=\"/files/2570-27262018-06-27pdf\"> </a> on page https://comptroller.baltimorecity.gov/minutes-2018: '' is not a parseable date\n",
      "There was an issue parsing tag <a href=\"/files/2570-27262018-06-27pdf\"> </a> on page https://comptroller.baltimorecity.gov/minutes-2018: '' is not a parseable date\n",
      "There was an issue parsing tag <a href=\"/files/2570-27262018-06-27pdf\" style=\";;color: rgb(9, 113, 138);\"> </a> on page https://comptroller.baltimorecity.gov/minutes-2018: '' is not a parseable date\n",
      "No new BOE minutes to download\n"
     ]
    }
   ],
   "source": [
    "# set to true if you'll be repeatedly running store_boe_pdfs()\n",
    "downloaded_files = get_boe_pdfs(minutes_url, base_url)\n",
    "pri"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Process and prepare sample data\n",
    "### Parse only a few sample pdfs\n",
    "While we're still in development mode, we can save some time by only parsing a few pdfs instead of all 500+ "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4680 BOARD OF ESITMATES November 20, 2013 MINUTES REGULAR MEETING Honorable Bernard C. \"Jack\" Young, President Honorable Stephanie Rawlings-Blake, Mayor - ABSENT Harry Black, Director of Finance Honorable Joan M. Pratt, Comptroller and Secretary George A. Nilson, City Solicitor Alfred H. Foxx, Director of Public Works David E. Ralph, Deputy City Solicitor Rudolph S. Chow, Deputy Director of Public Works Bernice H. Taylor, Deputy Comptroller and Clerk Pursuant to Article VI, Section 1(c) of the r\n"
     ]
    }
   ],
   "source": [
    "# specifies the paths for a couple sample pdfs\n",
    "meeting1_path = Path(\"pdf_files/2013/2013_11_20.pdf\")\n",
    "meeting2_path = Path(\"pdf_files/2010/2010_03_17.pdf\")\n",
    "\n",
    "# uses parse_pdf() to instantiate the Minutes class for each meeting\n",
    "meeting1 = parse_pdf(meeting1_path)\n",
    "meeting2 = parse_pdf(meeting2_path)\n",
    "\n",
    "print(meeting1.clean_text[:500])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Process and prepare data\n",
    "### Create a Pandas dataframe with full texts  \n",
    "Now that we have the .pdf files we need, we're ready to read them and store their text in a Pandas dataframe. \n",
    "\n",
    "This should take about one minute for each year of data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "text_df_raw = store_pdf_text_to_df(pdf_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### View a sample of the stored text"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>date</th>\n",
       "      <th>page_number</th>\n",
       "      <th>minutes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>376</th>\n",
       "      <td>2020-10-14</td>\n",
       "      <td>3960</td>\n",
       "      <td>3960\\n \\nBOARD OF ESTIMATES\\n \\n \\nOCTOBER 14,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289</th>\n",
       "      <td>2010-07-14</td>\n",
       "      <td>2341</td>\n",
       "      <td>2341  BOARD OF ESTIMATES                      ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>315</th>\n",
       "      <td>2019-02-06</td>\n",
       "      <td>584</td>\n",
       "      <td>584\\n \\nBOARD OF \\nESTIMATES\\n \\nFEBRUARY 06, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2013-11-20</td>\n",
       "      <td>4680</td>\n",
       "      <td>4680  BOARD OF ESITMATES  November 20, 2013 MI...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>285</th>\n",
       "      <td>2010-09-29</td>\n",
       "      <td>3357</td>\n",
       "      <td>3357  BOARD OF ESTIMATES                      ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>187</th>\n",
       "      <td>2009-08-19</td>\n",
       "      <td>3094</td>\n",
       "      <td>3094  BOARD OF ESTIMATES                      ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           date page_number                                            minutes\n",
       "376  2020-10-14        3960  3960\\n \\nBOARD OF ESTIMATES\\n \\n \\nOCTOBER 14,...\n",
       "289  2010-07-14        2341  2341  BOARD OF ESTIMATES                      ...\n",
       "315  2019-02-06         584  584\\n \\nBOARD OF \\nESTIMATES\\n \\nFEBRUARY 06, ...\n",
       "0    2013-11-20        4680  4680  BOARD OF ESITMATES  November 20, 2013 MI...\n",
       "285  2010-09-29        3357  3357  BOARD OF ESTIMATES                      ...\n",
       "187  2009-08-19        3094  3094  BOARD OF ESTIMATES                      ..."
