02_reasoning_behind_clean_raw_data.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Reasoning behind data cleaning\n",
    "\n",
    "## Summary\n",
    "\n",
    "This notebook explains the decisions made when cleaning SSNAP extract v2. To view the cleaning steps performed, please see \"01_clean_raw_data.ipynb\".\n",
    "\n",
    "## Aims\n",
    "\n",
    "(1) Explain decisions related to variables that were included in the clean dataset - for example:\n",
    "* Why and which results were considered likely to be errors/anomalous.\n",
    "* How decided on a threshold to identify results that were considered extreme or anomalous.\n",
    "\n",
    "(2) Explain why some variables from the SSNAP data were not included in the clean dataset."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Set up"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Import packages and functions\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import os\n",
    "import pandas as pd\n",
    "from dataclasses import dataclass\n",
    "\n",
    "# Linting\n",
    "%load_ext pycodestyle_magic\n",
    "%pycodestyle_on\n",
    "\n",
    "# Set the maximum number of columns and rows to 150\n",
    "pd.set_option('display.max_columns', 150)\n",
    "pd.set_option('display.max_rows', 150)\n",
    "\n",
    "\n",
    "# Set paths and filenames\n",
    "@dataclass(frozen=True)\n",
    "class Paths:\n",
    "    '''Singleton object for storing paths to data and files.'''\n",
    "\n",
    "    data_path = './output/'\n",
    "    data_filename = 'clean_samuel_ssnap_extract_v2.csv'\n",
    "    issues_filename = 'cleaning_issues_log.csv'\n",
    "    raw_path = './data'\n",
    "    raw_filename = 'SAMueL ssnap extract v2.csv'\n",
    "    notebook = '01'\n",
    "\n",
    "\n",
    "paths = Paths()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Load data\n",
    "data = pd.read_csv(os.path.join(paths.data_path, paths.data_filename))\n",
    "raw_data = pd.read_csv(os.path.join(paths.raw_path, paths.raw_filename),\n",
    "                       low_memory=False)\n",
    "issues = pd.read_csv(os.path.join(paths.data_path, paths.issues_filename))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def compare_raw_clean(raw_data, clean_data):\n",
    "    '''\n",
    "    Merges the raw and clean data using patient ID\n",
    "    Then presents counts for the remaining columns when ID is dropped\n",
    "    '''\n",
    "    compare = pd.merge(left=clean_data, left_on='id',\n",
    "                       right=raw_data, right_on='PatientId')\n",
    "    differences = (compare\n",
    "                   .drop(columns=['id', 'PatientId'], axis=1)\n",
    "                   .value_counts(dropna=False)\n",
    "                   .reset_index(name='count'))\n",
    "    return (differences.sort_values(differences.columns[0]))\n",
    "\n",
    "\n",
    "def mark_missing(df):\n",
    "    '''\n",
    "    Identifies NaN then marks each value as missing data or having data\n",
    "    '''\n",
    "    df_missing = (pd.DataFrame(df)\n",
    "                  .isnull()\n",
    "                  .apply(lambda x: x.map({True: 'Missing data',\n",
    "                                          False: 'Have data'})))\n",
    "    return (df_missing)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Included variables\n",
    "\n",
    "### Overview of dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(358993, 70)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show shape of the data\n",
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "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>Data type</th>\n",
       "      <th>Proportion complete</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>id</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>stroke_team</th>\n",
       "      <td>object</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>age</th>\n",
       "      <td>float64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>male</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>infarction</th>\n",
       "      <td>float64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onset_to_arrival_time</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.670144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onset_known</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precise_onset_known</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>onset_during_sleep</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>arrive_by_ambulance</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.999975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>call_to_ambulance_arrival_time</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.239782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ambulance_on_scene_time</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.239782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ambulance_travel_to_hospital_time</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.239782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ambulance_wait_time_at_hospital</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.239782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>month</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>year</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>weekday</th>\n",
       "      <td>object</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>arrival_time_3_hour_period</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>arrival_to_scan_time</th>\n",
       "      <td>float64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>scan_to_thrombolysis_time</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.114662</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombectomy</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>arrival_to_thrombectomy_time</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.012098</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>congestive_heart_failure</th>\n",
       "      <td>float64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>hypertension</th>\n",
       "      <td>float64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>atrial_fibrillation</th>\n",
       "      <td>float64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>diabetes</th>\n",
       "      <td>float64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>prior_stroke_tia</th>\n",
       "      <td>float64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>afib_antiplatelet</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.999997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>afib_anticoagulant</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.750268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>afib_vit_k_anticoagulant</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.999997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>afib_doac_anticoagulant</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.999997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>afib_heparin_anticoagulant</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.999997</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>new_afib_diagnosis</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.552183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>prior_disability</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>stroke_severity</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_complete</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_loc</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_loc_questions</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.970128</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_loc_commands</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.971038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_best_gaze</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.965949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_visual</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.959743</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_facial_palsy</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.970139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_motor_arm_left</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.970657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_motor_arm_right</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.970651</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_motor_leg_left</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.969905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_motor_leg_right</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.969813</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_limb_ataxia</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.957353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_sensory</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.960659</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_best_language</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.969428</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_dysarthria</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.965790</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nihss_arrival_extinction_inattention</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.964815</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>discharge_destination</th>\n",
       "      <td>object</td>\n",
       "      <td>0.999972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>death</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.999972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>discharge_disability</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.992844</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>disability_6_month</th>\n",
       "      <td>float64</td>\n",
       "      <td>0.304530</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_not_available</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_out_of_hours</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_scan_not_quick_enough</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_no_reason</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_haemorrhagic</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_time_window</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_comorbidity</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_medication</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_refusal</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_age</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_improving</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_too_mild_severe</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_time_unknown_wake_up</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>thrombolysis_no_but_other_medical</th>\n",
       "      <td>int64</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         Data type  Proportion complete\n",
       "id                                           int64             1.000000\n",
       "stroke_team                                 object             1.000000\n",
       "age                                        float64             1.000000\n",
       "male                                         int64             1.000000\n",
       "infarction                                 float64             1.000000\n",
       "onset_to_arrival_time                      float64             0.670144\n",
       "onset_known                                  int64             1.000000\n",
       "precise_onset_known                          int64             1.000000\n",
       "onset_during_sleep                           int64             1.000000\n",
       "arrive_by_ambulance                        float64             0.999975\n",
       "call_to_ambulance_arrival_time             float64             0.239782\n",
       "ambulance_on_scene_time                    float64             0.239782\n",
       "ambulance_travel_to_hospital_time          float64             0.239782\n",
       "ambulance_wait_time_at_hospital            float64             0.239782\n",
       "month                                        int64             1.000000\n",
       "year                                         int64             1.000000\n",
       "weekday                                     object             1.000000\n",
       "arrival_time_3_hour_period                   int64             1.000000\n",
       "arrival_to_scan_time                       float64             1.000000\n",
       "thrombolysis                                 int64             1.000000\n",
       "scan_to_thrombolysis_time                  float64             0.114662\n",
       "thrombectomy                                 int64             1.000000\n",
       "arrival_to_thrombectomy_time               float64             0.012098\n",
       "congestive_heart_failure                   float64             1.000000\n",
       "hypertension                               float64             1.000000\n",
       "atrial_fibrillation                        float64             1.000000\n",
       "diabetes                                   float64             1.000000\n",
       "prior_stroke_tia                           float64             1.000000\n",
       "afib_antiplatelet                          float64             0.999997\n",
       "afib_anticoagulant                         float64             0.750268\n",
       "afib_vit_k_anticoagulant                   float64             0.999997\n",
       "afib_doac_anticoagulant                    float64             0.999997\n",
       "afib_heparin_anticoagulant                 float64             0.999997\n",
       "new_afib_diagnosis                         float64             0.552183\n",
       "prior_disability                             int64             1.000000\n",
       "stroke_severity                              int64             1.000000\n",
       "nihss_complete                               int64             1.000000\n",
       "nihss_arrival_loc                            int64             1.000000\n",
       "nihss_arrival_loc_questions                float64             0.970128\n",
       "nihss_arrival_loc_commands                 float64             0.971038\n",
       "nihss_arrival_best_gaze                    float64             0.965949\n",
       "nihss_arrival_visual                       float64             0.959743\n",
       "nihss_arrival_facial_palsy                 float64             0.970139\n",
       "nihss_arrival_motor_arm_left               float64             0.970657\n",
       "nihss_arrival_motor_arm_right              float64             0.970651\n",
       "nihss_arrival_motor_leg_left               float64             0.969905\n",
       "nihss_arrival_motor_leg_right              float64             0.969813\n",
       "nihss_arrival_limb_ataxia                  float64             0.957353\n",
       "nihss_arrival_sensory                      float64             0.960659\n",
       "nihss_arrival_best_language                float64             0.969428\n",
       "nihss_arrival_dysarthria                   float64             0.965790\n",
       "nihss_arrival_extinction_inattention       float64             0.964815\n",
       "discharge_destination                       object             0.999972\n",
       "death                                      float64             0.999972\n",
       "discharge_disability                       float64             0.992844\n",
       "disability_6_month                         float64             0.304530\n",
       "thrombolysis_no_not_available                int64             1.000000\n",
       "thrombolysis_no_out_of_hours                 int64             1.000000\n",
       "thrombolysis_no_scan_not_quick_enough        int64             1.000000\n",
       "thrombolysis_no_no_reason                    int64             1.000000\n",
       "thrombolysis_no_but_haemorrhagic             int64             1.000000\n",
       "thrombolysis_no_but_time_window              int64             1.000000\n",
       "thrombolysis_no_but_comorbidity              int64             1.000000\n",
       "thrombolysis_no_but_medication               int64             1.000000\n",
       "thrombolysis_no_but_refusal                  int64             1.000000\n",
       "thrombolysis_no_but_age                      int64             1.000000\n",
       "thrombolysis_no_but_improving                int64             1.000000\n",
       "thrombolysis_no_but_too_mild_severe          int64             1.000000\n",
       "thrombolysis_no_but_time_unknown_wake_up     int64             1.000000\n",
       "thrombolysis_no_but_other_medical            int64             1.000000"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# List all columns and show their data types and proportion of completed data\n",
    "data_type_complete = pd.DataFrame(\n",
    "    {'Data type': data.dtypes,\n",
    "     'Proportion complete': data.count() / data.shape[0]})\n",
    "\n",
    "# Show all columns\n",
    "data_type_complete"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ID\n",
    "\n",
    "The patient ID from the SSNAP dataset is included, as this provides a consistent identifier between the raw data and different/cleaned versions of the data.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 26/05/2023).*\n",
    "\n",
    "### Stroke team\n",
    "\n",
    "Removed trailing whitespace, as Invicta Ward Kent and Canterbury Hospital has an extra space at the end.\n",
    "\n",
    "### Age\n",
    "\n",
    "The SSNAP data provides age and arrival time as categories. These were converted to numerical variables based on the middle value in each category. We do not have data on what the average age or arrival time was for each category, so this is an assumption. There is a chance that, for example, for ages 90 or over, the average age is actually lower than 92.5. However, in lieu of that information, this is considered a reasonable guess."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>age</th>\n",
       "      <th>raw_age</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>37.5</td>\n",
       "      <td>AgeUnder40</td>\n",
       "      <td>6110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>42.5</td>\n",
       "      <td>Age40to44</td>\n",
       "      <td>5072</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>47.5</td>\n",
       "      <td>Age45to49</td>\n",
       "      <td>9314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>52.5</td>\n",
       "      <td>Age50to54</td>\n",
       "      <td>15839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>57.5</td>\n",
       "      <td>Age55to59</td>\n",
       "      <td>21285</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>62.5</td>\n",
       "      <td>Age60to64</td>\n",
       "      <td>26881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>67.5</td>\n",
       "      <td>Age65to69</td>\n",
       "      <td>33609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>72.5</td>\n",
       "      <td>Age70to74</td>\n",
       "      <td>45959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>77.5</td>\n",
       "      <td>Age75to79</td>\n",
       "      <td>52542</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>82.5</td>\n",
       "      <td>Age80to84</td>\n",
       "      <td>56912</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>87.5</td>\n",
       "      <td>Age85to89</td>\n",
       "      <td>49194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>92.5</td>\n",
       "      <td>AgeOver90</td>\n",
       "      <td>36276</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     age     raw_age  count\n",
       "10  37.5  AgeUnder40   6110\n",
       "11  42.5   Age40to44   5072\n",
       "9   47.5   Age45to49   9314\n",
       "8   52.5   Age50to54  15839\n",
       "7   57.5   Age55to59  21285\n",
       "6   62.5   Age60to64  26881\n",
       "5   67.5   Age65to69  33609\n",
       "3   72.5   Age70to74  45959\n",
       "1   77.5   Age75to79  52542\n",
       "0   82.5   Age80to84  56912\n",
       "2   87.5   Age85to89  49194\n",
       "4   92.5   AgeOver90  36276"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert raw age to single column\n",
    "raw_ages = pd.DataFrame(raw_data['PatientId'])\n",
    "raw_ages['raw_age'] = raw_data[[\n",
    "    col for col in raw_data if col.startswith('Age')]].idxmax(1)\n",
    "\n",
    "# Merge with clean and present difference\n",
    "compare_raw_clean(raw_ages, data[['id', 'age']])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Gender\n",
    "\n",
    "No comment - as in SSNAP dataset.\n",
    "\n",
    "### Stroke type\n",
    "\n",
    "It is vital that we understand stroke type, so patients were missing stroke type (missing due to having not received a scan) were completely removed from the clean dataset.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 26/05/2023).*"
   ]
  },
  {