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "text_df_raw.sample(6, random_state=444)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Replace erroneous characters and consolidate white spaces\n",
    "Not sure if decision to transform all multiple white spaces to a single white space will work for the long term because we may need to use multiple spaces to detect certain fields.\n",
    "\n",
    "I hope we won't need to do that, though. Currently we're consolidating all multiple white spaces into just one white space."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "text_df = text_df_raw.copy()\n",
    "\n",
    "text_df['text'] = text_df['minutes'].apply(replace_chars)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4680  BOARD OF ESITMATES  November 20' 2013 MINUTES  REGULAR MEETING  Honorable Bernard C. \"Jack\" Young' President Honorable Stephanie Rawlings-Blake' Mayor - ABSENT Harry Black' Director of Finance Honorable Joan M. Pratt' Comptroller and Secretary George A. Nilson' City Solicitor \n",
      "Alfred H. Foxx' Director of Public Works David E. Ralph' Deputy City Solicitor Rudolph S. Chow' Deputy Director of Public Works \n",
      "Bernice H. Taylor' Deputy Comptroller and Clerk \n",
      "  Pursuant to Article VI' Section 1(c)\n"
     ]
    }
   ],
   "source": [
    "# view a sample of the transformed text\n",
    "print(text_df['text'][0][0:500])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "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>date</th>\n",
       "      <th>page_number</th>\n",
       "      <th>minutes</th>\n",
       "      <th>text</th>\n",
       "      <th>calendar_year</th>\n",
       "      <th>month</th>\n",
       "      <th>fiscal_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2013-11-20</td>\n",
       "      <td>4680</td>\n",
       "      <td>4680  BOARD OF ESITMATES  November 20, 2013 MI...</td>\n",
       "      <td>4680 BOARD OF ESITMATES November 20, 2013 MINU...</td>\n",
       "      <td>2013</td>\n",
       "      <td>11</td>\n",
       "      <td>2014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2013-01-30</td>\n",
       "      <td>280</td>\n",
       "      <td>280 BOARD OF ESTIMATES  JANUARY 30, 2013 MINUT...</td>\n",
       "      <td>280 BOARD OF ESTIMATES JANUARY 30, 2013 MINUTE...</td>\n",
       "      <td>2013</td>\n",
       "      <td>1</td>\n",
       "      <td>2013</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>3045</td>\n",
       "      <td>3045 BOARD OF ESTIMATES  August 14, 2013 MINUT...</td>\n",
       "      <td>3045 BOARD OF ESTIMATES August 14, 2013 MINUTE...</td>\n",
       "      <td>2013</td>\n",
       "      <td>8</td>\n",
       "      <td>2014</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        date page_number                                            minutes  \\\n",
       "0 2013-11-20        4680  4680  BOARD OF ESITMATES  November 20, 2013 MI...   \n",
       "1 2013-01-30         280  280 BOARD OF ESTIMATES  JANUARY 30, 2013 MINUT...   \n",
       "2 2013-08-14        3045  3045 BOARD OF ESTIMATES  August 14, 2013 MINUT...   \n",
       "\n",
       "                                                text  calendar_year  month  \\\n",
       "0  4680 BOARD OF ESITMATES November 20, 2013 MINU...           2013     11   \n",
       "1  280 BOARD OF ESTIMATES JANUARY 30, 2013 MINUTE...           2013      1   \n",
       "2  3045 BOARD OF ESTIMATES August 14, 2013 MINUTE...           2013      8   \n",
       "\n",
       "   fiscal_year  \n",
       "0         2014  \n",
       "1         2013  \n",
       "2         2014  "
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def add_fiscal_year(df):\n",
    "    df = df.copy()\n",
    "    df[\"calendar_year\"] = df[\"date\"].dt.year\n",
    "    df[\"month\"] = df[\"date\"].dt.month\n",
    "    c = pd.to_numeric(df[\"calendar_year\"])\n",
    "    df[\"fiscal_year\"] = np.where(df[\"month\"] >= 7, c + 1, c)\n",
    "    df[\"fiscal_year\"] = (pd.to_datetime(df[\"fiscal_year\"], format=\"%Y\")).dt.year\n",
    "    return df\n",
    "\n",
    "text_df['date'] = pd.to_datetime(text_df['date'])\n",