   "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>S2StrokeType</th>\n",
       "      <th>ArrivaltoBrainImagingMinutes</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Have data</td>\n",
       "      <td>Have data</td>\n",
       "      <td>358993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>Missing data</td>\n",
       "      <td>1388</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   S2StrokeType ArrivaltoBrainImagingMinutes   count\n",
       "0     Have data                    Have data  358993\n",
       "1  Missing data                 Missing data    1388"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show relationship between missing stoke type and missing arrival to scan time\n",
    "(mark_missing(raw_data[['S2StrokeType', 'ArrivaltoBrainImagingMinutes']])\n",
    " .value_counts()\n",
    " .reset_index(name='count'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Onset time\n",
    "\n",
    "This section provides details relating to:\n",
    "* Onset to arrival time\n",
    "* Onset known\n",
    "* Precise onset known\n",
    "* Onset during sleep"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Zero onset to arrival time**\n",
    "\n",
    "There are several individuals who have 0 minutes between onset and arrival, and it does not appear to be due to having not arrived by ambulance. Dataset itself only includes patients with out-of-hospital stroke onset, so onset out-of-hospital would not be a possible explanation. Hence, it is felt that we do not trust these times and replaced them as missing.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 26/05/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Frequency of onset to arrival = 0 in raw data...\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>S1ArriveByAmbulance</th>\n",
       "      <th>Count where onset to arrival = 0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>N</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Y</td>\n",
       "      <td>48</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  S1ArriveByAmbulance  Count where onset to arrival = 0\n",
       "0                   N                                11\n",
       "1                   Y                                48"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Frequency of onset to arrival = 0 in clean data...\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "False    358993\n",
       "Name: onset_to_arrival_time, dtype: int64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Count with 0 onset to arrival time in raw data, by whether they\n",
    "# arrived by ambulance\n",
    "print('Frequency of onset to arrival = 0 in raw data...')\n",
    "display(raw_data.groupby('S1ArriveByAmbulance')['OnsettoArrivalMinutes']\n",
    "        .apply(lambda x: (x == 0).sum())\n",
    "        .reset_index(name='Count where onset to arrival = 0'))\n",
    "\n",
    "# Demonstrating there are no times equal to 0 in clean data\n",
    "print('Frequency of onset to arrival = 0 in clean data...')\n",
    "display((data['onset_to_arrival_time'] == 0).value_counts())"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Unknown onset time**\n",
    "\n",
    "Problem: There are many cases where onset time is marked as \"unknown\" but an onset to arrival time was provided for every person in the dataset. Reasons that support lack of trust in data:\n",
    "* These typically appear to be the difference between midnight and the arrival time (e.g. of people with an onset to arrival of 1 to 2 hours, nearly all of them have an arrival time of 0000 to 3000).\n",
    "* Over 99% did not receive thrombolysis - but of those who received thrombolysis, the onset to arrival time was typically quite high and seems unusual (in the clean data, no-one has thrombolysis for an onset to arrival time greater than 12 hours).\n",
    "\n",
    "Decision: In these cases, onset to arrival time has been set as missing in the clean data...\n",
    "* Models (inc. outcome models) are typically requiring precise/best estimate (i.e. already dropping onset unknown).\n",
    "* Onset to arrival time is very important - e.g. models of onset to arrival time for haemorrhagic strokes.\n",
    "* These patients have not been removed entirely from the data, as this represents a very large proportion of people in the dataset, and could have valuable information - e.g. for descriptive analysis, to describe spread of stroke severity that arrive.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 26/05/2023).*\n",
    "\n",
    "There are no patients with onset during sleep who have a precise onset time - all are either best estimate or unknown - which is as we would anticipate."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of people with onset time not known in raw data\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "False    241138\n",
       "True     119243\n",
       "Name: S1OnsetTimeType, dtype: int64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Onset to arrival time when onset time not known in clean data\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "NaN    118416\n",
       "Name: onset_to_arrival_time, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Numbers with unknown onset time\n",
    "print('Number of people with onset time not known in raw data')\n",
    "display((raw_data['S1OnsetTimeType'] == 'NK').value_counts())\n",
    "\n",
    "print('Onset to arrival time when onset time not known in clean data')\n",
    "data.loc[data['onset_known'] == 0,\n",
    "         'onset_to_arrival_time'].value_counts(dropna=False)"
   ]
  },
  {
   "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>FirstArrivalTime</th>\n",
       "      <th>onset_to_arrival_hours</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>969</th>\n",
       "      <td>0900to1200</td>\n",
       "      <td>-12.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1111</th>\n",
       "      <td>0600to0900</td>\n",
       "      <td>-1.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>656</th>\n",
       "      <td>1800to2100</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>781</th>\n",
       "      <td>0300to0600</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>859</th>\n",
       "      <td>2100to2400</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>743</th>\n",
       "      <td>1200to1500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>474</th>\n",
       "      <td>0600to0900</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>0000to3000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>343</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>678</th>\n",
       "      <td>1500to1800</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>613</th>\n",
       "      <td>0900to1200</td>\n",
       "      <td>0.0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>455</th>\n",
       "      <td>0600to0900</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>462</th>\n",
       "      <td>1200to1500</td>\n",
       "      <td>1.0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>565</th>\n",
       "      <td>2100to2400</td>\n",
       "      <td>1.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>592</th>\n",
       "      <td>0300to0600</td>\n",
       "      <td>1.0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>825</th>\n",
       "      <td>1800to2100</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>694</th>\n",
       "      <td>1500to1800</td>\n",
       "      <td>1.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>0000to3000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>350</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>367</th>\n",
       "      <td>0900to1200</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>0000to3000</td>\n",
       "      <td>2.0</td>\n",
       "      <td>361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>526</th>\n",
       "      <td>0300to0600</td>\n",
       "      <td>2.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>742</th>\n",
       "      <td>1200to1500</td>\n",
       "      <td>2.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>858</th>\n",
       "      <td>2100to2400</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>800</th>\n",
       "      <td>1800to2100</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>453</th>\n",
       "      <td>0600to0900</td>\n",
       "      <td>2.0</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>288</th>\n",
       "      <td>0900to1200</td>\n",
       "      <td>2.0</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>278</th>\n",
       "      <td>0000to3000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>583</th>\n",
       "      <td>0600to0900</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>680</th>\n",
       "      <td>1500to1800</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>714</th>\n",
       "      <td>1200to1500</td>\n",
       "      <td>3.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>560</th>\n",
       "      <td>1800to2100</td>\n",
       "      <td>3.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>0300to0600</td>\n",
       "      <td>3.0</td>\n",
       "      <td>515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>406</th>\n",
       "      <td>0900to1200</td>\n",
       "      <td>3.0</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1146</th>\n",
       "      <td>1500to1800</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1110</th>\n",
       "      <td>0600to0900</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>700</th>\n",
       "      <td>0900to1200</td>\n",
       "      <td>4.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1068</th>\n",
       "      <td>0000to3000</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>0300to0600</td>\n",
       "      <td>4.0</td>\n",
       "      <td>654</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>502</th>\n",
       "      <td>1200to1500</td>\n",
       "      <td>4.0</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>527</th>\n",
       "      <td>1200to1500</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>517</th>\n",
       "      <td>1500to1800</td>\n",
       "      <td>5.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>732</th>\n",
       "      <td>0600to0900</td>\n",
       "      <td>5.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>970</th>\n",
       "      <td>0900to1200</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>806</th>\n",
       "      <td>1800to2100</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>0300to0600</td>\n",
       "      <td>5.0</td>\n",
       "      <td>820</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>713</th>\n",
       "      <td>1500to1800</td>\n",
       "      <td>6.0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>810</th>\n",
       "      <td>1800to2100</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>0600to0900</td>\n",
       "      <td>6.0</td>\n",
       "      <td>947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>234</th>\n",
       "      <td>0300to0600</td>\n",
       "      <td>6.0</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>971</th>\n",
       "      <td>0900to1200</td>\n",
       "      <td>6.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>506</th>\n",
       "      <td>1200to1500</td>\n",
       "      <td>6.0</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     FirstArrivalTime  onset_to_arrival_hours  count\n",
       "969        0900to1200                   -12.0      1\n",
       "1111       0600to0900                    -1.0      1\n",
       "656        1800to2100                     0.0      2\n",
       "781        0300to0600                     0.0      2\n",
       "859        2100to2400                     0.0      1\n",
       "743        1200to1500                     0.0      2\n",
       "474        0600to0900                     0.0      5\n",
       "68         0000to3000                     0.0    343\n",
       "678        1500to1800                     0.0      2\n",
       "613        0900to1200                     0.0      3\n",
       "455        0600to0900                     1.0      6\n",
       "462        1200to1500                     1.0      5\n",
       "565        2100to2400                     1.0      4\n",
       "592        0300to0600                     1.0      3\n",
       "825        1800to2100                     1.0      1\n",
       "694        1500to1800                     1.0      2\n",
       "67         0000to3000                     1.0    350\n",
       "367        0900to1200                     1.0      9\n",
       "64         0000to3000                     2.0    361\n",
       "526        0300to0600                     2.0      4\n",
       "742        1200to1500                     2.0      2\n",
       "858        2100to2400                     2.0      1\n",
       "800        1800to2100                     2.0      1\n",
       "453        0600to0900                     2.0      6\n",
       "288        0900to1200                     2.0     15\n",
       "278        0000to3000                     3.0     16\n",
       "583        0600to0900                     3.0      3\n",
       "680        1500to1800                     3.0      2\n",
       "714        1200to1500                     3.0      2\n",
       "560        1800to2100                     3.0      4\n",
       "54         0300to0600                     3.0    515\n",
       "406        0900to1200                     3.0      7\n",
       "1146       1500to1800                     4.0      1\n",
       "1110       0600to0900                     4.0      1\n",
       "700        0900to1200                     4.0      2\n",
       "1068       0000to3000                     4.0      1\n",
       "44         0300to0600                     4.0    654\n",
       "502        1200to1500                     4.0      5\n",
       "527        1200to1500                     5.0      4\n",
       "517        1500to1800                     5.0      4\n",
       "732        0600to0900                     5.0      2\n",
       "970        0900to1200                     5.0      1\n",
       "806        1800to2100                     5.0      1\n",
       "37         0300to0600                     5.0    820\n",
       "713        1500to1800                     6.0      2\n",
       "810        1800to2100                     6.0      1\n",
       "34         0600to0900                     6.0    947\n",
       "234        0300to0600                     6.0     26\n",
       "971        0900to1200                     6.0      1\n",
       "506        1200to1500                     6.0      4"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Relationship with midnight\n",
    "# Extract people with onset time not known\n",
    "unknown_time = raw_data.loc[\n",
    "    raw_data['S1OnsetTimeType'] == 'NK',\n",
    "    ['OnsettoArrivalMinutes', 'FirstArrivalTime', 'S2Thrombolysis']]\n",
    "\n",
    "# Convert onset to arrival to hours and round down\n",
    "unknown_time['onset_to_arrival_hours'] = (\n",
    "    np.floor(unknown_time['OnsettoArrivalMinutes']/60))\n",
    "\n",
    "(unknown_time[['FirstArrivalTime', 'onset_to_arrival_hours']]\n",
    " .value_counts()\n",
    " .reset_index(name='count')\n",
    " .sort_values(by='onset_to_arrival_hours')\n",
    " .head(50))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NB    118119\n",
      "Y        580\n",
      "N        544\n",
      "Name: S2Thrombolysis, dtype: int64\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAhYAAAHFCAYAAACuBbDPAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABBdElEQVR4nO3dd3hUZfrG8XuSkEZCgJBAEkISpSpBqlIUDEWkLRbWQhVWFwQLyrq7Noq7gsqK+hMRURd1EQQWC6LSsQIrVRDpEkDKooCEGiB5fn9wzWwmBRJ4IQG/n+uaC+adc877nPfMzLlz5pwZj5mZAAAAHAgo7gIAAMClg2ABAACcIVgAAABnCBYAAMAZggUAAHCGYAEAAJwhWAAAAGcIFgAAwBmCBQAAcIZgUQK89dZb8ng8vltQUJDi4uJ0xx13aOPGjcVdniQpOTlZd9111wXvNz09XR6PR2+99dZpp/vhhx80dOhQpaen53ns+uuvV+3atc9PgY55nwtLly51tsyhQ4fK4/E4W15Ohd0+RfHJJ5+oUaNGKl26tCpWrKgOHTrou+++K/T852MMcfaGDx+uDz/8sFDTHjlyREOHDtXnn3+e5zHvds3vNY6ShWBRgowfP16LFi3S3Llzdd9992n69Om69tprtX///uIuTR988IGefPLJ4i6jQD/88IOGDRvGm84FFhcXp0WLFqlDhw5OlrdkyRJ17txZFStW1AcffKDRo0crOjpaa9ascbJ8XHhFDRbDhg3LN1h06NBBixYtUlxcnNsC4VxQcReA/6ldu7YaNmwo6dRf2VlZWRoyZIg+/PBD9e7du1hrq1evXrH2X1xOnDjhO4qEvEJCQtS4cWNny/v3v/8tM9OUKVMUHh4uSfr973/vbPm4eMXExCgmJqa4y0AhcMSiBPOGjP/+979+7UuXLtXvfvc7lS9fXqGhoapXr56mTJmSZ/4dO3boj3/8oxITExUcHKz4+Hh16dLFb3kZGRn605/+pJSUFAUHByshIUEDBw7U4cOH/ZaV86OQn3/+WcHBwfkewVi3bp08Ho/+7//+z9e2e/du9e3bV5UrV1ZwcLBSUlI0bNgwnTx50m/enTt36rbbblNkZKSioqJ0++23a/fu3Wccp7feesu380lLS/N9pJT78PySJUt03XXXKTw8XJdddpmeeeYZZWdn+x7//PPP5fF49K9//UuDBg1SQkKCQkJCtGnTJknSP//5T1111VUKDQ1V+fLldfPNN2vt2rV+fdx1112KiIjQunXr1LZtW5UuXVpxcXF65plnJEmLFy/Wtddeq9KlS6t69ep6++23812n/fv3q3fv3ipfvrxKly6tTp066ccff8wzXWFqyu0Pf/iDypcvryNHjuR5rGXLlrryyit996dOnaprrrlGUVFRvnHr06eP7/H8Pgr5+eeffc+7kJAQxcTEqFmzZpo7d+5p65KkwMBAZWdnO/8IcNeuXWrQoIGqVavmW7Z3W23atEnt27dXRESEEhMTNWjQIGVmZvrNv2/fPvXv318JCQkKDg7WZZddpscff9xvut///vd+YydJnTp1ksfj0dSpU31ty5cvl8fj0ccffyzpf4f4FyxYoHvvvVcVKlRQdHS0brnlFu3cubNQ6zd9+nQ1adJE4eHhioyMVJs2bbRo0SK/abwfia1Zs0Z33nmnoqKiVLFiRfXp00cHDhzwm/ZM210q3HuHx+PR4cOH9fbbb/tel9dff32+65Cenu4LDsOGDfNN733fye+jEO/HnIsWLVLTpk0VFham5ORkjR8/XtKpj9Xq16+v8PBwpaamaubMmXn63bhxo7p27arY2FiFhISoVq1aeuWVVwo17iiAodiNHz/eJNmSJUv82kePHm2SbNq0ab62+fPnW3BwsF133XU2efJkmzlzpt11110mycaPH++b7qeffrK4uDirUKGCjRo1yubOnWuTJ0+2Pn362Nq1a83M7PDhw1a3bl2/aV566SWLioqyli1bWnZ2tm95SUlJ1qtXL9/9m2++2RITEy0rK8uv5j//+c8WHBxsv/zyi5mZ7dq1yxITEy0pKclee+01mzt3rv3tb3+zkJAQu+uuu3zzHTlyxGrVqmVRUVH28ssv26xZs+yBBx6wKlWq5Fm33Pbs2WPDhw83SfbKK6/YokWLbNGiRbZnzx4zM2vRooVFR0dbtWrVbOzYsTZnzhzr37+/SbK3337bt5wFCxaYJEtISLAuXbrY9OnTbcaMGbZ3717f8u+880775JNP7J133rHLLrvMoqKibMOGDb5l9OrVy4KDg61WrVr20ksv2Zw5c6x3794myR599FGrXr26vfnmmzZr1izr2LGjSbKlS5fmeS4kJiZanz597LPPPrNx48ZZbGysJSYm2v79+33TFramIUOGWM6X+nfffWeS7PXXX/cbxzVr1vjG0Mxs4cKF5vF47I477rBPP/3U5s+fb+PHj7cePXr45tmyZUue7dO2bVuLiYmxcePG2eeff24ffvihDR482N57770Ct6HX+vXrLTg42GrWrOnbfkWV+/W0evVqS0xMtCZNmtjPP//smy7ntvrHP/5hc+fOtcGDB5vH47Fhw4b5pjt69KjVqVPHSpcubf/4xz9s9uzZ9uSTT1pQUJC1b9/eN93YsWNNku3cudPMzE6cOGGRkZEWFhZm99xzj2+6Z5991oKCgiwjI8Ov3ssuu8zuv/9+mzVrlr3xxhtWrlw5S0tLO+P6vvvuuybJbrjhBvvwww9t8uTJ1qBBAwsODravvvrKN533eVCjRg0bPHiwzZkzx0aNGmUhISHWu3dv33SF2e6Ffe9YtGiRhYWFWfv27X2vyzVr1uS7HseOHbOZM2eaJPvDH/7gm37Tpk1+47RlyxbfPN7Xdo0aNfK8roYNG2apqak2adIk+/TTT61x48YWEhJiO3bs8M2/Zs0ai4qKstTUVHvnnXds9uzZNmjQIAsICLChQ4eeceyRP4JFCeB9wSxevNhOnDhhBw8etJkzZ1qlSpWsefPmduLECd+0NWvWtHr16vm1mZl17NjR4uLifDv6Pn36WKlSpeyHH34osN8RI0ZYQEBAnkDz73//2yTZp59+6mvLHSymT59ukmz27Nm+tpMnT1p8fLzdeuutvra+fftaRESEbd261a+Pf/zjHybJ9ybz6quvmiT76KOP/Ka75557zhgszMymTp1qkmzBggV5HmvRooVJsv/85z9+7VdccYW1bdvWd98bLJo3b+433f79+31vjjlt27bNQkJCrGvXrr62Xr165QmDJ06csJiYGJNky5cv97Xv3bvXAgMD7eGHH/a1eZ8LN998s19f33zzjUmyv//970WuKXew8I5J3bp1/druvfdeK1OmjB08eNDM/reNfv31VytIfsEiIiLCBg4cWOA8pzNu3DiLi4uz6Ohoq1Onji+gFkXOYDFnzhwrU6aMdenSxY4ePeo3nXdbTZkyxa+9ffv2VqNGDd99b2DIPd2zzz7r9xrYtGmTSbJ33nnHzMy+/vprk2R//vOfLSUlxTdfmzZtrGnTpnnq7d+/v9/yn3vuOZNku3btKnBds7KyLD4+3lJTU/1C/sGDBy02NtavH+/z4LnnnvNbRv/+/S00NNQXBgqz3Yvy3lG6dGm/947T+fnnn02SDRkyJM9jBQWL3OHc+7oKCwvzCxErV640SfZ///d/vra2bdta5cqV7cCBA3593XfffRYaGmr79u0rVN3wx0chJUjjxo1VqlQpRUZG6sYbb1S5cuX00Ucf+T7f37Rpk9atW6du3bpJkk6ePOm7tW/fXrt27dL69eslSZ999pnS0tJUq1atAvubMWOGateurbp16/otq23btvJ4PPmeQOXVrl07VapUyXfIUZJmzZqlnTt3+h0ynTFjhtLS0hQfH+/XR7t27SRJX3zxhSRpwYIFioyM1O9+9zu/frp27VqEESxYpUqVdPXVV/u11alTR1u3bs0z7a233up3f9GiRTp69Gieq2ISExPVsmVLzZs3z6/d4/Goffv2vvtBQUGqWrWq4uLi/M5VKV++vGJjY/OtwbuNvZo2baqkpCQtWLDgrGrK7cEHH9TKlSv1zTffSDp1WPtf//qXevXqpYiICElSo0aNJEm33XabpkyZoh07dpx2mV5XX3213nrrLf3973/X4sWLdeLEiULNN3XqVA0YMEAffvih5s2bp507d6p169bat2+fb5rWrVurRYsWhVre22+/rfbt2+vuu+/WlClTFBoammcaj8ejTp06+bXlfl7Mnz9fpUuXVpcuXfym8469d6wvv/xyJScn+z7ymTNnjlJTU9W9e3dt2bJFmzdvVmZmpr7++mu1bt06Ty25n/t16tSRpHyfH17r16/Xzp071aNHDwUE/O/tPCIiQrfeeqsWL16c5yOv/Po5duyY9uzZI6lw2/1c3jtci4uLU4MGDXz3va+runXrKj4+3tfufS/0juexY8c0b9483XzzzQoPD8/zfnrs2DEtXrz4gq3HpYRgUYK88847WrJkiebPn6++fftq7dq1uvPOO32Pe8+N+NOf/qRSpUr53fr37y9J+uWXXySd+py7cuXKp+3vv//9r1atWpVnWZGRkTIz37LyExQUpB49euiDDz7Qr7/+KunUZ6BxcXFq27atXx8ff/xxnj68n0V7+9i7d68qVqyYp59KlSqdadgKJTo6Ok9bSEiIjh49mqc991nne/fuzbddkuLj432Pe4WHh+fZiQUHB6t8+fJ55g8ODtaxY8fytOe33pUqVfL1VdSacuvcubOSk5N9nyW/9dZbOnz4sAYMGOCbpnnz5vrwww918uRJ9ezZU5UrV1bt2rU1adKk0y578uTJ6tWrl9544w01adJE5cuXV8+ePc94vszQoUPVtm1bXX311brqqqs0b948/fTTT2rdurX279+vw4cPa+nSpYW+AuW9995TWFiY7r777gIvt81vW4WEhPhtk71796pSpUp5lhEbG6ugoCC/sW7VqpUvaMydO1dt2rRRamqqKlasqLlz5+qbb77R0aNH8w0WuZ+jISEhkpTvczRnbVLBz4Ps7Ow8V5WdqZ/CbPdzee9wraDXVe724OBgSfJt27179+rkyZN6+eWX86yH9w+DC7kelxJOdS9BatWq5TthMy0tTVlZWXrjjTf073//W126dFGFChUkSY8++qhuueWWfJdRo0YNSafOoP7pp59O21+FChUUFhamf/7znwU+fjq9e/fWyJEj9d577+n222/X9OnTNXDgQAUGBvoto06dOnr66afzXYb3L4ro6Gh9++23eR4vzMmbruXegXjfiHft2pVn2p07d55xnM5Gfuu9e/duVa1a1UlNAQEBGjBggB577DE9//zzGjNmjFq1auV7/nh17txZnTt3VmZmphYvXqwRI0aoa9euSk5OVpMmTfJddoUKFfTiiy/qxRdf1LZt2zR9+nT99a9/1Z49e/I9ec5r8+bNqlu3ru9+nTp1tGDBArVs2VJt2rTRtddeq1KlSqlv376nXTevd999V08++aRatGih2bNn+y27KKKjo/Wf//xHZub33NizZ49OnjzpN9atWrXSm2++qW+//Vb/+c9/9MQTT0g6dVLsnDlztHXrVkVERDi7kuZMz4OAgACVK1euyMs903Y/1/eOkqBcuXIKDAxUjx49/AJ1TikpKRe4qksDRyxKsOeee07lypXT4MGDlZ2drRo1aqhatWr67rvv1LBhw3xvkZGRkk59VLFgwQLfRyP56dixozZv3qzo6Oh8l5WcnHza+mrVqqVrrrlG48eP18SJE5WZmZnnstiOHTvq+++/1+WXX55vH95gkZaWpoMHD2r69Ol+80+cOLFQY1WYv+7OVpMmTRQWFqYJEyb4tf/000+aP3++WrVq5bzPd9991+/+woULtXXrVt8Z9S5quvvuuxUcHKxu3bpp/fr1uu+++wqcNiQkRC1atNCzzz4rSVqxYkWh1qNKlSq677771KZNGy1fvvy009auXVtz5szxC1W1a9fWggULtHnzZr300ksaOXKkoqKiCtV3+fLlNXfuXNWqVUtpaWlnfVi7VatWOnToUJ7vYnjnnXd8j+ec1uPx6Mknn1RAQICaN28u6dRHOAsWLNCcOXPUvHlzlSpV6qxqya1GjRpKSEjQxIkTZWa+9sOHD2vatGm+K0XOVkHbvSjvHQUdGSyoP+n8vI5zCw8PV1pamlasWKE6derkux75HenEmXHEogQrV66cHn30Uf35z3/WxIkT1b17d7322mtq166d2rZtq7vuuksJCQnat2+f1q5dq+XLl/sua3vqqaf02WefqXnz5nrssceUmpqqX3/9VTNnztTDDz+smjVrauDAgZo2bZqaN2+uhx56SHXq1FF2dra2bdum2bNna9CgQbrmmmtOW2OfPn3Ut29f7dy5U02bNs3zF+9TTz2lOXPmqGnTpnrggQdUo0YNHTt2TOnp6fr00081duxYVa5cWT179tQLL7ygnj176umnn1a1atX06aefatasWYUaK+83a44bN06RkZEKDQ1VSkqKkzeGsmXL6sknn9Rjjz2mnj176s4779TevXs1bNgwhYaGasiQIefcR25Lly7V3Xffrd///vfavn27Hn/8cSUkJPg+8nJRU9myZdWzZ0+9+uqrSkpKynOuweDBg/XTTz+pVatWqly5sn799Ve99NJLKlWqVIHnORw4cEBpaWnq2rWratasqcjISC1ZskQzZ84s8Cib16hRo3TDDTeoSZMmeuSRR1SrVi3t2rVLU6ZM0ZEjR1SxYkWNGDFCbdu2LfSXJEVGRvr6btOmjaZPn660tLRCzevVs2dPvfLKK+rVq5fS09OVmpqqr7/+WsOHD1f79u39PtaIjY1V7dq1NXv2bKWlpfl26t5zRfbt26dRo0YVqf/TCQgI0HPPPadu3bqpY8eO6tu3rzIzMzVy5Ej9+uuvvsuci6Iw270o7x2pqan6/PPP9fHHHysuLk6RkZF53ie8IiMjlZSUpI8++kitWrVS+fLlVaFChTP+kXO2XnrpJV177bW67rrrdO+99yo5OVkHDx7Upk2b9PHHH2v+/Pnnpd9LXvGeOwqzgi83NTt1qVuVKlWsWrVqdvLkSTM7dbngbbfdZrGxsVaqVCmrVKmStWzZ0saOHes37/bt261Pnz5WqVIlK1WqlMXHx9ttt91m//3vf33THDp0yJ544gmrUaOGBQcH+y69euihh2z37t2+6XJfFeJ14MABCwsLy/fyRa+ff/7ZHnjgAUtJSbFSpUpZ+fLlrUGDBvb444/boUOHfNP99NNPduutt1pERIRFRkbarbfeagsXLizUVSFmZi+++KKlpKRYYGCg3zwtWrSwK6+8Ms/0vXr1sqSkJN9971UhU6dOzXf5b7zxhtWpU8c3Tp07d85z6VyvXr2sdOnSeeYtqIakpCTr0KGD7773uTB79mzr0aOHlS1b1nf1x8aNG8+qpvyuCvH6/PPPTZI988wzeR6bMWOGtWvXzhISEiw4ONhiY2Otffv2fpcw5r4q5NixY9avXz+rU6eOlSlTxsLCwqxGjRo2ZMgQO3z4cL415PTdd99Zly5dLCYmxoKCgqxy5crWu3dvW7Nmja1fv94qVKhgNWvW9Htu5pbf6ykzM9NuvfVWCw0NtU8++cTMCt5W+Y3X3r17rV+/fhYXF2dBQUGWlJRkjz76qB07dizP/A899JBJsqefftqvvVq1aibJVq1adcZ6zf73fMzvSqfcPvzwQ7vmmmssNDTUSpcuba1atbJvvvkm3/XKecltzv69V1sUZrubFf69Y+XKldasWTMLDw83SdaiRYvTrsvcuXOtXr16FhISYpJ87zsFXRVSmNeVlyQbMGCAX9uWLVusT58+lpCQYKVKlbKYmBhr2rSp7wosFJ3HLMfxMwC/KYMGDdKrr76q7du3c9gXgBN8FAL8Bi1evFgbNmzQmDFj1LdvX0IFAGc4YgH8Bnk8HoWHh6t9+/YaP36877srAOBcccQC+A3i7wkA5wuXmwIAAGcIFgAAwBmCBQAAcOaCn2ORnZ2tnTt3KjIyssDv7wcAACWLmengwYOKj4/3+9G73C54sNi5c6cSExMvdLcAAMCB7du3n/ZHLi94sPD+lsX27dtVpkyZC909AAA4CxkZGUpMTPTtxwtywYOF9+OPMmXKECwAALjInOk0Bk7eBAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4EFXcBv0UbN27UwYMHizyf5+QxhR7apmMRVWRBoeehslMiIyNVrVq187Z8AMCli2BxgW3cuFHVq1c/q3nrVQrQ8r4Rqv/aIa3Yne24Mn8bNmwgXAAAioxgcYF5j1RMmDBBtWrVKtK8Yb9ukL7sq3fffVdHy55dODmTtWvXqnv37md1RAUAAIJFMalVq5bq169ftJl2BkhfSrVq1pTi656XugAAOBecvAkAAJwhWAAAAGcIFgAAwBmCBQAAcIZgAQAAnCFYAAAAZwgWAADAGYIFAABwhmABAACcIVgAAABnLplgceTIES1fvlxHjhwp7lJwkeA5AwDuXTLBYt26dWrQoIHWrVtX3KXgIsFzBgDcu2SCBQAAKH4ECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOBBV3AUBxmTFjhiSpQYMGxVqHx+NRqVKldPz4cV9bWFiYQkJClJGRIY/Ho6CgICUnJ2vr1q3KzMyUmSkwMFBm5pvHzHzLysrKUqlSpRQREaGaNWtq+/btOnnypDIyMhQQEKBSpUopKChI2dnZioiIUFxcnL7//nsdO3ZMkhQUFKSsrCwFBwcrJiZGtWrVUrt27TRmzBjt379fpUqVUuvWrVW9enUdP35cEydO1J49exQeHq6yZcsqKytLkZGRCggIUGhoqDIzM5WamqqqVasqNTVVO3bs0JtvvqkDBw6oSpUqMjOtW7dO5cqVU4MGDRQYGKjAwEA1aNBA3377rVatWiWPx6OaNWvqyJEjSk9P1y+//KLs7GxFRUWpevXqSklJUVRUlFavXq1Dhw6pQoUK2rJliw4cOKD4+HjVqFFDK1eu1OHDhxUaGiqPx6OIiAjFx8eradOmSkxMVNOmTbVw4ULt2LFDu3fv1o4dOzR16lRlZmaqWrVq+vjjj7VixQrNnTtXS5cuVWhoqMxM+/btU0BAgDp37qwHH3xQwcHBvu2SlZWlzz//XJ9//rkk6frrr9f111+vrKwsjRkzRps3b9bll1+u/v37KzAwUF999ZV27dql6OhorV69Wps3b9ZPP/0kM9OuXbuUkJCg5s2b6/7771dwcLCysrJ888TFxem6665TYGDgWT8fC1pe7nbvWHnvX3PNNXrttdcKXJ/c83jX78cff5TH41G9evU0Y8YMHT58WNWrV9fIkSMVFhbm63fr1q2aNm2adu7cqXLlyunBBx9UWFiY9uzZ47fsHTt26Oeff1ZMTIwSEhJ03XXXSZJvG2RnZ6tcuXLav3+/JCk6OloxMTHau3dvnnm8tcfGxurEiROaMGGC0tPTVaVKFdWrV0/x8fG+6b1jlLOf6OhoVaxYMd9luthWhd12F5wV0RdffGEdO3a0uLg4k2QffPBBkeY/cOCASbIDBw4UtevTWrZsmUmyZcuWOV2ua+dU544VZkPKnPr3PLlYxvFcSeLGLc8tKCjonJfh8XjskUceMTOzadOmWWxsbJ5pwsPDLSAgwK8tICDAypQpU+h+AgICrHPnzpacnOzXnpycbNOmTTur18W0adPyXd4jjzySp/1MY5Xf+hR1fBs1apSn36Juu5iYmCKNq3ee/LZbQTfvGJ1unvyWeS7bqrDbztXyzQq//y7yRyGHDx/WVVddpdGjRxd1VqBE8Hg8xV3CRc3V+OX8q76kCAjI/y2xQoUKhZo/ODhYZqaRI0fqpptuUpcuXbRnzx5de+21mjdvnubNm6eEhAQdOXJE2dnZuvfee7Vr1y7de++9ys7OVkZGhmrVqiVJCg0N9Vu2d9y945adna2PPvpIgYGBWrRokQ4ePKhFixYpNTVVXbp00fvvv1+kdX///ffVpUsXpaam+i2vQoUKGjlypCpUqKBFixZpwoQJ8ng8io6OliR16NDBr6433njDb326deumCRMmSJJvHkmKiIjIt47ExERJp7bFkiVLtHPnTt9jCQkJuuyyy/Kdr3Tp0pKkevXq6fXXX1e7du0kST///LMyMjKUkJDgmzb32HrVq1fPN8+ePXvUrVu3fKeLjIz0ux8YGKiRI0dqz549qlmzpiSpcePGuuqqq3zTeJc5YsSIc95WuRW07Vwtv8jOJb1IHLEoKo5YFK933nnnnP8i5VbwrUKFCnnaPB5Pnracf0HGxMT4PdamTRvf/3P/VZ/7Vq5cuQIf83g8+fYtycLCwiwgIKDAx723gwcPWseOHS0wMDDfx2NjY619+/YWGhpqgYGBFhQU5Lc+YWFh1rFjR8vKyjIzs8zMTAsMDLTg4GALDw+35ORky8zMtOTkZOvQoYNv3tjYWPN4PL71DwgIsEOHDlnFihUtKCjIb5yDgoIsMzPT9xzPysqyTp06WUpKip08ebJQr4uTJ09acnKyderUyVertz0pKckqVqxoKSkpvlo7depkJ06csA4dOpgkq1ixomVmZlqnTp0sOTnZtz7eeqtUqWKdOnWyw4cPm8fjscDAQKtSpYq1a9fOtx4ZGRm+6du3b59nzNu3b2/Hjx+35ORkq1ixoiUlJVlwcLBJspCQEEtMTPTVefLkSTt+/LiFhoaax+OxmJgYCwoKsrCwMOvQoYMlJyf7jh5428LDw61KlSoWFhZm4eHh1r59e99REI/HY+3atfM9duzYMb/nRWBgoAUEBFhYWJhvXbOysiwrK8s6dOhgHo/H10/O7XI226qw287V8nMq7P77vJ9jkZmZqczMTN/9jIyM89LP0aNHJUlr1649L8t3xVuft96S5mIZx7PVs2fP4i7hkhYVFaVffvnFr81ynAfiVapUKd//r7/+ek2dOlVVqlTRtm3btHnzZt9j2dnZp+3vqquu8p27kFt+/XoV5vXXqFEjRURE6MYbb/Sdj5Pb008/rSuvvFKffvqpr827Pt5+Hn/8cd+RkDFjxigrK0uPPPKInnnmGaWnp2vMmDFKT0/XpEmTNGrUKE2dOlXx8fHas2ePbx2ys7O1ZMkSPfXUU+rbt6+uvPJKffHFF5KkkydPasyYMRo4cKCkU3/pP/roo2ratKm++uorXX/99Wdc16+++spXQ86jNt5zG8aNG6c//vGPfrUGBQXp8ssvlyT16NFDwcHBvn4ladKkSVq1apX69u2rbdu2afLkyRo3bpzMTFlZWdq2bZsaNWrk62vZsmW+9QsNDVVWVpYSEhK0Y8cOSdKTTz6pb775Runp6b56WrZsqfnz5yszM1Pbt2/3tX/11VeS5DtnKC0tTVOmTNHJkyfVvn17ffLJJxo0aJCef/55HT16VO3atdMnn3yibdu2+eqpWrWqb7uamTp27KjPPvtMkrRo0SI99thjvudFVlaWb3t719U7jt5l5+zHu13OZlsVdttJZ/dccOG8B4sRI0Zo2LBh57sbpaenS5K6d+9+3vtyIT09Xc2aNSvuMvK42MYRJUthf4I+547de0i8bt262rZtm++kusI4XXg4Vx07dpR06kTa003jPQTvlft+7dq1ff/3hqY//OEPeuaZZ/zaateu7RsL7w4xp127dvlqyr3eOcNYzj537dpVYO25l5271pzt3n5z1pqzTu/HEznnr127tqpUqeJ3/9133/Vbfs5tnXP9fv31V0n+H1nUrl1bH3/8sV89cXFxfsvztude75zbxLs9c36kkt82zr0Nck6Ts9b85ByHnPN5/5+zvqJuq9wK2nauln82znuwePTRR/Xwww/77mdkZPg+Q3MpOTlZkjRhwgTfZ5Ql0dq1a9W9e3dfvSXNxTKOZ6u4rwC51IWHhxdqurCwMF8IOXTokCRp5cqVkuR3xv6ZnM/zZWbMmKHBgwef9ujGjBkzdOWVV/q1HT582O/+999/r8aNG0uS7y/8N9980/e4t+3777/3jUV+5wDExcX5/kLOvd7eZeTs0ztPYXiny1lrznZvvzlrbdy4sa/OH3/80a9f7/9XrVrldz93neXKlct3/cqWLSvJf+f+/fff56kn987S2557vXNuE+/29Nacsy2n3Nsg5zTeq6gKknMcc87n/X/O+oq6rXIraNu5Wv5ZOZfPWyTOsSgqzrEoXpxjwTkWEudY5MY5FpxjURiF3X8TLC4wgkXxK+6d78V+O9POuLA3706hJN0Kqik6OrrI83fu3Nk3Vs2aNbM5c+bYnDlzLCEhwTdN3759bceOHdavXz9fW61atUyShYaG5jvuuWu8/PLLbeHChZaRkWELFy60Tp06mcfjKfJlhtOmTTOPx2OdOnXyW17Dhg1NkjVs2NAWLlxoEyZMMI/HYxUrVjRJvnDhrev111/3W5+uXbvahAkTTJJvHkkWERGR7xgmJiaa9L9QmXN94+PjLSUlJd/5oqKiTJLVq1fPXnvtNb/QIslv3HOPrfdWr149v/tdu3bNd7rIyMg828D7/5o1a5oka9y4sdWpUyfPvMOHDz/nbVXYbedq+V7nLVgcPHjQVqxYYStWrDBJNmrUKFuxYoVt3brVaWFFdbHsEAkWJUNx78C4lczbpfA9FikpKU6/xyIlJeWi/h6L2NjYIn+PRWxsbJG+x8I7RqebJ79lnsu2Kuy2K47vsSjyORZLly5VWlqa7773/IlevXrprbfeKurigGJhZnrqqac0ZMiQ4i6Fb968hL9585ZbblHnzp0vmm/e9Nab3/JGjBjBN2+e4Zs3R4wYUWzfvHm6bXeheczO42nV+cjIyFBUVJQOHDigMmXKOFvu8uXL1aBBAy1btkz169d3tlzXzqnOnSulcS2kP34hxdc9H+VdNOPowm9pXQHgXBV2/82PkAEAAGcIFgAAwBmCBQAAcIZgAQAAnCFYAAAAZwgWAADAGYIFAABwhmABAACcuWSCRc2aNbVs2TLVrFmzuEvBRYLnDAC4d95/Nv1CCQ8P59sTUSQ8ZwDAvUvmiAUAACh+BAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgzCXzld4XiyNHjkiSli9fXuR5w37doFqS1q5bp6O7sx1XdsratWvPy3IBAL8NBIsLbN26dZKke+65p8jz1qsUoOV9I9StWzetOE/BwisyMvK8Lh8AcGkiWFxgN910k6RTv6wZHh5epHk9J49p7aFterN9FVlQ6Hmo7pTIyEhVq1btvC0fAHDp8piZXcgOMzIyFBUVpQMHDqhMmTIXsmsAAHCWCrv/5uRNAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAgAAOEOwAAAAzhAsAACAMwQLAADgDMECAAA4Q7AAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4EzQhe7QzCRJGRkZF7prAABwlrz7be9+vCAXPFgcPHhQkpSYmHihuwYAAOfo4MGDioqKKvBxj50pejiWnZ2tnTt3KjIyUh6Px9lyMzIylJiYqO3bt6tMmTLOlnupYryKjjErOsas6BizomPMiu5sxszMdPDgQcXHxysgoOAzKS74EYuAgABVrlz5vC2/TJkyPLGKgPEqOsas6BizomPMio4xK7qijtnpjlR4cfImAABwhmABAACcuWSCRUhIiIYMGaKQkJDiLuWiwHgVHWNWdIxZ0TFmRceYFd35HLMLfvImAAC4dF0yRywAAEDxI1gAAABnCBYAAMAZggUAAHDmkggWY8aMUUpKikJDQ9WgQQN99dVXxV1SiTVixAg1atRIkZGRio2N1U033aT169cXd1kXlREjRsjj8WjgwIHFXUqJtmPHDnXv3l3R0dEKDw9X3bp1tWzZsuIuq8Q6efKknnjiCaWkpCgsLEyXXXaZnnrqKWVnZxd3aSXGl19+qU6dOik+Pl4ej0cffvih3+NmpqFDhyo+Pl5hYWG6/vrrtWbNmuIptoQ43ZidOHFCf/nLX5SamqrSpUsrPj5ePXv21M6dO8+pz4s+WEyePFkDBw7U448/rhUrVui6665Tu3bttG3btuIurUT64osvNGDAAC1evFhz5szRyZMndcMNN+jw4cPFXdpFYcmSJRo3bpzq1KlT3KWUaPv371ezZs1UqlQpffbZZ/rhhx/0/PPPq2zZssVdWon17LPPauzYsRo9erTWrl2r5557TiNHjtTLL79c3KWVGIcPH9ZVV12l0aNH5/v4c889p1GjRmn06NFasmSJKlWqpDZt2vh+o+q36HRjduTIES1fvlxPPvmkli9frvfff18bNmzQ7373u3Pr1C5yV199tfXr18+vrWbNmvbXv/61mCq6uOzZs8ck2RdffFHcpZR4Bw8etGrVqtmcOXOsRYsW9uCDDxZ3SSXWX/7yF7v22muLu4yLSocOHaxPnz5+bbfccot17969mCoq2STZBx984LufnZ1tlSpVsmeeecbXduzYMYuKirKxY8cWQ4UlT+4xy8+3335rkmzr1q1n3c9FfcTi+PHjWrZsmW644Qa/9htuuEELFy4spqouLgcOHJAklS9fvpgrKfkGDBigDh06qHXr1sVdSok3ffp0NWzYUL///e8VGxurevXq6fXXXy/uskq0a6+9VvPmzdOGDRskSd99952+/vprtW/fvpgruzhs2bJFu3fv9tsfhISEqEWLFuwPiuDAgQPyeDzndHTxgv8ImUu//PKLsrKyVLFiRb/2ihUravfu3cVU1cXDzPTwww/r2muvVe3atYu7nBLtvffe0/Lly7VkyZLiLuWi8OOPP+rVV1/Vww8/rMcee0zffvutHnjgAYWEhKhnz57FXV6J9Je//EUHDhxQzZo1FRgYqKysLD399NO68847i7u0i4L3PT+//cHWrVuLo6SLzrFjx/TXv/5VXbt2Pacfc7uog4VX7p9fNzOnP8l+qbrvvvu0atUqff3118VdSom2fft2Pfjgg5o9e7ZCQ0OLu5yLQnZ2tho2bKjhw4dLkurVq6c1a9bo1VdfJVgUYPLkyZowYYImTpyoK6+8UitXrtTAgQMVHx+vXr16FXd5Fw32B2fnxIkTuuOOO5Sdna0xY8ac07Iu6mBRoUIFBQYG5jk6sWfPnjypFf7uv/9+TZ8+XV9++eV5/Rn7S8GyZcu0Z88eNWjQwNeWlZWlL7/8UqNHj1ZmZqYCAwOLscKSJy4uTldccYVfW61atTRt2rRiqqjke+SRR/TXv/5Vd9xxhyQpNTVVW7du1YgRIwgWhVCpUiVJp45cxMXF+drZH5zZiRMndNttt2nLli2aP3/+Of/0/EV9jkVwcLAaNGigOXPm+LXPmTNHTZs2LaaqSjYz03333af3339f8+fPV0pKSnGXVOK1atVKq1ev1sqVK323hg0bqlu3blq5ciWhIh/NmjXLcxnzhg0blJSUVEwVlXxHjhxRQID/W3JgYCCXmxZSSkqKKlWq5Lc/OH78uL744gv2B6fhDRUbN27U3LlzFR0dfc7LvKiPWEjSww8/rB49eqhhw4Zq0qSJxo0bp23btqlfv37FXVqJNGDAAE2cOFEfffSRIiMjfUd7oqKiFBYWVszVlUyRkZF5zkEpXbq0oqOjOTelAA899JCaNm2q4cOH67bbbtO3336rcePGady4ccVdWonVqVMnPf3006pSpYquvPJKrVixQqNGjVKfPn2Ku7QS49ChQ9q0aZPv/pYtW7Ry5UqVL19eVapU0cCBAzV8+HBVq1ZN1apV0/DhwxUeHq6uXbsWY9XF63RjFh8fry5dumj58uWaMWOGsrKyfPuE8uXLKzg4+Ow6PevrSUqQV155xZKSkiw4ONjq16/PpZOnISnf2/jx44u7tIsKl5ue2ccff2y1a9e2kJAQq1mzpo0bN664SyrRMjIy7MEHH7QqVapYaGioXXbZZfb4449bZmZmcZdWYixYsCDf969evXqZ2alLTocMGWKVKlWykJAQa968ua1evbp4iy5mpxuzLVu2FLhPWLBgwVn3yc+mAwAAZy7qcywAAEDJQrAAAADOECwAAIAzBAsAAOAMwQIAADhDsAAAAM4QLAAAgDMECwAA4AzBAkAeb731lsqWLet0mXfddZduuummIs+XnJysF1980WkthXX8+HFVrVpV33zzjSQpPT1dHo9HK1euLJZ6Tmf16tWqXLmyDh8+XNyl4DeOYIFLyvbt2/WHP/xB8fHxCg4OVlJSkh588EHt3bv3gtYxdOhQ1a1b94zTne3O9ny7/fbbtWHDhgvaZ0FhZsmSJfrjH/94QWvxGjdunJKSktSsWbNi6b8oUlNTdfXVV+uFF14o7lLwG0ewwCXjxx9/VMOGDbVhwwZNmjRJmzZt0tixYzVv3jw1adJE+/btK+4SS5Tjx4/n237ixAmFhYUpNjb2AleUv5iYGIWHhxdL3y+//LLuvvvuYuk7txMnTpxxmt69e+vVV19VVlbWBagIKICrHzoBituNN95olStXtiNHjvi179q1y8LDw61fv36+tqSkJHv66aetd+/eFhERYYmJifbaa6/5Hs/MzLQBAwb4fswoKSnJhg8f7nv8119/tXvuucdiYmIsMjLS0tLSbOXKlWZmNn78+EL9yNuQIUMK/OGfVatWWVpamoWGhlr58uXtnnvusYMHDxa47idPnrQ+ffpYcnKyhYaGWvXq1e3FF1/0m6ZXr17WuXNnGz58uMXFxVlSUpLvR4gmT55sLVq0sJCQEPvnP/9p48ePt6ioKDMzW7dunUmytWvX+i3v+eeft6SkJMvOzi5S//nJ74eShgwZ4ttWL7zwgm9aSTZ27Fjr0KGDhYWFWc2aNW3hwoW2ceNGa9GihYWHh1vjxo1t06ZNfn1Mnz7d6tevbyEhIZaSkmJDhw61EydOFDimy5Yts4CAADtw4ICvzTte06ZNs+uvv97CwsKsTp06tnDhQr95//3vf9sVV1xhwcHBlpSUZP/4xz/8HpdkH3zwgV9bVFSU73lS0HZJT0+3jh07WtmyZS08PNyuuOIK++STT3zLyMzMtJCQEJs3b16B6wWcbwQLXBL27t1rHo/Hb+ef0z333GPlypWz7OxsMzu1sypfvry98sortnHjRhsxYoQFBAT4dp4jR460xMRE+/LLLy09Pd2++uormzhxopmd+gXFZs2aWadOnWzJkiW2YcMGGzRokEVHR9vevXvtyJEjNmjQILvyyitt165dtmvXrjxhx8zs4MGDdtttt9mNN97omy4zM9MOHz5s8fHxdsstt9jq1att3rx5lpKS4vsFx/wcP37cBg8ebN9++639+OOPNmHCBAsPD7fJkyf7punVq5dFRERYjx497Pvvv7fVq1f7dmDJyck2bdo0+/HHH23Hjh1+wcLMrEGDBvbEE0/49dmgQQN79NFHi9R/QcEiMzPTXnzxRStTpoxvLLxBKr9gkZCQYJMnT7b169fbTTfdZMnJydayZUubOXOm/fDDD9a4cWO78cYbffPMnDnTypQpY2+99ZZt3rzZZs+ebcnJyTZ06NACx/SFF16wmjVr+rV5x6tmzZo2Y8YMW79+vXXp0sWSkpJ8IWXp0qUWEBBgTz31lK1fv97Gjx9vYWFhfuGysMEi93bp0KGDtWnTxlatWmWbN2+2jz/+OM+vOV999dWnXS/gfCNY4JKwePHifN+svUaNGmWS7L///a+ZndpZde/e3fd4dna2xcbG2quvvmpmZvfff7+1bNnSF0RymjdvnpUpU8aOHTvm13755Zf7jnoMGTLErrrqqjPWnd/Odty4cVauXDk7dOiQr+2TTz6xgIAA27179xmX6dW/f3+79dZb/fqqWLGi389we3dguY8u5A4Wo0aNsssuu8x3f/369SbJ1qxZU6T+CwoW+fXplV+wyBlyFi1aZJLszTff9LVNmjTJQkNDffevu+66PKHzX//6l8XFxRVYz4MPPmgtW7b0a/OO1xtvvOFrW7Nmjd8Rna5du1qbNm385nvkkUfsiiuu8FuHwgSL3NslNTX1jKHh5ptvtrvuuuu00wDnE+dY4DfBzCRJHo/H11anTh3f/z0ejypVqqQ9e/ZIOnVS5cqVK1WjRg098MADmj17tm/aZcuW6dChQ4qOjlZERITvtmXLFm3evPmca127dq2uuuoqlS5d2tfWrFkzZWdna/369QXON3bsWDVs2FAxMTGKiIjQ66+/rm3btvlNk5qaquDg4DzzNmzY8LQ13XHHHdq6dasWL14sSXr33XdVt25dXXHFFUXq35Wc265ixYqSTq1bzrZjx44pIyND0qlt9tRTT/ltr3vuuUe7du3SkSNH8u3j6NGjCg0NPWP/cXFxkuR77qxduzbPyZ7NmjXTxo0bi3zuQ+7t8sADD+jvf/+7mjVrpiFDhmjVqlV55gkLCytwnYALgWCBS0LVqlXl8Xj0ww8/5Pv4unXrVK5cOVWoUMHXVqpUKb9pPB6PsrOzJUn169fXli1b9Le//U1Hjx7Vbbfdpi5dukiSsrOzFRcXp5UrV/rd1q9fr0ceeeSc18XM/AJQ7hrzM2XKFD300EPq06ePZs+erZUrV6p37955TtDMGVYK0+4VFxentLQ0TZw4UZI0adIkde/evcj9u5Jz23nHJL827/bMzs7WsGHD/LbX6tWrtXHjxgLDQ4UKFbR///5C9+/tK7/t5w22OefJ3ZbfyZm5t8vdd9+tH3/8UT169NDq1avVsGFDvfzyy37T7Nu3TzExMfnWDVwIBAtcEqKjo9WmTRuNGTNGR48e9Xts9+7devfdd3X77bcXuGPOT5kyZXT77bfr9ddf1+TJkzVt2jTt27dP9evX1+7duxUUFKSqVav63bzBJTg4uFB/neY33RVXXKGVK1f6fR/BN998o4CAAFWvXj3f5Xz11Vdq2rSp+vfvr3r16qlq1apOjp7k1K1bN02ePFmLFi3S5s2bdccddzjtv7Bjdjbq16+v9evX59leVatWVUBA/m+D9erV07p16/IEgDO54oor9PXXX/u1LVy4UNWrV1dgYKCkU1e67Nq1y/f4xo0bC32UITExUf369dP777+vQYMG6fXXX/d7/Pvvv1e9evWKVDPgEsECl4zRo0crMzNTbdu21Zdffqnt27dr5syZatOmjRISEvT0008XelkvvPCC3nvvPa1bt04bNmzQ1KlTValSJZUtW1atW7dWkyZNdNNNN2nWrFlKT0/XwoUL9cQTT2jp0qWSTn2p05YtW7Ry5Ur98ssvyszMzLef5ORkrVq1SuvXr9cvv/yiEydOqFu3bgoNDVWvXr30/fffa8GCBbr//vvVo0cP32H/3KpWraqlS5dq1qxZ2rBhg5588kktWbKk6IN4GrfccosyMjJ07733Ki0tTQkJCU77T05O1qFDhzRv3jz98ssvTg/nDx48WO+8846GDh2qNWvWaO3atZo8ebKeeOKJAudJS0vT4cOHtWbNmiL1NWjQIM2bN09/+9vftGHDBr399tsaPXq0/vSnP/mmadmypUaPHq3ly5dr6dKl6tevX54jaPkZOHCgZs2apS1btmj58uWaP3++atWq5Xs8PT1dO3bsUOvWrYtUM+ASwQKXjGrVqmnp0qW6/PLLdfvtt+vyyy/XH//4R6WlpWnRokUqX758oZcVERGhZ599Vg0bNlSjRo2Unp6uTz/9VAEBAfJ4PPr000/VvHlz9enTR9WrV9cdd9yh9PR0347/1ltv1Y033qi0tDTFxMRo0qRJ+fZzzz33qEaNGr5zE7755huFh4dr1qxZ2rdvnxo1aqQuXbqoVatWGj16dIH19uvXT7fccotuv/12XXPNNdq7d6/69+9ftAE8gzJlyqhTp0767rvv1K1bN+f9N23aVP369dPtt9+umJgYPffcc85qb9u2rWbMmKE5c+aoUaNGaty4sUaNGqWkpKQC54mOjtYtt9yid999t0h91a9fX1OmTNF7772n2rVra/DgwXrqqad01113+aZ5/vnnlZiYqObNm6tr167605/+VKjv6sjKytKAAQNUq1Yt3XjjjapRo4bGjBnje3zSpEm64YYbTrtewPnmsaIe5wOA34jVq1erdevW2rRpkyIjI4u7nNPKzMxUtWrVNGnSpIvim0Jx6SJYAMBpvP3226pfv77fVScl0YYNG7RgwQL17du3uEvBbxzBAgAAOMM5FgAAwBmCBQAAcIZgAQAAnCFYAAAAZwgWAADAGYIFAABwhmABAACcIVgAAABnCBYAAMCZ/wcdlJvidRvmHgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Number of patients with unknown time who received thrombolysis\n",
    "print(unknown_time['S2Thrombolysis'].value_counts(dropna=False))\n",
    "\n",
    "# Extract people who received thrombolysis\n",
    "onset = raw_data.loc[\n",
    "    raw_data['S2Thrombolysis'] == 'Y',\n",
    "    ['S1OnsetTimeType', 'OnsettoArrivalMinutes']]\n",
    "\n",
    "# Convert onset to arrival from minutes to hours\n",
    "onset['onset_to_arrival_hours'] = onset['OnsettoArrivalMinutes']/60\n",
    "\n",
    "# Create dataframes where onset time is known or not\n",
    "onset_known = onset.loc[onset['S1OnsetTimeType'] != 'NK',\n",
    "                        'onset_to_arrival_hours']\n",
    "onset_unknown = onset.loc[((onset['S1OnsetTimeType'] == 'NK') &\n",
    "                           (onset['onset_to_arrival_hours'] < 168)),\n",
    "                          'onset_to_arrival_hours']\n",
    "\n",
    "plt.boxplot(onset_known, vert=False)\n",
    "plt.xlabel('Onset to arrival time (hours)')\n",
    "plt.title('Received thrombolysis & known onset time')\n",
    "plt.show()\n",
    "\n",
    "plt.boxplot(onset_unknown, vert=False)\n",
    "plt.xlabel('Onset to arrival time (hours)')\n",
    "plt.title('Received thrombolysis & unknown onset time (hiding > 1 week)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "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>onset_during_sleep</th>\n",
       "      <th>onset_known</th>\n",
       "      <th>precise_onset_known</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>120661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>99647</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>89865</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>28551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>20269</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   onset_during_sleep  onset_known  precise_onset_known   count\n",
       "0                   0            1                    1  120661\n",
       "1                   0            1                    0   99647\n",
       "2                   0            0                    0   89865\n",
       "3                   1            0                    0   28551\n",
       "4                   1            1                    0   20269"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(data[['onset_during_sleep', 'onset_known', 'precise_onset_known']]\n",
    " .value_counts()\n",
    " .reset_index(name='count'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Long onset to arrival time**\n",
    "\n",
    "Problem: A large onset to arrival time is plausible, but some results are so large that they are considered to be invalid and likely errors in the data, or incredibly anomalous results.\n",
    "\n",
    "Context: The outcome model requires an onset-to-arrival time of less than a day.\n",
    "\n",
    "Decision: We have not implemented any measures to deal with this issue when cleaning the dataset, although we are aware that the most extreme values (e.g. 76 years and 118 years) are quite likely to be invalid. In reality, this data is not used in analysis, as analysis focusses only on people who have an onset to arrival within four hours.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 02/06/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Onset to arrival in years\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Extract onset to arrival time when onset time is known\n",
    "onset_arrival = raw_data.loc[\n",
    "    ((raw_data['S1OnsetTimeType'] == 'P') |\n",
    "     (raw_data['S1OnsetTimeType'] == 'BE')),\n",
    "    'OnsettoArrivalMinutes']\n",
    "\n",
    "# Years\n",
    "# Convert time from minutes to years then present counts (rounded)\n",
    "onset_arrival_years = onset_arrival/525600\n",
    "print('Onset to arrival in years')\n",
    "(round(onset_arrival_years).value_counts().sort_index())\n",
    "\n",
    "# Days\n",
    "onset_arrival_days = onset_arrival/1440\n",
    "onset_arrival_days_count = (round(onset_arrival_days[onset_arrival_days < 30])\n",
    "                            .value_counts()\n",
    "                            .sort_index()\n",
    "                            .reset_index(name='count'))\n",
    "fig = plt.figure()\n",
    "plt.bar(onset_arrival_days_count['index'], onset_arrival_days_count['count'])\n",
    "plt.xlabel('Days (rounded to nearest day)')\n",
    "plt.ylabel('Number of stroke patients')\n",
    "plt.title('Onset to arrival time (<30 days)')\n",
    "plt.show()\n",
    "\n",
    "# Hours\n",
    "onset_arrival_hours = onset_arrival/60\n",
    "onset_arrival_hours_count = (\n",
    "    round(onset_arrival_hours[onset_arrival_hours < 168])\n",
    "    .value_counts()\n",
    "    .sort_index()\n",
    "    .reset_index(name='count'))\n",
    "\n",
    "plt.bar(onset_arrival_hours_count['index'],\n",
    "        onset_arrival_hours_count['count'])\n",
    "plt.xlabel('Onset to arrival time in hours (rounded to nearest hour)')\n",
    "plt.ylabel('Number of stroke patients')\n",
    "plt.title('Onset to arrival time (<7 days/168 hours)')\n",
    "plt.show()\n",
    "\n",
    "onset_arrival_hours_shorter = onset_arrival_hours_count[\n",
    "    onset_arrival_hours_count['index'] < 72]\n",
    "plt.bar(onset_arrival_hours_shorter['index'],\n",
    "        onset_arrival_hours_shorter['count'])\n",
    "plt.xlabel('Onset to arrival time in hours (rounded to nearest hour)')\n",
    "plt.ylabel('Number of stroke patients')\n",
    "plt.title('Onset to arrival time (<3 days/72 hours)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Statistical detection of outliers**\n",
    "\n",
    "Examples of methods to detect outliers:\n",
    "* Z-score (but assumes normal distribution)\n",
    "* 1.5 * interquartile range (IQR) (but assumes normal distribution - as in a normal distribution, 0.28% of the data lies outside 3SD of the mean - and for the IQR method, 1.5 means 2.7 SD from the mean)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "21.708333333333332"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Classical method: 1.5 times\n",
    "q1 = onset_arrival_hours.quantile(0.25)\n",
    "q3 = onset_arrival_hours.quantile(0.75)\n",
    "iqr = q3 - q1\n",
    "upper = q3 + 1.5*iqr\n",