    "text_df = add_fiscal_year(text_df)\n",
    "\n",
    "text_df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Store the date of meeting as the index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>date</th>\n",
       "      <th>page_number</th>\n",
       "      <th>minutes</th>\n",
       "      <th>text</th>\n",
       "      <th>calendar_year</th>\n",
       "      <th>month</th>\n",
       "      <th>fiscal_year</th>\n",
       "      <th>word_count</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2013-11-20</th>\n",
       "      <td>2013-11-20</td>\n",
       "      <td>4680</td>\n",
       "      <td>4680  BOARD OF ESITMATES  November 20, 2013 MI...</td>\n",
       "      <td>4680 BOARD OF ESITMATES November 20, 2013 MINU...</td>\n",
       "      <td>2013</td>\n",
       "      <td>11</td>\n",
       "      <td>2014</td>\n",
       "      <td>22448</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-01-30</th>\n",
       "      <td>2013-01-30</td>\n",
       "      <td>280</td>\n",
       "      <td>280 BOARD OF ESTIMATES  JANUARY 30, 2013 MINUT...</td>\n",
       "      <td>280 BOARD OF ESTIMATES JANUARY 30, 2013 MINUTE...</td>\n",
       "      <td>2013</td>\n",
       "      <td>1</td>\n",
       "      <td>2013</td>\n",
       "      <td>11455</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013-08-14</th>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>3045</td>\n",
       "      <td>3045 BOARD OF ESTIMATES  August 14, 2013 MINUT...</td>\n",
       "      <td>3045 BOARD OF ESTIMATES August 14, 2013 MINUTE...</td>\n",
       "      <td>2013</td>\n",
       "      <td>8</td>\n",
       "      <td>2014</td>\n",
       "      <td>42186</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 date page_number  \\\n",
       "date                                \n",
       "2013-11-20 2013-11-20        4680   \n",
       "2013-01-30 2013-01-30         280   \n",
       "2013-08-14 2013-08-14        3045   \n",
       "\n",
       "                                                      minutes  \\\n",
       "date                                                            \n",
       "2013-11-20  4680  BOARD OF ESITMATES  November 20, 2013 MI...   \n",
       "2013-01-30  280 BOARD OF ESTIMATES  JANUARY 30, 2013 MINUT...   \n",
       "2013-08-14  3045 BOARD OF ESTIMATES  August 14, 2013 MINUT...   \n",
       "\n",
       "                                                         text  calendar_year  \\\n",
       "date                                                                           \n",
       "2013-11-20  4680 BOARD OF ESITMATES November 20, 2013 MINU...           2013   \n",
       "2013-01-30  280 BOARD OF ESTIMATES JANUARY 30, 2013 MINUTE...           2013   \n",
       "2013-08-14  3045 BOARD OF ESTIMATES August 14, 2013 MINUTE...           2013   \n",
       "\n",
       "            month  fiscal_year  word_count  \n",
       "date                                        \n",
       "2013-11-20     11         2014       22448  \n",
       "2013-01-30      1         2013       11455  \n",
       "2013-08-14      8         2014       42186  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "text_df = text_df.set_index('date', drop=False)\n",
    "text_df['word_count'] = text_df['text'].apply(lambda x: len(x.split()))\n",
    "\n",
    "text_df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Compute and visualize a few basic features\n",
    "Let's see how two very basic features change over time:\n",
    "\n",
    "- Meetings per year\n",
    "- Total word count of all meetings per year"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Filter to show only complete fiscal years"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "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>fiscal_year</th>\n",
       "      <th>minutes</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2010</td>\n",
       "      <td>44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2011</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012</td>\n",
       "      <td>46</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   fiscal_year  minutes\n",