    "upper"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You could choose a threshold - for example, excluding the top 1% of times."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "45.96666666666667"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "onset_arrival_hours.quantile(0.95)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Applying a log transformation data could make it more symmetric, and maybe then you could look anew at outliers, but applying it to left skewed data would just make it more left skewed.\n",
    "\n",
    "Our main concern here is that, if you calculate the mean, outliers will have a large impact on the result. Some options of more robust estimators of location include:\n",
    "* Median\n",
    "* 5% trimmed average (mean after removing top 5% and bottom 5% of values)\n",
    "* Hodges-Lehmann estimator (computers median)\n",
    "* Interquartile mean (average of first and third quartiles)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Ambulance times\n",
    "\n",
    "In cases where any ambulance times are missing or inappropriate, all ambulance times are set as missing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Extract ambulance times\n",
    "raw_amb_times = raw_data[[\n",
    "    'CallConnectedtoArrivalMinutes',\n",
    "    'ArrivalPatientLocationtoArrivalMinutes',\n",
    "    'DeparturePatientLocationtoArrivalMinutes',\n",
    "    'WheelsStoptoArrivalMinutes']]\n",
    "clean_amb_times = data[[\n",
    "    'call_to_ambulance_arrival_time',\n",
    "    'ambulance_on_scene_time',\n",
    "    'ambulance_travel_to_hospital_time',\n",
    "    'ambulance_wait_time_at_hospital']]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Missing times**\n",
    "\n",
    "The clean dataset either requires all times to be complete or all to be missing.\n",
    "\n",
    "Although there are apx. 15,000 patients only missing WheelsStoptoArrivalMinutes, this means they are missing ambulance_travel_to_hospital_time and ambulance_wait_time_at_hospital, so this was not included. This also allows consistency in decision on what to keep when data on ambulance times is missing.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 26/05/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "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>call_to_ambulance_arrival_time</th>\n",
       "      <th>ambulance_on_scene_time</th>\n",
       "      <th>ambulance_travel_to_hospital_time</th>\n",
       "      <th>ambulance_wait_time_at_hospital</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>272913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>86080</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  call_to_ambulance_arrival_time ambulance_on_scene_time  \\\n",
       "0                   Missing time            Missing time   \n",
       "1                       Has time                Has time   \n",
       "\n",
       "  ambulance_travel_to_hospital_time ambulance_wait_time_at_hospital   count  \n",
       "0                      Missing time                    Missing time  272913  \n",
       "1                          Has time                        Has time   86080  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Presence of missing ambulance times in clean data\n",
    "(clean_amb_times\n",
    " .isnull()\n",
    " .apply(lambda x: x.map({True: 'Missing time', False: 'Has time'}))\n",
    " .value_counts()\n",
    " .reset_index(name='count'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "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>CallConnectedtoArrivalMinutes</th>\n",
       "      <th>ArrivalPatientLocationtoArrivalMinutes</th>\n",
       "      <th>DeparturePatientLocationtoArrivalMinutes</th>\n",
       "      <th>WheelsStoptoArrivalMinutes</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>251882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>92958</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>15421</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>87</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Has time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Has time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>Missing time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Missing time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Has time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Missing time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Has time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Missing time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>Missing time</td>\n",
       "      <td>Has time</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   CallConnectedtoArrivalMinutes ArrivalPatientLocationtoArrivalMinutes  \\\n",
       "0                   Missing time                           Missing time   \n",
       "1                       Has time                               Has time   \n",
       "2                       Has time                               Has time   \n",
       "3                       Has time                               Has time   \n",
       "4                       Has time                           Missing time   \n",
       "5                       Has time                           Missing time   \n",
       "6                   Missing time                               Has time   \n",
       "7                   Missing time                               Has time   \n",
       "8                       Has time                           Missing time   \n",
       "9                   Missing time                               Has time   \n",
       "10                      Has time                               Has time   \n",
       "11                  Missing time                               Has time   \n",
       "\n",
       "   DeparturePatientLocationtoArrivalMinutes WheelsStoptoArrivalMinutes   count  \n",
       "0                              Missing time               Missing time  251882  \n",
       "1                                  Has time                   Has time   92958  \n",
       "2                                  Has time               Missing time   15421  \n",
       "3                              Missing time               Missing time      87  \n",
       "4                              Missing time                   Has time       8  \n",
       "5                              Missing time               Missing time       7  \n",
       "6                                  Has time                   Has time       5  \n",
       "7                              Missing time               Missing time       5  \n",
       "8                                  Has time                   Has time       3  \n",
       "9                                  Has time               Missing time       3  \n",
       "10                             Missing time                   Has time       1  \n",
       "11                             Missing time                   Has time       1  "
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Missing ambulance times in raw data\n",
    "(raw_amb_times\n",
    " .isnull()\n",
    " .apply(lambda x: x.map({True: 'Missing time', False: 'Has time'}))\n",
    " .value_counts()\n",
    " .reset_index(name='count'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Negative times**\n",
    "\n",
    "We get negative times when:\n",
    "* A time is greater than the subsequent time in the sequence (e.g. call_to_ambulance_arrival is ArrivalPatient - CallConnected, but will be negative if CallConnected is larger than ArrivalPatient)\n",
    "* WheelsStoptoArrivalMinutes is positive (as calculated by 0 - WheelsStop)\n",
    "\n",
    "The clean dataset contains no negative times.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 26/05/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    360309\n",
       "True         72\n",
       "dtype: int64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(raw_data['CallConnectedtoArrivalMinutes'] >\n",
    " raw_data['ArrivalPatientLocationtoArrivalMinutes']).value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    360229\n",
       "True        152\n",
       "dtype: int64"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(raw_data['ArrivalPatientLocationtoArrivalMinutes'] >\n",
    " raw_data['DeparturePatientLocationtoArrivalMinutes']).value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    360011\n",
       "True        370\n",
       "dtype: int64"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(raw_data['DeparturePatientLocationtoArrivalMinutes'] >\n",
    " raw_data['WheelsStoptoArrivalMinutes']).value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    354343\n",
       "True       6038\n",
       "Name: WheelsStoptoArrivalMinutes, dtype: int64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(raw_data['WheelsStoptoArrivalMinutes'] > 0).value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>call_to_ambulance_arrival_time</th>\n",
       "      <th>ambulance_on_scene_time</th>\n",
       "      <th>ambulance_travel_to_hospital_time</th>\n",
       "      <th>ambulance_wait_time_at_hospital</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>358993</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   call_to_ambulance_arrival_time  ambulance_on_scene_time  \\\n",
       "0                           False                    False   \n",
       "\n",
       "   ambulance_travel_to_hospital_time  ambulance_wait_time_at_hospital   count  \n",
       "0                              False                            False  358993  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(clean_amb_times < 0).value_counts().reset_index(name='count')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Times equal to zero**\n",
    "\n",
    "It is plausible that the wait time at the hospital might be 0 minutes. However, a time of 0 minutes is not plausible for:\n",
    "* Call received to ambulance arrival\n",
    "* Length of time that ambulance is on scene\n",
    "* Travel time to hospital\n",
    "\n",
    "Hence, in those cases, all times are set as NaN.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 26/05/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>CallConnectedtoArrivalMinutes</th>\n",
       "      <th>ArrivalPatientLocationtoArrivalMinutes</th>\n",
       "      <th>DeparturePatientLocationtoArrivalMinutes</th>\n",
       "      <th>WheelsStoptoArrivalMinutes</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>349050</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>11158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>True</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   CallConnectedtoArrivalMinutes  ArrivalPatientLocationtoArrivalMinutes  \\\n",
       "0                          False                                   False   \n",
       "1                          False                                   False   \n",
       "2                          False                                   False   \n",
       "3                          False                                   False   \n",
       "4                          False                                    True   \n",
       "5                           True                                   False   \n",
       "\n",
       "   DeparturePatientLocationtoArrivalMinutes  WheelsStoptoArrivalMinutes  \\\n",
       "0                                     False                       False   \n",
       "1                                     False                        True   \n",
       "2                                      True                       False   \n",
       "3                                      True                        True   \n",
       "4                                     False                       False   \n",
       "5                                     False                       False   \n",
       "\n",
       "    count  \n",
       "0  349050  \n",
       "1   11158  \n",
       "2     118  \n",
       "3      19  \n",
       "4      19  \n",
       "5      17  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(raw_amb_times == 0).value_counts().reset_index(name='count')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>call_to_ambulance_arrival_time</th>\n",
       "      <th>ambulance_on_scene_time</th>\n",
       "      <th>ambulance_travel_to_hospital_time</th>\n",
       "      <th>ambulance_wait_time_at_hospital</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>347866</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>True</td>\n",
       "      <td>11127</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   call_to_ambulance_arrival_time  ambulance_on_scene_time  \\\n",
       "0                           False                    False   \n",
       "1                           False                    False   \n",
       "\n",
       "   ambulance_travel_to_hospital_time  ambulance_wait_time_at_hospital   count  \n",
       "0                              False                            False  347866  \n",
       "1                              False                             True   11127  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(clean_amb_times == 0).value_counts().reset_index(name='count')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Long call to ambulance time**\n",
    "\n",
    "Problem: Although a time at 98 hours may be a valid and real time that happened, it is very anomalous and is not what we would reasonably expect to happen in future - doesn't reflect normal variation in system - and so is considered appropriate to set that time as missing. There is one very high time (120 years) that is also definietely anomalous.\n",
    "\n",
    "Decision: Set a threshold of 24 hours (relatively arbritrary - based on spread of results and judgement on what is a very extreme and potentially implausible result). If a time is over that threshold, all ambulance times are set as missing.\n",
    "\n",
    "*(Threshold agreed between MA, KP, AL & AH on 02/06/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "0.0     74679\n",
       "1.0     26572\n",
       "2.0      4266\n",
       "3.0      1341\n",
       "4.0       638\n",
       "5.0       306\n",
       "6.0       196\n",
       "7.0        99\n",
       "8.0        86\n",
       "9.0        67\n",
       "10.0       30\n",
       "11.0       21\n",
       "12.0       20\n",
       "13.0       16\n",
       "14.0       15\n",
       "15.0        8\n",
       "16.0        4\n",
       "17.0        6\n",
       "18.0        4\n",
       "19.0        2\n",
       "20.0        1\n",
       "21.0        2\n",
       "23.0        3\n",
       "24.0        7\n",
       "25.0        2\n",
       "27.0        1\n",
       "48.0        1\n",
       "98.0        1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract call to ambulance arrival time from raw data\n",
    "# and convert to hours\n",
    "call_amb_hours = (\n",
    "    raw_data['ArrivalPatientLocationtoArrivalMinutes'] -\n",
    "    raw_data['CallConnectedtoArrivalMinutes'])/60\n",
    "\n",
    "# Drop NA, really high value (1061774), and negative\n",
    "call_amb_hours_filter = call_amb_hours[\n",
    "    (call_amb_hours < 100) & ~(call_amb_hours < 0)].dropna()\n",
    "\n",
    "plt.boxplot(call_amb_hours_filter, vert=0)\n",
    "plt.xlabel('Hours')\n",
    "plt.ylabel('Call to ambulance arrival time')\n",
    "plt.title('Call to ambulance arrival (hiding > 100 hours)')\n",
    "plt.show()\n",
    "\n",
    "round(call_amb_hours_filter).value_counts().sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1061773.9333333333"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Demonstration of max time in cleaned data\n",
    "np.nanmax(data['call_to_ambulance_arrival_time'])/60"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Long ambulance on scene time**\n",
    "\n",
    "Problem: In the raw data, time on scene increments from 0 to 6 hours, then leaps to 24, then two totally implausible, although those are now dropped due to other data cleaning actions.\n",
    "\n",
    "Decision: Cut-off at 12 hours (arbritrary - based on range of values and judgement of what is so long that it is likely to be invalid). Since these are times that are considered invalid, all ambulance times for these patients were set as missing (since times are incremental, so invalid times have knock-on effect) - rather than being dropped from the dataset entirely.\n",
    "\n",
    "*(Threshold agreed between MA, KP, AL & AH on 26/05/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0    40360\n",
       "1.0    44613\n",
       "2.0     1009\n",
       "3.0       80\n",
       "4.0       10\n",
       "5.0        4\n",
       "6.0        3\n",
       "7.0        1\n",
       "Name: ambulance_on_scene_time, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Convert time from minutes, filter from non-negative,\n",
    "# and present counts (rounded)\n",
    "amb_sc_hours = data['ambulance_on_scene_time']/60\n",
    "amb_sc_hours = amb_sc_hours[amb_sc_hours >= 0]\n",
    "round(amb_sc_hours).value_counts().sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0        51582\n",
       "1.0        55290\n",
       "2.0         1241\n",
       "3.0           94\n",
       "4.0           14\n",
       "5.0            4\n",
       "6.0            3\n",
       "7.0            1\n",
       "8.0            1\n",
       "10.0           2\n",
       "24.0           1\n",
       "8760.0         1\n",
       "17520.0        1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# From raw data\n",
    "raw_scene_time = (\n",
    "    raw_data['DeparturePatientLocationtoArrivalMinutes'] -\n",
    "    raw_data['ArrivalPatientLocationtoArrivalMinutes'])/60\n",
    "raw_scene_time = raw_scene_time[raw_scene_time >= 0]\n",
    "round(raw_scene_time).value_counts().sort_index()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Long ambulance travel to hospital time**\n",
    "\n",
    "Problem: Most times are around 0 or 1 hours, and some are around 2 to 5 hours - but then there are times of 10 to 24 hours, as well as some totally implausible times.\n",
    "\n",
    "Decision: Cut-off at 6 hours (arbritrary - based on range of values and decision on what times are so large that they are likely to be invalid/erronaeous). Setting times as missing (rather than dropping entirely) as this is on basis of them being invalid (rather than extreme but plausible).\n",
    "\n",
    "*(Threshold agreed between MA, KP, AL & AH on 02/06/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0          78514\n",
       "1.0          13997\n",
       "2.0             57\n",
       "3.0              4\n",
       "4.0              1\n",
       "5.0              2\n",
       "10.0             6\n",
       "11.0             2\n",
       "14.0             1\n",
       "23.0             1\n",
       "24.0             5\n",
       "240.0            1\n",
       "1488.0           1\n",
       "8760.0           1\n",
       "1048835.0        1\n",
       "1052038.0        1\n",
       "1060339.0        1\n",
       "dtype: int64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# From raw data\n",
    "raw_travel_time = (\n",
    "    raw_data['WheelsStoptoArrivalMinutes'] -\n",
    "    raw_data['DeparturePatientLocationtoArrivalMinutes'])/60\n",
    "raw_travel_time = raw_travel_time[raw_travel_time >= 0]\n",
    "round(raw_travel_time).value_counts().sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "24.216666666666665"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Demonstration of max time in cleaned data\n",
    "np.nanmax(data['ambulance_travel_to_hospital_time'])/60"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Long ambulance wait at hospital time**\n",
    "\n",
    "Problem: Majority of patients have quite a short wait time, but there are wait times that increment up and up in small steps (e.g. from a few hours through to 14, 15, 16 ..20, 22, 23... 43...). Many of these times are implausible (including times where it was weeks)\n",
    "\n",
    "Decision: Set threshold of 12 hours (arbritrary - based on evaluation of range of times, and judgement on what is likely to be invalid).\n",
    "\n",
    "*(Threshold agreed between MA, KP, AL & AH on 02/06/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Find ambulance wait time at hospital in hours\n",
    "amb_wait_hours = (0 - raw_data['WheelsStoptoArrivalMinutes'])/60\n",
    "\n",
    "# Remove negative\n",
    "amb_wait_hours = amb_wait_hours[amb_wait_hours >= 0]\n",
    "\n",
    "plt.boxplot(amb_wait_hours[amb_wait_hours <= 100], vert=0)\n",
    "plt.xlabel('Hours')\n",
    "plt.ylabel('Ambulance wait time at hospital')\n",
    "plt.title('Ambulance hospital wait time (hiding > 100 hours)')\n",
    "plt.show()\n",
    "\n",
    "plt.boxplot(amb_wait_hours[amb_wait_hours <= 24], vert=0)\n",
    "plt.xlabel('Hours')\n",
    "plt.ylabel('Ambulance wait time at hospital')\n",
    "plt.title('Ambulance hospital wait time (hiding > 24 hours)')\n",