       "1         2010       44\n",
       "2         2011       46\n",
       "3         2012       46"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "meetings_per_year = text_df[['minutes', 'fiscal_year']].groupby('fiscal_year').count().reset_index()\n",
    "\n",
    "cond_1 = meetings_per_year['fiscal_year'] < 2021\n",
    "cond_2 = meetings_per_year['fiscal_year'] > 2009\n",
    "\n",
    "meetings_per_year = meetings_per_year[cond_1 & cond_2]\n",
    "meetings_per_year.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.ticker as ticker\n",
    "\n",
    "plt.figure(figsize=(12,6))\n",
    "ax = sns.lineplot(x='fiscal_year', y='minutes', data=meetings_per_year)\n",
    "\n",
    "ax.set(title=\"Meetings Per Fiscal Year\", ylabel=\"Number of meetings\", xlabel = \"Fiscal year\", ylim=(35, 50))\n",
    "ax.yaxis.set_major_locator(ticker.MultipleLocator(5))\n",
    "\n",
    "sns.despine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "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>fiscal_year</th>\n",
       "      <th>word_count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2010</td>\n",
       "      <td>788800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2011</td>\n",
       "      <td>788521</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2012</td>\n",
       "      <td>903824</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   fiscal_year  word_count\n",
       "1         2010      788800\n",
       "2         2011      788521\n",
       "3         2012      903824"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "words_per_year = text_df.groupby('fiscal_year')['word_count'].sum().reset_index()\n",
    "\n",
    "words_per_year = words_per_year[cond_1 & cond_2]\n",
    "\n",
    "words_per_year.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,6))\n",
    "ax = sns.lineplot(x='fiscal_year', y='word_count', data=words_per_year)\n",
    "\n",
    "ax.set(title=\"Word Count Per Year\", ylabel=\"Words\", xlabel=\"Fiscal Year\", ylim=(400_000, 1_250_000))\n",
    "\n",
    "sns.despine()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tabulate data\n",
    "### Create empty dataframes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_dateframes():\n",
    "    meetings_df = pd.DataFrame(\n",
    "        columns=[\"date\", \"president\", \"mayor\", \"no_of_protests\", \"no_of_settlements\",]\n",
    "    )\n",
    "    agreements_df = pd.DataFrame(\n",
    "        columns=[\"date\", \"department\", \"contractor\", \"account_number\", \"agreement\",]\n",
    "    )\n",
    "    accounts_df = pd.DataFrame(\n",
    "        columns=[\"date\", \"account_number\",]\n",
    "    )\n",
    "    \n",
    "    return meetings_df, agreements_df, accounts_df\n",
    "\n",
    "agreements_df, meetings_df, accounts_df = create_dateframes()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test out account regex"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1"
      ]
     },
     "execution_count": 84,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "account_lookup = r\"\\d{4}-\\d{6}-\\d{4}-\\d{6}-\\d{6}\"\n",
    "sample_text = text_df['text']\n",
    "sample_text = 'blah blah 5000-585919-1191-719000-603096 blah blah'\n",
    "list_of_accounts = list(re.findall(re.compile(account_lookup), sample_text))\n",
    "\n",
    "len(list_of_accounts)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### List all account numbers for each meeting"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "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>date</th>\n",
       "      <th>page_number</th>\n",
       "      <th>minutes</th>\n",
       "      <th>text</th>\n",
       "      <th>accounts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2013-11-20</td>\n",
       "      <td>4680</td>\n",
       "      <td>4680  BOARD OF ESITMATES  November 20, 2013 MI...</td>\n",
       "      <td>4680 BOARD OF ESITMATES November 20, 2013 MINU...</td>\n",
       "      <td>[4000-496313-3573-591219-603051, 2089-208914-5...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2013-01-30</td>\n",
       "      <td>280</td>\n",