    "plt.show()\n",
    "\n",
    "plt.boxplot(amb_wait_hours[amb_wait_hours <= 12], vert=0)\n",
    "plt.xlabel('Hours')\n",
    "plt.ylabel('Ambulance wait time at hospital')\n",
    "plt.title('Ambulance hospital wait time (hiding > 12 hours)')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8784.066666666668"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Max wait time in cleaned data (hours)\n",
    "np.nanmax(data['ambulance_wait_time_at_hospital'])/60"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Times when didn't arrive by ambulance**\n",
    "\n",
    "There are some individuals who have ambulance times yet are marked as having not arrived by ambulance. In these cases, we are not certain which of these times are true. Hence, all times have been marked as NaN for these individuals.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 26/05/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    360372\n",
       "True          9\n",
       "dtype: int64"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Number of times this is the case in the raw data\n",
    "times_but_no_amb = (\n",
    "    raw_data['S1ArriveByAmbulance'] == 'N') & (\n",
    "        (raw_data['CallConnectedtoArrivalMinutes'].notnull()) |\n",
    "        (raw_data['ArrivalPatientLocationtoArrivalMinutes'].notnull()) |\n",
    "        (raw_data['DeparturePatientLocationtoArrivalMinutes'].notnull()) |\n",
    "        (raw_data['WheelsStoptoArrivalMinutes'].notnull()))\n",
    "times_but_no_amb.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    358993\n",
       "dtype: int64"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Demonstrating these are removed from the clean data\n",
    "clean_times_but_no_amb = (\n",
    "    data['arrive_by_ambulance'] == 0) & (\n",
    "        (data['call_to_ambulance_arrival_time'].notnull()) |\n",
    "        (data['ambulance_on_scene_time'].notnull()) |\n",
    "        (data['ambulance_travel_to_hospital_time'].notnull()) |\n",
    "        (data['ambulance_wait_time_at_hospital'].notnull()))\n",
    "clean_times_but_no_amb.value_counts()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Comparing onset to arrival time to ambulance times**\n",
    "\n",
    "**What?** There are some cases where call-to-wheels-stop and call-to-arrival is greater than onset-to-arrival. This is not due to estimate type (all onset known, apx. half best estimate and half precise estimate).\n",
    "\n",
    "**Why?** The ambulance crew have to trigger when wheels stop occur (so these differences may occur because they didn't trigger arrival). The arrival time for onset to arrival is record when the patient arrives at the hospital. The use of ambulance times if just being used for indicative stuff like checking travel times.\n",
    "\n",
    "For now, we are not doing anything when cleaning with this, and are viewing it as noise in the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    57182\n",
       "True      2438\n",
       "Name: onset_vs_wheels, dtype: int64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "count    2438.000000\n",
       "mean       30.680066\n",
       "std        68.396046\n",
       "min         1.000000\n",
       "25%         5.000000\n",
       "50%        14.000000\n",
       "75%        37.000000\n",
       "max      1428.000000\n",
       "Name: onset_vs_wheels, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "False    55231\n",
       "True      4389\n",
       "Name: onset_vs_arrival, dtype: int64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "count      4389.000000\n",
       "mean        474.134655\n",
       "std        9365.259832\n",
       "min           1.000000\n",
       "25%           3.000000\n",
       "50%           9.000000\n",
       "75%          31.000000\n",
       "max      527012.000000\n",
       "Name: onset_vs_arrival, dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Select relevant columns and drop NA\n",
    "check_times = data[['call_to_ambulance_arrival_time',\n",
    "                    'ambulance_on_scene_time',\n",
    "                    'ambulance_travel_to_hospital_time',\n",
    "                    'ambulance_wait_time_at_hospital',\n",
    "                    'onset_to_arrival_time',\n",
    "                    'onset_known',\n",
    "                    'precise_onset_known']].dropna()\n",
    "\n",
    "# Find total time from call_to_hospital\n",
    "check_times['call_to_wheels_stop'] = check_times[[\n",
    "    'call_to_ambulance_arrival_time',\n",
    "    'ambulance_on_scene_time',\n",
    "    'ambulance_travel_to_hospital_time']].sum(axis=1)\n",
    "check_times['call_to_arrival'] = check_times[[\n",
    "    'call_to_ambulance_arrival_time',\n",
    "    'ambulance_on_scene_time',\n",
    "    'ambulance_travel_to_hospital_time',\n",
    "    'ambulance_wait_time_at_hospital']].sum(axis=1)\n",
    "\n",
    "# Find difference between onset to arrival time and call to wheels stop\n",
    "# and describe those greater than 0\n",
    "check_times['onset_vs_wheels'] = (\n",
    "    check_times['call_to_wheels_stop'] - check_times['onset_to_arrival_time'])\n",
    "wheels_greater = check_times['onset_vs_wheels'] > 0\n",
    "display(wheels_greater.value_counts())\n",
    "display(check_times.loc[wheels_greater, 'onset_vs_wheels'].describe())\n",
    "\n",
    "# Repeat for call to arrival\n",
    "check_times['onset_vs_arrival'] = (\n",
    "    check_times['call_to_arrival'] - check_times['onset_to_arrival_time'])\n",
    "wheels_greater = check_times['onset_vs_arrival'] > 0\n",
    "display(wheels_greater.value_counts())\n",
    "display(check_times.loc[wheels_greater, 'onset_vs_arrival'].describe())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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>onset_known</th>\n",
       "      <th>precise_onset_known</th>\n",
       "      <th>wheels_greater</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>False</td>\n",
       "      <td>29190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>False</td>\n",
       "      <td>27992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>True</td>\n",
       "      <td>1538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>True</td>\n",
       "      <td>900</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   onset_known  precise_onset_known  wheels_greater  count\n",
       "0            1                    1           False  29190\n",
       "1            1                    0           False  27992\n",
       "2            1                    1            True   1538\n",
       "3            1                    0            True    900"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Check if this is due to being \"best estimate\" (rather than precise)\n",
    "check_times['wheels_greater'] = check_times['onset_vs_wheels'] > 0\n",
    "(check_times[['onset_known', 'precise_onset_known', 'wheels_greater']]\n",
    " .value_counts()\n",
    " .reset_index(name='count'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Day, month, year and time of arrival\n",
    "\n",
    "No comment for day, month and year (just pulled from dataset).\n",
    "\n",
    "For time of arrival, to convert from a categorical to numeric variable, we set arrival time to the start of each bracket. It is not a concern that this is different to age (where set to middle of the bracket), as the meaningful information for arrival time is the difference between each time (i.e. 3 units), which would be the same regardless of whether the start, middle or end of the bracket was used. The start was chosen as this was felt easier to understand than a half time, and as this is consistent with SAMueL1.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 19/05/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "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>arrival_time_3_hour_period</th>\n",
       "      <th>FirstArrivalTime</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0</td>\n",
       "      <td>0000to3000</td>\n",
       "      <td>15628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>3</td>\n",
       "      <td>0300to0600</td>\n",
       "      <td>11481</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>6</td>\n",
       "      <td>0600to0900</td>\n",
       "      <td>24980</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>9</td>\n",
       "      <td>0900to1200</td>\n",
       "      <td>81705</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>12</td>\n",
       "      <td>1200to1500</td>\n",
       "      <td>78339</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>15</td>\n",
       "      <td>1500to1800</td>\n",
       "      <td>64213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>18</td>\n",
       "      <td>1800to2100</td>\n",
       "      <td>50236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>21</td>\n",
       "      <td>2100to2400</td>\n",
       "      <td>32411</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   arrival_time_3_hour_period FirstArrivalTime  count\n",
       "6                           0       0000to3000  15628\n",
       "7                           3       0300to0600  11481\n",
       "5                           6       0600to0900  24980\n",
       "0                           9       0900to1200  81705\n",
       "1                          12       1200to1500  78339\n",
       "2                          15       1500to1800  64213\n",
       "3                          18       1800to2100  50236\n",
       "4                          21       2100to2400  32411"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Extract relevant columns and merge\n",
    "compare_time = pd.merge(\n",
    "    left=data[['id', 'arrival_time_3_hour_period']], left_on='id',\n",
    "    right=raw_data[['PatientId', 'FirstArrivalTime']], right_on='PatientId')\n",
    "\n",
    "# Table illustrating change from categorical to numerical variable\n",
    "(compare_time\n",
    " .drop(['PatientId', 'id'], axis=1)\n",
    " .value_counts(dropna=False)\n",
    " .reset_index(name='count')\n",
    " .sort_values(by='arrival_time_3_hour_period'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Scan, thrombolysis and thrombectomy"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Arrival to scan time**\n",
    "\n",
    "Problem: Can be very long (e.g. 2000 hours = 83 days, 8000 hours = 333 days).\n",
    "\n",
    "Context:\n",
    "* We only request out-of-hospital onset, so this is not due to in-hospital stroke where arrived at hospital prior to having had a stroke.\n",
    "* However, it might be the case that people had a stroke out-of-hospital, arrived (e.g. by ambulance) for another reason, and only later realised had a stroke (or there was something more severe being prioritised).\n",
    "* It shouldn't really be more than four hours - and it should hopefully be a matter of minutes. Often, we filter to onset-to-arrival < 4 hours and arrival-to-scan < 4 hours.\n",
    "\n",
    "Decision: As with onset to arrival, we have not implemented any measures to deal with this issue when cleaning the dataset. In reality, this data is not used in analysis, as analysis focusses only on people who have an arrival to scan time of within four hours.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 02/06/2023).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True     0.99868\n",
      "False    0.00132\n",
      "Name: ArrivaltoBrainImagingMinutes, dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True     0.981749\n",
      "False    0.018251\n",
      "Name: ArrivaltoBrainImagingMinutes, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# Convert time from minutes to hours and drop missing\n",
    "arr_scan_hours = (raw_data['ArrivaltoBrainImagingMinutes']/60).dropna()\n",
    "\n",
    "# Create boxplot\n",
    "plt.boxplot(arr_scan_hours, vert=0)\n",
    "plt.xlabel('Hours')\n",
    "plt.ylabel('Arrival to scan time')\n",
    "plt.title('Arrival to scan time')\n",
    "plt.show()\n",
    "\n",
    "# Filter to time within one week\n",
    "plt.boxplot(arr_scan_hours[arr_scan_hours < 168], vert=0)\n",
    "plt.xlabel('Hours')\n",
    "plt.ylabel('Arrival to scan time')\n",
    "plt.title('Arrival to scan time (hiding >168 hours/7 days)')\n",
    "plt.show()\n",
    "print((arr_scan_hours < 168).value_counts(normalize=True))\n",
    "\n",
    "# Filter to time within one day\n",
    "plt.boxplot(arr_scan_hours[arr_scan_hours < 24], vert=0)\n",
    "plt.xlabel('Hours')\n",
    "plt.ylabel('Arrival to scan time')\n",
    "plt.title('Arrival to scan time (hiding >24 hours/1 day)')\n",
    "plt.show()\n",
    "print((arr_scan_hours < 24).value_counts(normalize=True))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "False    0.619341\n",
      "True     0.380659\n",
      "Name: onset_to_arrival_time, dtype: float64\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True     0.990589\n",
      "False    0.009411\n",
      "Name: arrival_to_scan_time, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# Filter to onset to arrival minutes < 4 hours (240 minutes)\n",
    "onset_four_arr_scan = data.loc[\n",
    "    data['onset_to_arrival_time'] < 240,\n",
    "    'arrival_to_scan_time']/60\n",
    "\n",
    "# Create boxplot\n",
    "plt.boxplot(onset_four_arr_scan, vert=0)\n",
    "plt.xlabel('Hours')\n",
    "plt.ylabel('Arrival to scan time')\n",
    "plt.title('Arrival to scan time (if onset to arrival <4h)')\n",
    "plt.show()\n",
    "print((data['onset_to_arrival_time'] < 240).value_counts(normalize=True))\n",
    "\n",
    "# Filter to arrival to scan < 20 hours\n",
    "plt.boxplot(onset_four_arr_scan[onset_four_arr_scan < 20], vert=0)\n",
    "plt.xlabel('Hours')\n",
    "plt.ylabel('Arrival to scan time')\n",
    "plt.title('Arrival to scan time (if <20h & onset to arrival <4h)')\n",
    "plt.show()\n",
    "print((onset_four_arr_scan < 20).value_counts(normalize=True))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Scan to thrombolysis time**\n",
    "\n",
    "A scan to thrombolysis time of 0 is considered unrealistic but not completely invalid. It would require input of the same time for scan and thrombolysis. Although technically impossible to have 0 minutes, it is part of noise of data but not invalid results, so these have not been removed from the clean data.\n",
    "\n",
    "*(Decision agreed between MA, KP, AL & AH on 02/06/2023)*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "False    358702\n",
       "True        291\n",
       "Name: scan_to_thrombolysis_time, dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(data['scan_to_thrombolysis_time'] == 0).value_counts()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Thrombolysis**\n",
    "\n",
    "No comment - Y (yes) to 1, and N or NB (no or no but) to 0."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Thrombectomy**\n",
    "\n",
    "If ArrivaltoArterialPunctureMinutes is NaN, we conclude they did not have a thrombectomy. Is there a possibility that this could be missing for anyone who did have a thrombectomy? No, the SSNAP team suggested that this is how we identify people who have had a thrombectomy, and that everyone who had one should have a time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "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>thrombectomy</th>\n",
       "      <th>ArrivaltoArterialPunctureMinutes</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>Missing data</td>\n",
       "      <td>354650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>Have data</td>\n",
       "      <td>4343</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   thrombectomy ArrivaltoArterialPunctureMinutes   count\n",
       "0             0                     Missing data  354650\n",
       "1             1                        Have data    4343"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show where raw data had or was missing data, then compare with clean data\n",
    "raw_thrombectomy = mark_missing(raw_data['ArrivaltoArterialPunctureMinutes'])\n",
    "raw_thrombectomy['PatientId'] = raw_data['PatientId']\n",
    "compare_raw_clean(raw_thrombectomy,\n",
    "                  data[['id', 'thrombectomy']])"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Comorbidities\n",
    "\n",
    "**Diagnoses.** New atrial fibrillation diagnosis is missing data (should be complete for all cases where AtrialFibrillation = N), but still included in cleaned dataset for some descriptive analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>S2CoMAtrialFibrillation</th>\n",
       "      <th>S2NewAFDiagnosis</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "      <td>185842</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>N</td>\n",
       "      <td>Y</td>\n",
       "      <td>13133</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>N</td>\n",
       "      <td>NaN</td>\n",
       "      <td>95809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Y</td>\n",
       "      <td>NaN</td>\n",
       "      <td>65597</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  S2CoMAtrialFibrillation S2NewAFDiagnosis   count\n",
       "0                       N                N  185842\n",
       "3                       N                Y   13133\n",
       "1                       N              NaN   95809\n",
       "2                       Y              NaN   65597"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "((raw_data[['S2CoMAtrialFibrillation', 'S2NewAFDiagnosis']]\n",
    "  .value_counts(dropna=False)\n",
    "  .reset_index(name='count')\n",
    "  .sort_values(by='S2NewAFDiagnosis')))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Anticoagulants and antiplatelets.** We deal with missing values in different ways for each column.\n",
    "* **Antiplatelets** - It is only possible to indicate antiplatelet use when a diagnosis of atrial fibrillation is present, and hence, missing antiplatelet values when atrial fibrillation is false can be set as 0. However, when this is not the case, they remain as missing.\n",
    "* **Anticoagulants** - Marking the use of anticoagulents is not dependent on any other variables, so for that variable, all values that were NaN in the raw data are marked as NaN in the clean data.\n",
    "* **Specific types of anticoagulant** - This is different to the individual, specific types of anticoagulent, for which the data dictionary states that a missing value means they were not on that type of anticoagulant, and hence for those types, NaN was set as 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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>atrial_fibrillation</th>\n",
       "      <th>afib_antiplatelet</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>293618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>55384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9990</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   atrial_fibrillation  afib_antiplatelet   count\n",
       "0                  0.0                0.0  293618\n",
       "1                  1.0                0.0   55384\n",
       "2                  1.0                1.0    9990\n",
       "3                  1.0                NaN       1"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Counts where missing antiplatelet in the clean data\n",
    "(data[['atrial_fibrillation', 'afib_antiplatelet']]\n",
    " .value_counts(dropna=False)\n",
    " .reset_index(name='count')\n",
    " .sort_values(by=['atrial_fibrillation', 'afib_antiplatelet']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "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>atrial_fibrillation</th>\n",
       "      <th>afib_anticoagulant</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>197186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>6782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>89650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>25501</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>39872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   atrial_fibrillation  afib_anticoagulant   count\n",
       "0                  0.0                 0.0  197186\n",
       "4                  0.0                 1.0    6782\n",
       "1                  0.0                 NaN   89650\n",
       "3                  1.0                 0.0   25501\n",
       "2                  1.0                 1.0   39872\n",
       "5                  1.0                 NaN       2"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Counts where missing anticoagulant in the clean data\n",
    "(data[['atrial_fibrillation', 'afib_anticoagulant']]\n",
    " .value_counts(dropna=False)\n",
    " .reset_index(name='count')\n",
    " .sort_values(by=['atrial_fibrillation', 'afib_anticoagulant']))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "If anticoagulent use marked as no, but a specific anticoagulent is marked as true, all anticoagulent results are set as NaN, as we would not know which result is true."
   ]
  },
  {
   "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>S2CoMAFAnticoagulent</th>\n",
       "      <th>S2CoMAFAnticoagulentVitK</th>\n",
       "      <th>S2CoMAFAnticoagulentDOAC</th>\n",
       "      <th>S2CoMAFAnticoagulentHeparin</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>127274</th>\n",
       "      <td>N</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       S2CoMAFAnticoagulent  S2CoMAFAnticoagulentVitK  \\\n",
       "127274                    N                       0.0   \n",
       "\n",
       "        S2CoMAFAnticoagulentDOAC  S2CoMAFAnticoagulentHeparin  \n",
       "127274                       1.0                          0.0  "
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Show patients with discrepancy between any anticoagulant use and\n",
    "# use of a specific anticoagulent columns\n",
    "raw_data.loc[\n",
    "    (((raw_data['S2CoMAFAnticoagulent'] == 'N') |\n",
    "      (raw_data['S2CoMAFAnticoagulent'] == 'NB')\n",
    "      ) &\n",
    "     ((raw_data['S2CoMAFAnticoagulentVitK'] == 1) |\n",
    "      (raw_data['S2CoMAFAnticoagulentDOAC'] == 1) |\n",
    "      (raw_data['S2CoMAFAnticoagulentHeparin'] == 1))),\n",
    "    ['S2CoMAFAnticoagulent', 'S2CoMAFAnticoagulentVitK',\n",
    "     'S2CoMAFAnticoagulentDOAC', 'S2CoMAFAnticoagulentHeparin']]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Prior disability\n",
    "\n",
    "No comment.\n",
    "\n",
    "### NIHSS data\n",
    "\n",
    "No comment.\n",
    "\n",
    "### Death and outcome data\n",
    "\n",
    "Nearly everyone who is marked to have died also had a Modified Rankin Scale (mRS) score of 6, which indicates death. There are some inconsistencies in this data:\n",
    "* mrs=6 and discharge=died but death=0\n",
    "* death=0 and discharge=community team but mrs=6\n",
    "* discharge=died and death=1 but mrs=missing\n",
    "\n",
    "In the first two cases, as we cannot be certain which result is true, propose both variables are set as missing.\n",
    "\n",
    "In the final case, as it is not inconsistency but instead discharge disability missing (which is true for lots of people), and as there is agreement between death and discharge destination, could either set discharge as 6 (but that would falsely change number with mrs reported or leave as is, or set all as missing)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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>death</th>\n",
       "      <th>discharge_disability</th>\n",
       "      <th>discharge_destination</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>died</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>community_team_or_esd</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   death  discharge_disability  discharge_destination  count\n",
       "0      0                   6.0                   died      9\n",
       "1      0                   6.0  community_team_or_esd      1"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Demonstrate issue in raw data\n",
    "# Create dataframe with death columns\n",
    "death_compare = pd.DataFrame()\n",
    "\n",
    "death_compare['death'] = (raw_data['ArrivalToDeathDays'] >= 0) * 1\n",
    "\n",
    "death_compare['discharge_disability'] = raw_data['S7RankinDischarge']\n",
    "\n",
    "discharge: dict = {\n",
    "    'CH': 'care_home',\n",
    "    'D': 'died',\n",
    "    'H': 'home',\n",
    "    'SE': 'somewhere_else',\n",
    "    'TC': 'community_team_or_esd',\n",
    "    'TCN': 'community_team_or_esd',\n",
    "    'TN': 'non_ssnap_hospital_team',\n",
    "    'T': 'ssnap_hospital_team',\n",
    "    np.NaN: 'missing'}\n",
    "death_compare['discharge_destination'] = (\n",
    "    raw_data['S7DischargeType'].map(discharge))\n",
    "\n",
    "# Identify patients who have any marker indicating death\n",
    "any_death = ((death_compare['death'] == 1) |\n",
    "             (death_compare['discharge_disability'] == 6) |\n",
    "             (death_compare['discharge_destination'] == 'died'))\n",
    "\n",
    "# Identify patients who have all markers indicating death\n",
    "# (or, if not, missing a result)\n",
    "all_or_missing_death = (\n",
    "    ((death_compare['death'] == 1) |\n",
    "     (death_compare['death'].isnull())) &\n",
    "    ((death_compare['discharge_disability'] == 6) |\n",
    "     (death_compare['discharge_disability'].isnull())) &\n",
    "    ((death_compare['discharge_destination'] == 'died') |\n",
    "     (death_compare['discharge_destination'].isnull())))\n",
    "\n",
    "# Extract columns where have at least one marker indicating death,\n",
    "# but then others do not (don't extract if its just because others\n",
    "# are missing) - and set all their death columns as NaN\n",
    "(death_compare.loc[\n",
    "    any_death & ~all_or_missing_death,\n",
    "    ['death', 'discharge_disability', 'discharge_destination']]\n",
    "    .value_counts(dropna=False)\n",
    "    .reset_index(name='count'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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>death</th>\n",
       "      <th>discharge_disability</th>\n",
       "      <th>discharge_destination</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.0</td>\n",
       "      <td>6.0</td>\n",
       "      <td>died</td>\n",
       "      <td>49839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>died</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   death  discharge_disability discharge_destination  count\n",
       "0    1.0                   6.0                  died  49839\n",
       "1    1.0                   NaN                  died      2"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Demonstration that these values are removed in the clean data\n",
    "(data.loc[((data['death'] == 1) |\n",
    "          (data['discharge_disability'] == 6) |\n",
    "          (data['discharge_destination'] == 'died')),\n",
    "          ['death', 'discharge_disability', 'discharge_destination']]\n",
    " .value_counts(dropna=False).reset_index(name='count'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reasons for no thrombolysis\n",
    "\n",
    "No comment - as in SSNAP dataset - includes \"no\" and \"no but\" reasons."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Summary: number of patients removed during data cleaning\n",
    "\n",
    "Below is a summary of the total number of patients completely removed from the full dataset during the data cleaning process."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "    Original dataset contained 360381 patients.\n",
      "    We have removed 1388 patients.\n",
      "    Cleaned dataset contains 358993 patients.\n",
      "    \n"
     ]
    }
   ],
   "source": [
    "# Print change to number of patients in dataset\n",
    "original_n = len(raw_data.index)\n",
    "clean_n = len(data.index)\n",
    "drop_n = original_n - clean_n\n",
    "print(\n",
    "    f'''\n",
    "    Original dataset contained {original_n} patients.\n",
    "    We have removed {drop_n} patients.\n",
    "    Cleaned dataset contains {clean_n} patients.\n",
    "    ''')"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Patterns in issues"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "time_some_missing       4.309051\n",
       "time_negative           1.833892\n",
       "time_zero               0.079083\n",
       "times_but_no_amb        0.002497\n",
       "onset_negative          0.016927\n",
       "onset_unknown          33.088037\n",
       "anticoag_discrep        0.000277\n",
       "inconsistent_death      0.002775\n",
       "missing_stroke_type     0.385148\n",
       "dtype: float64"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Percentage with each issue overall\n",
    "issues.drop('stroke_team', axis=1).mean(axis=0)*100"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    360381.000000\n",
       "mean          0.397177\n",
       "std           0.526028\n",
       "min           0.000000\n",
       "25%           0.000000\n",
       "50%           0.000000\n",
       "75%           1.000000\n",
       "max           3.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "count    360381.000000\n",
       "mean          0.066297\n",
       "std           0.251010\n",
       "min           0.000000\n",
       "25%           0.000000\n",
       "50%           0.000000\n",
       "75%           0.000000\n",
       "max           2.000000\n",
       "dtype: float64"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Frequency of issues per individual\n",
    "issues_patient = (\n",
    "    issues.drop('stroke_team', axis=1, inplace=False).sum(axis=1))\n",
    "display(issues_patient.describe())\n",
    "issues_patient.hist()\n",
    "plt.xlabel('Number of issues')\n",
    "plt.ylabel('Number of patients')\n",
    "plt.show()\n",
    "\n",
    "# Dropping onset unknown (as so many have that issue) and filtering to 1+\n",
    "issues_patient = (\n",
    "    issues.drop(['stroke_team', 'onset_unknown'], axis=1, inplace=False)\n",
    "          .sum(axis=1))\n",
    "display(issues_patient.describe())\n",
    "issues_patient[issues_patient > 0].hist()\n",
    "plt.title('Excluding onset unknown and filtering to 1+')\n",
    "plt.xlabel('Number of issues')\n",
    "plt.ylabel('Number of patients')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 700x700 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "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>stroke_team</th>\n",
       "      <th>time_some_missing</th>\n",
       "      <th>time_negative</th>\n",
       "      <th>time_zero</th>\n",
       "      <th>times_but_no_amb</th>\n",
       "      <th>onset_negative</th>\n",
       "      <th>onset_unknown</th>\n",
       "      <th>anticoag_discrep</th>\n",
       "      <th>inconsistent_death</th>\n",
       "      <th>missing_stroke_type</th>\n",
       "      <th>norm_sum</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>84</th>\n",
       "      <td>Russells Hall Hospital</td>\n",
       "      <td>0.174886</td>\n",
       "      <td>2.938090</td>\n",
       "      <td>0.524659</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>31.024834</td>\n",
       "      <td>0.034977</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.973068</td>\n",
       "      <td>3.551336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>New Cross Hospital</td>\n",
       "      <td>0.044484</td>\n",
       "      <td>2.201957</td>\n",
       "      <td>0.511566</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.066726</td>\n",
       "      <td>54.915480</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.044484</td>\n",
       "      <td>1.623665</td>\n",
       "      <td>3.202248</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>Cumberland Infirmary</td>\n",
       "      <td>29.197397</td>\n",
       "      <td>0.780911</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.216920</td>\n",
       "      <td>23.947939</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.086768</td>\n",
       "      <td>2.425406</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>Northampton General Hospital</td>\n",
       "      <td>0.064336</td>\n",
       "      <td>1.158053</td>\n",
       "      <td>0.021445</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.128673</td>\n",
       "      <td>52.069483</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.021445</td>\n",
       "      <td>1.586961</td>\n",
       "      <td>2.356433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Blackpool Victoria Hospital</td>\n",
       "      <td>28.782161</td>\n",
       "      <td>0.377358</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.034305</td>\n",
       "      <td>55.162950</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.480274</td>\n",
       "      <td>2.226348</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>Leighton Hospital</td>\n",
       "      <td>18.929254</td>\n",
       "      <td>0.191205</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>40.439771</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>2.198853</td>\n",
       "      <td>2.062902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>Royal Hampshire County Hospital</td>\n",
       "      <td>11.741974</td>\n",
       "      <td>10.582640</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.029727</td>\n",
       "      <td>33.026159</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.386445</td>\n",
       "      <td>2.057234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>Royal Hallamshire Hospital</td>\n",
       "      <td>0.061741</td>\n",
       "      <td>5.577279</td>\n",
       "      <td>0.144063</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.061741</td>\n",
       "      <td>52.665157</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.514509</td>\n",
       "      <td>2.010493</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>Queen Elizabeth Hospital Edgbaston</td>\n",
       "      <td>0.031319</td>\n",
       "      <td>2.724710</td>\n",
       "      <td>0.626370</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>44.754150</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.281867</td>\n",
       "      <td>2.009905</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>Invicta Ward Kent and Canterbury Hospital</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>8.280828</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.090009</td>\n",
       "      <td>25.112511</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>1.530153</td>\n",
       "      <td>2.008879</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  stroke_team  time_some_missing  \\\n",
       "84                     Russells Hall Hospital           0.174886   \n",
       "47                         New Cross Hospital           0.044484   \n",
       "11                       Cumberland Infirmary          29.197397   \n",
       "51               Northampton General Hospital           0.064336   \n",
       "2                 Blackpool Victoria Hospital          28.782161   \n",
       "37                          Leighton Hospital          18.929254   \n",
       "75            Royal Hampshire County Hospital          11.741974   \n",
       "74                 Royal Hallamshire Hospital           0.061741   \n",
       "64         Queen Elizabeth Hospital Edgbaston           0.031319   \n",
       "28  Invicta Ward Kent and Canterbury Hospital           0.000000   \n",
       "\n",
       "    time_negative  time_zero  times_but_no_amb  onset_negative  onset_unknown  \\\n",
       "84       2.938090   0.524659               0.0        0.000000      31.024834   \n",
       "47       2.201957   0.511566               0.0        0.066726      54.915480   \n",
       "11       0.780911   0.000000               0.0        0.216920      23.947939   \n",
       "51       1.158053   0.021445               0.0        0.128673      52.069483   \n",
       "2        0.377358   0.000000               0.0        0.034305      55.162950   \n",
       "37       0.191205   0.000000               0.0        0.000000      40.439771   \n",
       "75      10.582640   0.000000               0.0        0.029727      33.026159   \n",
       "74       5.577279   0.144063               0.0        0.061741      52.665157   \n",
       "64       2.724710   0.626370               0.0        0.000000      44.754150   \n",
       "28       8.280828   0.000000               0.0        0.090009      25.112511   \n",
       "\n",
       "    anticoag_discrep  inconsistent_death  missing_stroke_type  norm_sum  \n",
       "84          0.034977            0.000000             2.973068  3.551336  \n",
       "47          0.000000            0.044484             1.623665  3.202248  \n",
       "11          0.000000            0.000000             0.086768  2.425406  \n",
       "51          0.000000            0.021445             1.586961  2.356433  \n",
       "2           0.000000            0.000000             0.480274  2.226348  \n",
       "37          0.000000            0.000000             2.198853  2.062902  \n",
       "75          0.000000            0.000000             0.386445  2.057234  \n",
       "74          0.000000            0.000000             0.514509  2.010493  \n",
       "64          0.000000            0.000000             0.281867  2.009905  \n",
       "28          0.000000            0.000000             1.530153  2.008879  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Percentage with each issue by team\n",
    "issues_team = issues.groupby('stroke_team').mean()*100\n",
    "\n",
    "# Use min max scaling to normalise columns (so each range from 0 to 1)\n",
    "issues_team_norm = issues_team.apply(\n",
    "    lambda x: (x - x.min()) / (x.max() - x.min())).reset_index()\n",
    "\n",
    "# Sort by sum of each row\n",
    "issues_team_norm['norm_sum'] = (\n",
    "    issues_team_norm.drop('stroke_team', axis=1).sum(axis=1))\n",
    "issues_team_norm_sort = (\n",
    "    issues_team_norm\n",
    "    .sort_values(by='norm_sum', ascending=False)\n",
    "    .drop('norm_sum', axis=1))\n",
    "\n",
    "# Show plot\n",
    "fig, ax = plt.subplots(figsize=(7, 7))\n",
    "plt.imshow(issues_team_norm_sort.drop('stroke_team', axis=1), aspect='auto')\n",
    "plt.xlabel('Issue')\n",
    "plt.ylabel('Stroke team')\n",
    "plt.show()\n",
    "\n",
    "# Append sorting to non-normalised version and sort it\n",
    "issues_team = issues_team.merge(issues_team_norm[['stroke_team', 'norm_sum']],\n",
    "                                on='stroke_team')\n",
    "display(issues_team.sort_values(by='norm_sum', ascending=False).head(10))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Excluded variables"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### INR\n",
    "\n",
    "**Comparing INR, INR High and INR not known**\n",
    "\n",
    "* INR High is only marked as true in 31 cases\n",
    "* In all these cases, they are missing INR result and marked as INR not checked.\n",
    "\n",
    "Hence, there are serious inconsistencies between the three INR columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "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>S2INR</th>\n",
       "      <th>S2INRHigh</th>\n",
       "      <th>S2INRNK</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>238845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>112376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Have data</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>8506</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          S2INR  S2INRHigh  S2INRNK   count\n",
       "0  Missing data        0.0      0.0  238845\n",
       "1  Missing data        NaN      NaN  112376\n",
       "2     Have data        0.0      0.0    8506\n",
       "3  Missing data        0.0      1.0     623\n",
       "4  Missing data        1.0      0.0      31"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Look at proportion with INR high and INR result\n",
    "inr_high = mark_missing(raw_data['S2INR'])\n",
    "inr_high['S2INRHigh'] = raw_data['S2INRHigh']\n",
    "inr_high['S2INRNK'] = raw_data['S2INRNK']\n",
    "(inr_high[['S2INR', 'S2INRHigh', 'S2INRNK']]\n",
    " .value_counts(dropna=False)\n",
    " .reset_index(name='count'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Mobile data to arrival\n",
    "\n",
    "This is missing for the majority of patients. Everyone who has MobileDatatoArrivalMinutes also have data for CallConnectedtoArrivalMinutes, so this variable is not needed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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>MobileDatatoArrivalMinutes</th>\n",
       "      <th>CallConnectedtoArrivalMinutes</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>Missing data</td>\n",
       "      <td>251896</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>Have data</td>\n",
       "      <td>92414</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Have data</td>\n",
       "      <td>Have data</td>\n",
       "      <td>16071</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  MobileDatatoArrivalMinutes CallConnectedtoArrivalMinutes   count\n",
       "0               Missing data                  Missing data  251896\n",
       "1               Missing data                     Have data   92414\n",
       "2                  Have data                     Have data   16071"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(mark_missing(raw_data[['MobileDatatoArrivalMinutes',\n",
    "                        'CallConnectedtoArrivalMinutes']])\n",
    " .value_counts()\n",
    " .reset_index(name='count'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### TIA in last month\n",
    "\n",
    "This is missing for the majority of patients. Also, it should only be completed if 2.1.5 (had stroke/TIA) is marked as yes, but this is not the case. Hence, not included in clean dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>S2CoMStrokeTIA</th>\n",
       "      <th>S2TIAInLastMonth</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>N</td>\n",
       "      <td>NaN</td>\n",
       "      <td>268486</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>N</td>\n",
       "      <td>N</td>\n",
       "      <td>342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>N</td>\n",
       "      <td>NK</td>\n",
       "      <td>42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Y</td>\n",
       "      <td>NaN</td>\n",
       "      <td>64115</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Y</td>\n",
       "      <td>N</td>\n",
       "      <td>22205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Y</td>\n",
       "      <td>NK</td>\n",
       "      <td>3845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Y</td>\n",
       "      <td>Y</td>\n",
       "      <td>1346</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  S2CoMStrokeTIA S2TIAInLastMonth   count\n",
       "0              N              NaN  268486\n",
       "5              N                N     342\n",
       "6              N               NK      42\n",
       "1              Y              NaN   64115\n",
       "2              Y                N   22205\n",
       "3              Y               NK    3845\n",
       "4              Y                Y    1346"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(raw_data[['S2CoMStrokeTIA', 'S2TIAInLastMonth']]\n",
    " .value_counts(dropna=False)\n",
    " .reset_index(name='count')\n",
    " .sort_values(by='S2CoMStrokeTIA'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stroke unit death\n",
    "\n",
    "Not included in clean dataset, as we are not currently interested in the location of death."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pre-hospital impression\n",
    "\n",
    "S1PreHospitalImpression - not certain what it is (might be to do with what paramedic think is the diagnosis - e.g. fast positive) and did not ask for it, not included in cleaned dataset."