       "      <td>280 BOARD OF ESTIMATES  JANUARY 30, 2013 MINUT...</td>\n",
       "      <td>280 BOARD OF ESTIMATES JANUARY 30, 2013 MINUTE...</td>\n",
       "      <td>[9905-922004-9188-900000-709047, 4000-807513-6...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2013-08-14</td>\n",
       "      <td>3045</td>\n",
       "      <td>3045 BOARD OF ESTIMATES  August 14, 2013 MINUT...</td>\n",
       "      <td>3045 BOARD OF ESTIMATES August 14, 2013 MINUTE...</td>\n",
       "      <td>[2032-000000-1330-158400-605008, 2044-000000-1...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2013-08-28</td>\n",
       "      <td>3346</td>\n",
       "      <td>3346 BOARD OF ESTIMATES  AUGUST 28, 2013 MINUT...</td>\n",
       "      <td>3346 BOARD OF ESTIMATES AUGUST 28, 2013 MINUTE...</td>\n",
       "      <td>[5000-598014-2252-688000-600000, 5000-598314-2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2013-06-05</td>\n",
       "      <td>2018</td>\n",
       "      <td>2018 BOARD OF ESTIMATES  JUNE 05, 2013 MINUTES...</td>\n",
       "      <td>2018 BOARD OF ESTIMATES JUNE 05, 2013 MINUTES ...</td>\n",
       "      <td>[2036-000000-1752-175200-603070, 1001-000000-1...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         date page_number                                            minutes  \\\n",
       "0  2013-11-20        4680  4680  BOARD OF ESITMATES  November 20, 2013 MI...   \n",
       "1  2013-01-30         280  280 BOARD OF ESTIMATES  JANUARY 30, 2013 MINUT...   \n",
       "2  2013-08-14        3045  3045 BOARD OF ESTIMATES  August 14, 2013 MINUT...   \n",
       "3  2013-08-28        3346  3346 BOARD OF ESTIMATES  AUGUST 28, 2013 MINUT...   \n",
       "4  2013-06-05        2018  2018 BOARD OF ESTIMATES  JUNE 05, 2013 MINUTES...   \n",
       "\n",
       "                                                text  \\\n",
       "0  4680 BOARD OF ESITMATES November 20, 2013 MINU...   \n",
       "1  280 BOARD OF ESTIMATES JANUARY 30, 2013 MINUTE...   \n",
       "2  3045 BOARD OF ESTIMATES August 14, 2013 MINUTE...   \n",
       "3  3346 BOARD OF ESTIMATES AUGUST 28, 2013 MINUTE...   \n",
       "4  2018 BOARD OF ESTIMATES JUNE 05, 2013 MINUTES ...   \n",
       "\n",
       "                                            accounts  \n",
       "0  [4000-496313-3573-591219-603051, 2089-208914-5...  \n",
       "1  [9905-922004-9188-900000-709047, 4000-807513-6...  \n",
       "2  [2032-000000-1330-158400-605008, 2044-000000-1...  \n",
       "3  [5000-598014-2252-688000-600000, 5000-598314-2...  \n",
       "4  [2036-000000-1752-175200-603070, 1001-000000-1...  "
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def find_all_accounts(row):\n",
    "    account_lookup = r\"\\d{4}-\\d{6}-\\d{4}-\\d{6}-\\d{6}\"\n",
    "    meeting_date = row['date']\n",
    "    list_of_accounts = list(re.findall(re.compile(account_lookup), row['text']))\n",
    "    return list_of_accounts\n",
    "    \n",
    "\n",
    "text_df['accounts'] = text_df.apply(find_all_accounts, axis=1)\n",
    "text_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Unnest account numbers\n",
    "This takes all the account numbers that were stored together in a list and makes each one its own row."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "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>accounts</th>\n",
       "      <th>date</th>\n",
       "      <th>page_number</th>\n",
       "      <th>minutes</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>4000-496313-3573-591219-603051</td>\n",
       "      <td>2013-11-20</td>\n",
       "      <td>4680</td>\n",
       "      <td>4680  BOARD OF ESITMATES  November 20, 2013 MI...</td>\n",
       "      <td>4680 BOARD OF ESITMATES November 20, 2013 MINU...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2089-208914-5930-437923-603051</td>\n",
       "      <td>2013-11-20</td>\n",
       "      <td>4680</td>\n",
       "      <td>4680  BOARD OF ESITMATES  November 20, 2013 MI...</td>\n",
       "      <td>4680 BOARD OF ESITMATES November 20, 2013 MINU...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2089-208914-5930-426526-603051</td>\n",
       "      <td>2013-11-20</td>\n",
       "      <td>4680</td>\n",
       "      <td>4680  BOARD OF ESITMATES  November 20, 2013 MI...</td>\n",
       "      <td>4680 BOARD OF ESITMATES November 20, 2013 MINU...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         accounts        date page_number  \\\n",
       "0  4000-496313-3573-591219-603051  2013-11-20        4680   \n",
       "0  2089-208914-5930-437923-603051  2013-11-20        4680   \n",
       "0  2089-208914-5930-426526-603051  2013-11-20        4680   \n",
       "\n",
       "                                             minutes  \\\n",
       "0  4680  BOARD OF ESITMATES  November 20, 2013 MI...   \n",
       "0  4680  BOARD OF ESITMATES  November 20, 2013 MI...   \n",
       "0  4680  BOARD OF ESITMATES  November 20, 2013 MI...   \n",
       "\n",
       "                                                text  \n",
       "0  4680 BOARD OF ESITMATES November 20, 2013 MINU...  \n",
       "0  4680 BOARD OF ESITMATES November 20, 2013 MINU...  \n",
       "0  4680 BOARD OF ESITMATES November 20, 2013 MINU...  "
      ]
     },
     "execution_count": 86,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def unnesting(df, explode):\n",
    "    idx = df.index.repeat(df[explode[0]].str.len())\n",
    "    df1 = pd.concat(\n",
    "        [pd.DataFrame({x: np.concatenate(df[x].values)}) for x in explode], axis=1\n",
    "    )\n",
    "    df1.index = idx\n",
    "    return df1.join(df.drop(explode, 1), how=\"left\")\n",
    "\n",
    "        \n",
    "accounts_df = unnesting(text_df,['accounts'])\n",
    "\n",
    "accounts_df.head(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Count the account numbers mentioned per year\n",
    "Not sure if this is the right approach to dropping duplicate account numbers, but for now I'm dropping anything where the same account is mentioned in the same meeting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "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>fiscal_year</th>\n",
       "      <th>accounts</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2010</td>\n",
       "      <td>1669</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2011</td>\n",
       "      <td>1515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2012</td>\n",
       "      <td>1486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2013</td>\n",
       "      <td>1509</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2014</td>\n",
       "      <td>1182</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   fiscal_year  accounts\n",
       "0         2010      1669\n",
       "1         2011      1515\n",
       "2         2012      1486\n",
       "3         2013      1509\n",
       "4         2014      1182"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "accounts_df = accounts_df.drop_duplicates(subset=['fiscal_year', 'accounts'], keep='last')\n",
    "\n",
    "cond_1 = accounts_df['fiscal_year'] > 2009\n",
    "cond_2 = accounts_df['fiscal_year'] < 2021\n",
    "accounts_df = accounts_df[cond_1 & cond_2]\n",
    "accounts_per_year = accounts_df.groupby('fiscal_year')['accounts'].count().reset_index()\n",
    "\n",
    "accounts_per_year.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hmm. That's weird.\n",
    "Looks like accounts dropped off quickly in 2015."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,6))\n",
    "ax = sns.lineplot(x='fiscal_year', y='accounts', data=accounts_per_year)\n",
    "\n",
    "ax.set(title=\"Unique Account Numbers Per Year\", ylabel=\"Account Numbers\", xlabel=\"Fiscal Year\", ylim=(0, 2200))\n",
    "ax.xaxis.set_major_locator(ticker.MultipleLocator(1))\n",
    "sns.despine()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "#^.+?(?=(\\s–|-\\s).*Agreements)\n",
    "# department_lookup = r\"^.+?(?=–|-.*Agreements)\"\n",
    "# department_lookup = r\"(?<=MINUTES).+?(?=(\\s–|-\\s).*)\"\n",
    "# department_lookup = r\"(?<=MINUTES).+(?=(\\s-\\s))\""
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": true,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "239.549px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}