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Appendices"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Looking further into INR\n",
    "\n",
    "Measurement of INR is contingent on whether patients were on Vitamin K antagonists (inc. Warfarin) before their stroke.\n",
    "\n",
    "*In relation to Mike's comment that expect people with INR measurements to be on anticoagulants*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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>S2INR</th>\n",
       "      <th>anticoag_vitk</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>0.0</td>\n",
       "      <td>238845</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>NaN</td>\n",
       "      <td>112376</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Have data</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8506</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>1.0</td>\n",
       "      <td>654</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          S2INR  anticoag_vitk   count\n",
       "0  Missing data            0.0  238845\n",
       "1  Missing data            NaN  112376\n",
       "2     Have data            1.0    8506\n",
       "3  Missing data            1.0     654"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Find whether missing INR or not, and VitK value\n",
    "inr = mark_missing(raw_data['S2INR'])\n",
    "inr['anticoag_vitk'] = raw_data['S2CoMAFAnticoagulentVitK']\n",
    "\n",
    "# Look at those counts\n",
    "inr_counts = inr.value_counts(dropna=False).reset_index(name='count')\n",
    "inr_counts"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are equal numbers with best estimate v.s. precise v.s. not known onset, who have INR measurement\n",
    "\n",
    "*In relation to Mike's comment that wouldn't expect INR measurement when don't know onset*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "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>S2INR</th>\n",
       "      <th>anticoag_vitk</th>\n",
       "      <th>S1OnsetTimeType</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Have data</td>\n",
       "      <td>1.0</td>\n",
       "      <td>BE</td>\n",
       "      <td>2949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Have data</td>\n",
       "      <td>1.0</td>\n",
       "      <td>P</td>\n",
       "      <td>2876</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Have data</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NK</td>\n",
       "      <td>2681</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>1.0</td>\n",
       "      <td>BE</td>\n",
       "      <td>258</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NK</td>\n",
       "      <td>231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>1.0</td>\n",
       "      <td>P</td>\n",
       "      <td>165</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          S2INR  anticoag_vitk S1OnsetTimeType  count\n",
       "0     Have data            1.0              BE   2949\n",
       "1     Have data            1.0               P   2876\n",
       "2     Have data            1.0              NK   2681\n",
       "3  Missing data            1.0              BE    258\n",
       "4  Missing data            1.0              NK    231\n",
       "5  Missing data            1.0               P    165"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inr_onset = inr.copy()\n",
    "inr_onset['S1OnsetTimeType'] = raw_data['S1OnsetTimeType']\n",
    "(inr_onset[inr_onset['anticoag_vitk'] == 1]\n",
    " .value_counts(dropna=False)\n",
    " .reset_index(name='count'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For patients who were on Vitamin K antagonists, 92.8% have INR measurements and 7.1% are missing data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2    0.928603\n",
       "3    0.071397\n",
       "Name: count, dtype: float64"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Look at proportion of people on VitK with INR data\n",
    "vitk = inr_counts[inr_counts['anticoag_vitk'] == 1]\n",
    "vitk['count'] / vitk['count'].sum()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "INR measurements are present for patients with both stroke types.\n",
    "\n",
    "*In relation to Mike's comment asking if they measure when its haemorrhage*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "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>S2INR</th>\n",
       "      <th>stroke_type</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>I</td>\n",
       "      <td>308504</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>Have data</td>\n",
       "      <td>I</td>\n",
       "      <td>6637</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>PIH</td>\n",
       "      <td>42012</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Have data</td>\n",
       "      <td>PIH</td>\n",
       "      <td>1840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Missing data</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1359</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Have data</td>\n",
       "      <td>NaN</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          S2INR stroke_type   count\n",
       "0  Missing data           I  308504\n",
       "2     Have data           I    6637\n",
       "1  Missing data         PIH   42012\n",
       "3     Have data         PIH    1840\n",
       "4  Missing data         NaN    1359\n",
       "5     Have data         NaN      29"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inr_stroke = inr.drop('anticoag_vitk', axis=1)\n",
    "inr_stroke['stroke_type'] = raw_data['S2StrokeType']\n",
    "\n",
    "(inr_stroke\n",
    " .value_counts(dropna=False)\n",
    " .reset_index(name='count')\n",
    " .sort_values(by='stroke_type'))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Can see that individuals with INR measurement only also had thrombolysis when values were less than 4. Individuals with higher INR are mostly in thrombolysis=NB, with one in thrombolysis=N.\n",
    "\n",
    "*In relation to Mike's comment about relationship between INR measurement, haemorrhage and thrombolysis - and that would expect high INR to not have thrombolysis*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkUAAAG6CAYAAAAGUjKQAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABSa0lEQVR4nO3deVgT1/4G8DckEDZBQRYVBAQroijuaxX3pe6Xtm7VVu3ivlVb29uiXvfdam3rctVqtVZLbavUK1dxq1C3WhVXFBEUqyIKgkAC5/eHv8w1BDCELIDv53ny6JycOfPNSSAvk5mJTAghQERERPSSs7J0AURERERlAUMRERERERiKiIiIiAAwFBEREREBYCgiIiIiAsBQRERERASAoYiIiIgIAEMREREREQCGIiIiIiIADEVURm3atAkymUzr5ubmhtDQUOzZs8fS5Ul8fX3x9ttvl3i9rKwszJw5E4cOHTJ6TRVVaGgoQkNDX9hPJpNh3LhxJq8nKSkJY8aMwSuvvAI7Ozu4uLggODgY7777LpKSkqR+ERERGDRoEAICAmBnZwdfX18MGTIE165dk/rMnDlT5/Ve2E3z+ENDQ1G/fn2TP0Zj0Pwsnzp1ymhjaubLFG7evAmZTIZNmzaZZHwq2xSWLoCoOBs3bkRgYCCEELh79y5Wr16N3r1745dffkHv3r0tXZ7BsrKyMGvWLADQ642eypbk5GQ0btwYlStXxtSpU1GnTh08fvwYFy9exA8//IAbN27A29sbALBw4UJ4enri008/Ra1atZCUlIR58+ahcePGiI2NRb169TBq1Ch0795dGj8lJQUDBgzA+PHjMXjwYKndycnJ7I/1ZVOtWjXExMTA39/f0qWQBTAUUZlWv359NG3aVFru3r07qlSpgu3bt5frUGRuWVlZsLe3t3QZFca6devw4MEDnDhxAn5+flJ7v3798MknnyA/P19q+/XXX+Hu7q61fseOHeHr64vly5dj/fr18PLygpeXl3T/zZs3AQA1a9ZEy5YtjVq7SqWCTCaDQsFf/4VRKpVGn3MqP/jxGZUrtra2sLGxgbW1tVb7w4cPMWbMGNSoUQM2NjaoVasWPv30U+Tk5AAAsrOz0ahRIwQEBODx48fSenfv3oWnpydCQ0ORl5cHAHj77bfh6OiIuLg4dOrUCQ4ODnBzc8O4ceOQlZX1whpv3bqFoUOHwt3dHUqlEnXr1sXSpUulN8qbN2/Czc0NADBr1izpo5EXfQwXFxeHrl27wt7eHm5ubhg7diz27t0LmUym9TGc5qOVI0eOoHXr1rC3t8eIESP0qg0ADh06pDOmpu6CHyuUZK6EEFizZg1CQkJgZ2eHKlWqICwsDDdu3NDpt2jRIvj4+MDW1haNGzfGb7/99sJ5L+ibb77BK6+8AqVSiaCgIHz//fdaj0WhUGD+/Pk66x05cgQymQw7d+4scuzU1FRYWVnphB0NK6v//WotrE/16tXh5eWl9TGbIU6ePIlXX30V9vb2qFWrFhYsWFDoc7llyxZMnToVNWrUgFKpRHx8PADg3//+Nxo2bAhbW1u4uLigf//+uHTpktY2NM/x5cuX0a1bNzg4OKBatWpYsGABACA2NhZt27aFg4MDXnnlFWzevLnQWtPS0vDOO+/AxcUFDg4O6N27t85zr29NBY0cORIuLi6F/nx27NgR9erVk5Z37tyJFi1awNnZWZo3zc8HUPjr/P79+3jvvffg7e0NpVIJNzc3tGnTBv/973+LrYvKIUFUBm3cuFEAELGxsUKlUonc3FyRlJQkJkyYIKysrMS+ffukvk+fPhUNGjQQDg4OYsmSJWL//v3is88+EwqFQvTs2VPqd/XqVVGpUiUxYMAAIYQQeXl5omPHjsLd3V3cuXNH6jd8+HBhY2MjatasKebOnSv2798vZs6cKRQKhejVq5dWnT4+PmL48OHS8r1790SNGjWEm5ub+Prrr8W+ffvEuHHjBAAxevRoIYQQ2dnZYt++fQKAGDlypIiJiRExMTEiPj6+yPm4c+eOcHV1FTVr1hSbNm0SkZGR4q233hK+vr4CgIiOjpb6tm/fXri4uAhvb2+xatUqER0dLQ4fPqxXbUIIER0drTOmEEIkJCQIAGLjxo0GzdW7774rrK2txdSpU8W+ffvEtm3bRGBgoPDw8BB3796V+oWHh0tz89tvv4m1a9eKGjVqCE9PT9G+ffsi50gDgPD29hZBQUFi+/bt4pdffhHdu3cXAMTOnTulfv379xc1a9YUarVaa/3XX39dVK9eXahUqiK3sXXrVgFAdO3aVezbt088fvz4hXU97/r168LKykpMnjy50Ps1c7148eJC72/fvr1wdXUVtWvXFl9//bWIiooSY8aMEQDE5s2bpX6a57JGjRoiLCxM/PLLL2LPnj0iNTVVzJs3TwAQgwYNEnv37hXffvutqFWrlnB2dhZXr16VxtA8x3Xr1hUrV64UUVFR4p133hEAxIwZM8Qrr7wiNmzYIP7zn/+IXr16CQDi1KlT0vqan2Vvb28xYsQI6Tl1d3cX3t7eIi0tTeqrb02a14jGX3/9JQCIdevWac1TXFycACC+/PJLIYQQx48fFzKZTAwcOFBERkaKgwcPio0bN4q33npLZ+6ff51369ZNuLm5ibVr14pDhw6J3bt3i88//1x8//33hT4/VH4xFFGZpPlFWvCmVCrFmjVrtPp+/fXXAoD44YcftNoXLlwoAIj9+/dLbTt27BAAxIoVK8Tnn38urKystO4X4tmbAACxcuVKrfa5c+cKAOLYsWNSW8FQ9PHHHwsA4o8//tBad/To0UImk4krV64IIYS4f/++ACDCw8P1mo9p06YJmUwm4uLitNq7detWaCgCIA4cOKDVV9/aShqK9JmrmJgYAUAsXbpUq19SUpKws7MT06dPF0IIkZaWJmxtbUX//v21+v3+++8CgN6hyM7OTitoqdVqERgYKAICAqQ2zeP86aefpLbbt28LhUIhZs2aVew28vPzxfvvvy+srKwEACGTyUTdunXF5MmTRUJCQrHrqlQqERoaKpycnMStW7cK7aNPKCrsuQwKChLdunXTeYzt2rXT6peWlibs7Oy0/mgQQohbt24JpVIpBg8eLLVpnuMff/xR6zG4ubkJAOLMmTNSe2pqqpDL5WLKlClSm+ZnuajndM6cOSWuqWAo0sxJSEiIVtvo0aOFk5OTyMjIEEIIsWTJEgFAPHr0SBSlsNe5o6OjmDRpUpHrUMXBj8+oTPv2229x8uRJnDx5Er/99huGDx+OsWPHYvXq1VKfgwcPwsHBAWFhYVrraj6OOnDggNT2xhtvYPTo0Zg2bRrmzJmDTz75BF26dCl020OGDNFa1hzwGh0dXWS9Bw8eRFBQEJo3b65TixACBw8efPGDLsThw4dRv359BAUFabUPGjSo0P5VqlRBx44dzVIb8OK52rNnD2QyGYYOHQq1Wi3dPD090bBhQ+mjupiYGGRnZ+uM17p1a/j4+OhdT6dOneDh4SEty+VyvPnmm4iPj0dycjKAZx8zNmzYEF9++aXU7+uvv4ZMJsN7771X7PgymQxff/01bty4gTVr1uCdd96BSqXC8uXLUa9ePRw+fLjQ9YQQGDlyJI4ePYpvv/1WOhjbEJ6enjrPZYMGDZCYmKjT9x//+IfWckxMDJ4+farzka23tzc6duyo9TMDPHu8PXv2lJYVCgUCAgJQrVo1NGrUSGp3cXGBu7t7oTUU9ZxqXiMlramgiRMn4uzZs/j9998BAOnp6diyZQuGDx8OR0dHAECzZs0APPs98MMPP+D27dvFjqnRvHlzbNq0CXPmzEFsbCxUKpVe61H5w1BEZVrdunXRtGlTNG3aFN27d8c333yDrl27Yvr06Xj06BGAZ8d3eHp66pyi6+7uDoVCgdTUVK32ESNGQKVSQaFQYMKECYVuV6FQwNXVVavN09NT2l5RUlNTUa1aNZ326tWrv3Dd4qSmpmq9yWsU1gag0BpMVZs+c/X3339DCAEPDw9YW1tr3WJjY/HgwQOt/pr1CxtTH8Wt//zjnDBhAg4cOIArV65ApVJh3bp1CAsL03tbPj4+GD16NDZs2IBr165hx44dyM7OxrRp03T6CiEwatQobN26FZs2bULfvn31fjyFKTjnwLODhJ8+farTXvB518xBUa+Hgq8Fe3t72NraarXZ2NjAxcVFZ30bGxtkZ2frtBf1nGi2VdKaCurbty98fX2lkLtp0yZkZmZi7NixUp927dph9+7dUKvVGDZsGLy8vFC/fn1s37692LF37NiB4cOHY/369WjVqhVcXFwwbNgw3L17t9j1qPxhKKJyp0GDBnj69CmuXr0K4Nmbg+ZN93n37t2DWq1G1apVpbbMzEy89dZb0rVlRo0aVeg21Gq1zi9hzS/Awt6MNFxdXZGSkqLTfufOHQDQqqUkNI+xoKJ+KRd2DRd9a9O8+WkOUtfQBJeC9JmrqlWrQiaT4dixY9Kev+dvu3fv1upf2OMqyRtQces///wNHjwYrq6u+PLLL7Fz507cvXtX6020pN544w00aNAAFy5c0GrXBKKNGzdi/fr1GDp0qMHbMETB14NmDop6PRj6Oi1OUc+JppbS1mRlZYWxY8di165dSElJwZo1a9CpUyfUqVNHq1/fvn1x4MABPH78GIcOHYKXlxcGDx6MmJiYIseuWrUqVqxYgZs3byIxMRHz589HRESEQdcoo7KNoYjKnbNnzwKAdAZXp06d8OTJE+mNVePbb7+V7tf44IMPcOvWLURERGDDhg345ZdfsHz58kK3891332ktb9u2DUDx1xXq1KkTLl68iDNnzujUIpPJ0KFDBwDP/qIHUOhf9YVp3749Lly4gIsXL2q1P39G1YvoW5uvry8A4Ny5c1r9fvnllyLHftFc9erVC0II3L59W9rz9/wtODgYANCyZUvY2trqjHf8+PFCP5IpyoEDB7RCZF5eHnbs2AF/f3+tU99tbW3x3nvvYfPmzVi2bBlCQkLQpk2bF45f2Bs3ADx58gRJSUnS3jfgWSB69913sXHjRnzzzTd455139H4cptKqVSvY2dlh69atWu3Jyck4ePCg1s+MsRT1nGpeI8aoadSoUbCxscGQIUNw5cqVYi/iqVQq0b59eyxcuBAA8Oeff+r1OGrWrIlx48ahS5cuOj9LVP7xQhVUpl24cAFqtRrAs93rERERiIqKQv/+/aXrwwwbNgxffvklhg8fjps3byI4OBjHjh3DvHnz0LNnT3Tu3BkAsH79emzduhUbN25EvXr1UK9ePYwbNw4fffQR2rRpo3V8ho2NDZYuXYonT56gWbNmOH78OObMmYMePXqgbdu2RdY7efJkfPvtt3jttdcwe/Zs+Pj4YO/evVizZg1Gjx6NV155BQBQqVIl+Pj44Oeff0anTp3g4uKCqlWrSoGkoEmTJuHf//43evTogdmzZ8PDwwPbtm3D5cuXAWifAl7a2jw9PdG5c2fMnz8fVapUgY+PDw4cOICIiIhCx9Vnrtq0aYP33nsP77zzDk6dOoV27drBwcEBKSkpOHbsGIKDgzF69GhUqVIFH374IebMmYNRo0bh9ddfR1JSEmbOnFmij8+qVq2Kjh074rPPPoODgwPWrFmDy5cvFxoix4wZg0WLFuH06dNYv369XuPPnTsXv//+O958803pEgMJCQlYvXo1UlNTsXjxYqnvhAkTsGHDBowYMQLBwcGIjY2V7lMqlVrH5JhL5cqV8dlnn+GTTz7BsGHDMGjQIKSmpmLWrFmwtbVFeHi40bd56tQpref0008/RY0aNTBmzBij1VS5cmUMGzYMX331FXx8fHSuZfb5558jOTkZnTp1gpeXFx49eoSVK1fC2toa7du3L3TMx48fo0OHDhg8eDACAwNRqVIlnDx5Evv27cOAAQNKPzFUtljsEG+iYhR29pmzs7MICQkRy5YtE9nZ2Vr9U1NTxQcffCCqVasmFAqF8PHxETNmzJD6nTt3TtjZ2WmdKSbEs9PjmzRpInx9faVTg4cPHy4cHBzEuXPnRGhoqLCzsxMuLi5i9OjR4smTJ1rrFzz7TAghEhMTxeDBg4Wrq6uwtrYWderUEYsXLxZ5eXla/f773/+KRo0aCaVSKQDojFPQhQsXROfOnYWtra1wcXERI0eOFJs3bxYAxF9//SX1a9++vahXr16hY+hbW0pKiggLCxMuLi7C2dlZDB06VJw6darQs8/0nSshhPj3v/8tWrRoIRwcHISdnZ3w9/cXw4YN0zqFOz8/X8yfP194e3sLGxsb0aBBA/Hrr7+K9u3b63322dixY8WaNWuEv7+/sLa2FoGBgeK7774rcp3Q0FDh4uIisrKyXji+EELExsaKsWPHioYNGwoXFxchl8uFm5ub6N69u4iMjNTq6+PjU+iZlACEj49PoePrc/ZZYc/x8OHDtcbUnH32/KUInrd+/XrRoEEDYWNjI5ydnUXfvn11znDUPMf61uDj4yNee+01aVnzs7x//37x1ltvicqVK0tnmV27ds2gmgo7+0zj0KFDAoBYsGCBzn179uwRPXr0EDVq1BA2NjbC3d1d9OzZUxw9elTqU/Dss+zsbPHBBx+IBg0aCCcnJ2FnZyfq1KkjwsPDRWZmZqE1UPklE6LAgRhEL7m3334bu3btwpMnTyxdygu999572L59O1JTU2FjY2P27ZenuSrKvXv34OPjg/Hjx2PRokWWLodKaerUqfjqq6+QlJRU7PF/RIXhx2dE5cTs2bNRvXp11KpVC0+ePMGePXuwfv16/POf/7RIICrvkpOTcePGDSxevBhWVlaYOHGipUuiUoiNjcXVq1exZs0avP/++wxEZBCGIqJywtraGosXL0ZycjLUajVq166NZcuW8c3cQOvXr8fs2bPh6+uL7777DjVq1LB0SVQKrVq1gr29PXr16oU5c+ZYuhwqp/jxGRERERF4Sj4RERERAIYiIiIiIgAMRURUhvzxxx/o378/atasCaVSCQ8PD7Rq1QpTp04F8OwijMuWLUP37t3h5eUFe3t71K1bFx9//LH0tS8aN2/ehEwmw5IlS6S2Q4cOQSaTQSaTFXoF47ffflv6niyN0NBQaR2ZTAZbW1sEBQVhzpw5yM3NNf4kEJHFMBQRUZmwd+9etG7dGunp6Vi0aBH279+PlStXok2bNtixYweAZ1cAnzlzJnx8fLBixQpERkbi3Xffxdq1a9GmTRu9rxAOANOnT9e7b61atRATE4OYmBjs3LkTtWvXxmeffVbsFZOJqPzh2WdEVCYsWrQIfn5++M9//gOF4n+/mgYOHChdP0hz5ejnT7cODQ1FzZo18frrr+PHH3/U63vFunfvjn379uHXX3/VuepxYezs7NCyZUtpuUePHggKCsLmzZvxxRdf6HxZKhGVT9xTRERlQmpqKqpWraoViDQ0X2Mil8sLvf6M5itakpKS9NrW22+/jaCgIMyYMQN5eXklrlWhUCAkJAS5ubk6H9sRUfnFUEREZUKrVq3wxx9/YMKECfjjjz+gUqn0XvfgwYMAgHr16unVXy6XY/78+YiLi8PmzZsNqjchIQGVK1eWvpiYiMo/hiIiKhMWLFiAtm3bYtWqVWjZsiUcHBzQpk0bLFiwoNivEbl9+zY+/vhjNG3aFL169dJ7e3369EHbtm0RHh6O7OzsF/ZXq9VQq9W4e/cuwsPDcerUKSxYsAByuVzvbRJR2cZQRERlgqurK44ePYqTJ09iwYIF6Nu3L65evYoZM2YgODgYDx480Fnn4cOH6NmzJ4QQ2LFjh/Qxm74WLlyI5ORkrFy5sth+cXFxsLa2hrW1NapVq4bZs2djxowZeP/990u0PSIq2xiKiKhMadq0KT766CPs3LkTd+7cweTJk3Hz5k2dL2tNS0tDly5dcPv2bURFRaFWrVol3lbr1q3Rr18/LFiwAGlpaUX28/f3x8mTJ3HixAns3LkTDRs2xPz58/H999+XeJtEVHYxFBFRmWVtbY3w8HAAwIULF6T2tLQ0dO7cGQkJCYiKikKDBg0M3sb8+fORkZGBefPmFdnH1tYWTZs2RbNmzRAWFoYDBw7Aw8MDkyZNKvajPSIqXxiKiKhMSElJKbT90qVLAIDq1asD+F8gunHjBvbv349GjRqVaruBgYEYMWIEVq1ahVu3bum1jqurKxYsWIC///4bq1atKtX2iajs4HWKiKhM6NatG7y8vNC7d28EBgYiPz8fZ8+exdKlS+Ho6IiJEyfi6dOn6NatG/7880+sWLECarUasbGx0hhubm7w9/cv8bZnzpyJ7777DtHR0XBwcNBrnWHDhmHZsmVYsmQJxo4dCycnpxJvl4jKFu4pIqIy4Z///CeqVKmC5cuXo0+fPujRowe++OILdO7cGSdOnEBwcDD+/vtvnDx5EkIITJw4Ea1atdK6/etf/zJo29WrV8ekSZNKtI6VlRUWLFiAhw8fYsWKFQZtl4jKFpkQQli6CCIiIiJL454iIiIiIjAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERgJfo4o35+fm4c+cOKlWqBJlMZulyiIiIyAyEEMjIyED16tVf+KXRL00ounPnDry9vS1dBhEREVlAUlISvLy8iu3z0oSiSpUqAXg2Kaa4HL9KpcL+/fvRtWtXWFtbG3180sU5Nz/Ouflxzs2Pc25+ppzz9PR0eHt7SzmgOC9NKNJ8ZObk5GSyUGRvbw8nJyf+EJkJ59z8OOfmxzk3P865+ZljzvU5dIYHWhMRERGBoYiIiIgIAEMREREREQCGIiIiIiIADEVEREREABiKiIiIiAAwFBEREREBYCgiIiIiAsBQRERERASAoYiIiIgIAEMREREREQCGIiIiIiIADEVEREREABiKiIiIiAAwFBEREREBYCgiIiIiAsBQRERERASAoYiIiIgIAEMREREREQCGIiIiIiIADEVEREREABiKiIiIiAAwFBEREREBYCgiIiIiAsBQRERERASAoYiIiIgIAEMREREREQCGIiIiIiIADEVEREREABiKiIiIiAAwFBEREREBYCgiIiIiAsBQRERERASAoYiIiIgIAEMREREREQCGIiIiIiIADEVEREREABiKiIiIiAAACksXQOVDwoNMZOaopeWcvGzcyUxEdQcfKOW2JRrLQamAX1UHY5dIRERUKgxF9EIJDzLRYckhrTYr29tw8FuFzITxyM+uUeIxoz8MZTAiIqIyhaGIXkizh2jFmyEIcHcEACSkX8EnJ4CVb4bAz6mO3mPF33uCSTvOau11IiIiKgsYikhvAe6OqF/DGQBgZfssHPm7OyLI1dmSZRERERkFD7QmIiIiAkMREREREQCGIiIiIiIADEVEREREABiKiIiIiAAwFFVoWVlZOHPmDLKysixdisVwDoiISF8MRRXY5cuX0aRJE1y+fNnSpVgM54CIiPTFUEREREQEhiIiIiIiAAxFRERERAD4NR/0ksnNzcWaNWtw/fp1+Pv7Y8yYMbCxsTHK2Hl5eTh69ChSUlJQrVo1vPrqq5DL5SXuY0zm3h4RUXlWLvYUCSHQuXNndOvWTee+NWvWwNnZGbdu3bJAZVSeTJ8+HQ4ODpg8eTJWr16NyZMnw8HBAdOnTy/12BEREQgICECHDh0wePBgdOjQAQEBAYiIiChRH2My9/aIiMq7chGKZDIZNm7ciD/++APffPON1J6QkICPPvoIK1euRM2aNS1YIZV1K1euxOLFi+Hq6op169YhJSUF69atg6urKxYvXlyqYBQREYGwsDAEBwcjJiYGGRkZiImJQXBwMMLCwhAREaFXH2My9/aIiCoEUY5s2rRJODo6ihs3boj8/HzRoUMH0bdvX73Wffz4sQAgHj9+bJLacnNzxe7du0Vubq5JxjfE6dOnBQBx+vTpUo1zPvmR8Plojzif/Ehqi3sQJ+pvqi/iHsSVeixD6TPnmjmQy+XCw8NDqFQqrftVKpXw8PAQCoVC5OTklLgGtVotfH19Re/evUVeXp7WfXl5eaJ3797C19f3hX38/PyEWq0u8fYNrcnQ7ZXF13lFxzk3P865+Zlyzkvy/l+ujikaPnw4fvrpJ7zzzjv4xz/+gQsXLuDChQuF9s3JyUFOTo60nJ6eDgBQqVRQqVRGr00zpinGNtSTp88e/5WUR7BKTDV4nOv3MwEAmU9zpMenVqulf0vymDOlmh5LYxhKrVYj6Qnw162HUCgKfylfSXkE4NmxNbNmzYIQQqfe8PBwjBkzBqtWrcKECRNKVMPhw4dx8+ZNbNmyBXl5ecjLy9O6f9q0aWjXrh0AvLBPdHQ02rdvX6Ltl6YmQ7ZXFl/nFR3n3Pw45+ZnyjkvyZjlKhQBwNq1a1G/fn0cPXoUu3btgru7e6H95s+fj1mzZum079+/H/b29iarLyoqymRjl9Tx89cBAFN2nofy6NNSj7fnUAzuuj37/x31HQDA78d+R4IiQe8xTt0HAAWm7jpf6nqeUQDnTxV5b87deOn/SqUSkZGROn1sbW0BAAcPHkRAQECJtn7kyBEAQHJyMlJTdYPn06f/m/cX9fntt9+QmZlZou2XpqbSbK8svc5fFpxz8+Ocm58p5rwk32hQ7kKRu7s73nvvPezevRv9+/cvst+MGTMwZcoUaTk9PR3e3t7o2rUrnJycjF6XSqVCVFQUunTpAmtra6OPb4gqVU9hEYBlrwejbv2GBo9z/X4mpu46j16hrdDYpwoA4NLDS1izbw3atG2Dui519R7LMzENW+JPYmlYMPzdHAyuCXi2pyg2NhYtW7Ysck/RpQt2GLz52f9zcnLQs2dPnT7r168HAHTs2LHQ+4vj4OCAZcuWwcvLCy1atNC5PzY2Vvr/i/r06NHDKHuK9K3JkO2Vxdd5Rcc5Nz/OufmZcs41nxTpo9yFIgBQKBRFvglqKJVKKJVKnXZra2uTvshNPX5JONo9e/x1qlVGiI+rweNo5trBTik9Nk2bQqEo0eN1kGpyRv0azgbXBDz7Ibp9AWhY06XIGvJTKwMA5HI5wsPDMXLkSK3XjlqtxqxZs6BQKDB+/PgSP3cdOnSAr68vFi1ahN27d8PK6n/nLuTn52Px4sXw9fUFgGL7+Pn5oUOHDkY5XV6fmkq7vbL0On9ZcM7Nj3NufqaY85KMVy7OPiMqrSFDhuDvv/+Gl5cX1q5dizt37mDt2rXw8vLC33//jcmTJxt0vSK5XI6lS5diz5496Nevn9aZXv369cOePXuwdOnSF/ZZsmSJ0a4fpE9NxtweEVFFUS73FBGV1MSJE+Hh4YHly5fj/fffl9oVCgWmTZuGRYsWGTz2gAEDsGvXLkydOhWtW7eW2v38/LBr1y4MGDAAAPTqYyz61kRERP/DUEQvjUWLFmHOnDkmuaL1gAED0Ldv32KvHq1PH2My9/aIiMq7chmKZs6ciZkzZ1q6DCqHbGxsMGnSJJOMLZfLERoaWuo+xmTu7RERlWc8poiIiIgIDEVEREREABiKiIiIiAAwFFVogYGBOH36NAIDAy1disVwDoiISF/l8kBr0o+9vT0aN25s6TIsinNARET64p4iIiIiIjAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAzz4jPTxV5QEALtx+LLUlpD8BAFy/9wT52Y8LXa8w8feeGLc4IiIiI2Eoohe6/v9B5uOI81Kble1tOPgBE3ecRX72/RKP6aDkS4+IiMoWvjPRC3Wt5wkA8Hd3hJ31s29Yz8nLxp3MEFTv4AOl3LZE4zkoFfCr6mD0OomIiEqDoYheyMXBBgOb1yzQ6owm8LBIPURERKbAA62JiIiIwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAoNC3o4uLS4kGlslkOHPmDHx8fEpcFBEREZG56R2KHj16hBUrVsDZ2fmFfYUQGDNmDPLy8kpVHBEREZG56B2KAGDgwIFwd3fXq+/48eMNKoiIiIjIEvQORfn5+SUaOCMjo8TFEBEREVkKD7QmIiIigoGhaPPmzdi7d6+0PH36dFSuXBmtW7dGYmKi0YojIiIiMheDQtG8efNgZ2cHAIiJicHq1auxaNEiVK1aFZMnTzZqgURERETmUKIDrTWSkpIQEBAAANi9ezfCwsLw3nvvoU2bNggNDTVmfURERERmYdCeIkdHR6SmpgIA9u/fj86dOwMAbG1t8fTpU+NVR0RERGQmBu0p6tKlC0aNGoVGjRrh6tWreO211wAAcXFx8PX1NWZ9RERERGZh0J6iL7/8Eq1atcL9+/fx448/wtXVFQBw+vRpDBo0yKgFEhEREZmDQXuKKleujNWrV+u0z5o1q9QFEREREVmC3qHo3LlzqF+/PqysrHDu3Lli+zZo0KDUhRERERGZk96hKCQkBHfv3oW7uztCQkIgk8kghJDu1yzLZDJ+5xkRERGVO3qHooSEBLi5uUn/JyIiIqpI9A5FPj4+hf6fiIiIqCLg13wQERERgV/zQURERASAX/NBREREBIBf80FEREQEgF/zQURERASAX/NBREREBIBf80FEREQEwMA9Rb6+vpg9ezaSkpKMXQ8RERGRRRgUiqZOnYqff/4Zfn5+6NKlC77//nvk5OQYuzYiIiIiszEoFI0fPx6nT5/G6dOnERQUhAkTJqBatWoYN24czpw5Y+waiYiIiEzOoFCk0bBhQ6xcuRK3b99GeHg41q9fj2bNmqFhw4b497//rfWFsURERERlmUEHWmuoVCr89NNP2LhxI6KiotCyZUuMHDkSd+7cwaeffor//ve/2LZtm7FqJSIiIjIZg0LRmTNnsHHjRmzfvh1yuRxvvfUWli9fjsDAQKlP165d0a5dO6MVSkRERGRKBoWiZs2aoUuXLvjqq6/Qr18/WFtb6/QJCgrCwIEDS10gERERkTkYFIpu3LgBHx+fYvs4ODhg48aNBhVFREREZG4GHWj9okBEREREVN7ovaeoSpUqkMlkevV9+PChwQURERERWYLeoWjFihUmLIOIiIjIsvQORcOHDzdlHUREREQWZfB1ivLy8rB7925cunQJMpkMQUFB6NOnD+RyuTHrIyIiIjILg0JRfHw8evbsidu3b6NOnToQQuDq1avw9vbG3r174e/vb+w6iYiIiEzKoFA0YcIE+Pv7IzY2Fi4uLgCA1NRUDB06FBMmTMDevXuNWiRVLAkPMpGZoy71OGq1GklPgLg76VAoFMjJy8adzERUd/CBUm5rhEpfzEGpgF9VB7Nsi4iITMugUHT48GGtQAQArq6uWLBgAdq0aWO04qjiSXiQiQ5LDhlxRAWWnI8FAFjZ3oaD3ypkJoxHfnYNI26jeNEfhjIYERFVAAaFIqVSiYyMDJ32J0+ewMbGptRFUcWl2UO04s0QBLg7lmostVqNY8eOoW3btlAoFEhIv4JPTgAr3wyBn1MdY5RbrPh7TzBpx1mj7PUiIiLLMygU9erVC++99x42bNiA5s2bAwD++OMPfPDBB+jTp49RC6SKKcDdEfVrOJdqDJVKhURHoF51J1hbW8PK9lnI8nd3RJBr6cYmIqKXj0FXtP7iiy/g7++PVq1awdbWFra2tmjTpg0CAgKwcuVKY9dIREREZHIG7SmqXLkyfv75Z1y7dg2XLl0C8OwLYAMCAoxaHBEREZG5GHydIgCoXbu2FIT0/QoQIiIiorLIoI/PAGDDhg2oX7++9PFZ/fr1sX79emPWRkRERGQ2Bu0p+uyzz7B8+XKMHz8erVq1AgDExMRg8uTJuHnzJubMmWPUIomIiIhMzaBQ9NVXX2HdunUYNGiQ1NanTx80aNAA48ePfylDUVZWFq5fv46srCw4O/PMJ6LnZWVl4fLlywgMDIS9vb2lyyEiKpRBH5/l5eWhadOmOu1NmjSBWv1yXrPlypUrmDp1Kq5cuWLpUojKnMuXL6NJkya4fPmypUshIiqSQaFo6NCh+Oqrr3Ta165diyFDhpS6KCIiIiJz0/vjsylTpkj/l8lkWL9+Pfbv34+WLVsCAGJjY5GUlIRhw4YZv0oiIiIiE9M7FP35559ay02aNAEAXL9+HQDg5uYGNzc3xMXFGbE8IiIiIvPQOxRFR0ebsg4iekns2rULr7/+urS8c+dO1KhRA61bt5bajh8/jlu3bmHgwIFS2+eff47Zs2dLy4cOHQIAhIaGarXVqFEDQUFBUKlUsLa2xsaNGzF06FCtsWvWrIng4GBkZGSgUqVKOH/+PJ4+faq13k8//YS+ffsiLy8PcrkccXFxcHZ2RkhICB49eoTKlSvj7NmzcHR0xFtvvYXr16/D398fc+fORYsWLZCVlQV7e3vExcUhNzdXa+yLFy/C0dFRa6xjx45h2rRp0jhbtmzBkydPEBISgocPH8LFxaXQPk+fPkXz5s1x//59uLm54cSJEwCg1bZv3z6EhYXhzp07qF69Og4fPozc3Fydx2JjY4P27dtL/fbu3YuJEydqbU8ul2PatGm4du0aateujcWLF0Mul2PNmjVSv4EDB6JLly5a27Ozs9NZD4BW29y5c7Fx40ZpnDFjxiA3N1drfrds2QIbGxut7b3zzjv49NNPix178eLFsLGxwdGjR5GSkoJq1aqhdevWOH78uLT86quvIj4+HvXq1dN63gMCAl64nlwu13mt5+bmatU5ZswYne8HzcvL0xq7RYsW+Oabb7TWkcvlWn0K297Tp091Hq+dnZ1BNRWmYJ2F1aBPn6LGPnz4MI4cOQIHBwd06NBBr/VMQljQ8OHDBQAxf/58rfaffvpJaEqLjo4WAKSbra2tCAoKEt98802JtvX48WMBQDx+/Nho9T/vjz/+EADEH3/8YZLxK4rzyY+Ez0d7xPnkR6UeKzc3V+zevVvk5uYKIYSIexAn6m+qL+IexJV6bH0Y87GUFwXnXF+nT5/W+jnmrWLdZDKZxWvQ52Zvb6+1rFAoDBqn4Hq+vr7ixx9/1HrNT5s2TaefQqEQ06ZNk/r8+OOPwtfXt9htWVlZCWdn52K317dv30LX7du3b4lrKkxhdRasQZ8+ho5dWiV5/zfoQOvs7GwsXrwYPXv2RNOmTdG4cWOtW0nY2tpi4cKFSEtLK7bflStXkJKSgosXL+L999/H6NGjceDAAUPKJ6Iy4Pk9Qxpvv/22QWO98cYbBq337bffGrTezp07tf66lsvlsLLS/nXq4+Ojs55mD8bz/P39tZaVSqXWX8k2Njbw8PDQ6aNQKLSW9+zZo/XNAlZWVqhcubLWenZ2djp9du7cqVW7TCZDnTp1tNZzcXGBl5eXtOzp6Yn4+HjpC8GFEGjYsCHc3Ny01nv99de1arC3t0d8fDxq164ttdWuXRujRo3SWq9bt27w9fWVloOCgvDXX38hKChIavP19UW3bt201hs1apTO2PHx8ejfvz+AZ5eGePXVV7F161YAgKurK2QymbSsYW1tjUGDBsHa2lqrfevWrZDJZHB1dZWWY2JiEBwcjLCwMERERAAApk+fjsWLF8PV1RXr1q1DSkoK1q1bB1dXVyxevBjTp09HREQEwsLCEBwcjJiYGEycOBHAs+dbc9zu6NGjkZ+fj8ePH2PIkCHIyMjQ2V6/fv3w888/w8bGBh9//DHi4+Px8ccfw8bGBj///DP69eund02FKVhnYTXo0+dFYx89ehTbt2/H0aNHX7ieSRmSugYNGiSqVq0qPvjgAxEeHi5mzpypddPX8OHDRa9evURgYKBWUi1sT1FaWprWurVq1RKLFi3Se1vcU1Q2cE9R+WbonqKFCxdKfwXGxMQIIYQ4fvy4zl+2kyZNKvIv5nXr1hV536ZNm4q87/Dhw0Xet3Xr1iLvK27v1qlTpwRQ+B6SpKSkIvecaH5PFLzJZDKRkpJS7B6DF/W5ceNGkffdvn27yPuuXbtW5Pbu37+v1SaXy6XHlpqaKhQKhXB3dxfW1tZa/dLS0oSHh4eQy+XSepr7kpOTBQBhY2MjrSeXy4WHh4fIzMwUNjY2WmO5u7sLhUIhMjIypO09f7+NjY3IzMwUHh4ewsrKSmfsjIwM4evrK1577TVp7Jo1a4revXsLlUolevfuLWrUqCGN17lzZ+Hn5yd++OEH4evrKzp27CjdV716da31/Pz8hFqtFnl5edJyVlaWUCgUwsPDQ6hUKq2fA5VKJc2Lj4+P6N27t8jLyxM5OTnSOjk5OdJYPj4+olevXtIc5OTkCCGEtD0fHx/p8Wru08jJyZEe76NHj15Y0/Pja6jVauHr6yvV+bznH/Pzj6WoPmq1utixn//dUtx6hijJ+79BF2/cu3cvIiMj0aZNG0NW1yKXyzFv3jwMHjwYEyZM0PprpDBCCPznP/9BUlISWrRoUWS/nJwc5OTkSMvp6ekAAJVKBZVKVeq6C3ry9Nm2rqY8hk1iqtHHryiu388EAGQ+zSn186BZX/Ov5hpZarXaJM9xQZn//5xfSXn80lyfS61WI+kJ8Neth1p7Kl7ko48+kv7fpEkTqFSqQvcUrVixosgxhg8fjnfffbfQ+zR7mOzt7ZGVlaV1X6tWrbTaPT09cffuXQCQjjWyt7eHk5OT1K5UKhEcHKy1nq+vL1JSUpCTkyNdp23IkCE4evQoEhMTATzbo+Lh4YFBgwZh27Zt0ti2trZ4+PCh9DvL19cX9+7dk8YeMmQIXF1d4ePjI43l6+uL1q1bS+MMGjRIp09QUBBCQkKwbds2aS9KUFAQhBDSl3W3aNECbm5uqFu3rtTm6+sLNzc3nDx5UlqvRYsWuHv3rjT2kCFD4OzsjGbNmuHkyZMAgKlTpyI5ORnbtm3Dq6++CrVajdmzZyM+Ph5LliyRxnFwcEB4eDjGjBkjrRcdHY2TJ08iODgYADBp0iTk5+djyZIlyMvLw6xZs2BtbY0JEyZIYw0dOhStW7fGmDFjMHjwYGl7R44ckeZl0qRJsLa21tre82MPHjwYN2/exJYtW1C3bl0sWbIEt27dwnfffQchBKZNm4Z27dpJz/vnn3+Odu3aYd++fdJ6v//+O3JycnDnzh3s2LFDa73o6Gi0b99eWp4yZQrUajVmzZoFIYTO7yJNnYmJidi6dSvy8vKwatUqaR2ZTKZV09atW3H+/HmMGTMGq1atwoQJEwBAq8+kSZMgk8m0tiWTyaS57NGjh141PT8+ABw+fFiag7y8POTl5WmtV7DO4vpo5qmosQv+Pi9qPUOU5P3AoFBUo0YNVKpUyZBVC9W/f3+EhIQgPDwcGzZsKLSPJizl5OQgPz8fs2fPlp6MwsyfPx+zZs3Sad+/f79Jrqh7/Pyzs/Cm/XQRyphco49f0ew5FIO7bi/up4+oqCgAwB31HQDA78d+R4IiwTiDF+PUfQBQYOqu8ybfVtmiAM6fMmhN9+o1ERkZqdUWGhqKxMREJCT87zkLCAiAu7s7jh8/DuBZsImMjNQKBK1atYJKpcKpU/+rJSwsDBcvXpTaQkNDERkZib59+2L79u0Ann3UdurUKWlszXqOjo5Ys2YNAGDAgAE664WFheHmzZvYtWuXtF6TJk1QtWpVKcy98cYbiIyMRJMmTaQ37bCwMNjY2Gh9N2RYWBhu374tjd2kSRNERkaif//+0lhhYWGoVq2aNE6zZs10+vTr1w+urq7Ytm0bsrOzpba8vDwpAPXs2RORkZHo06eP1BYWFgZ7e3ucPHlSesPo2bMn0tLSpLE1NfXo0UMKRX5+fvDw8MC2bduQlJQE4FmQ8PPzkx6bZnu2trZSm5+fH2xtbXHy5EnpD1Q/Pz/k5+dLfZRKJSIjI7XGatKkiTTOuXPnpH7Pz6+fn1+h29OMrVkvOTlZa+zk5GSkpqbi6dOnUtuAAQNw+/ZtAJACcnJyMnr37i097wXX++2335CZmSktx8bGaj2egp6vUzPWwYMHtdZ5vqbk5GRpnYMHD0pfwv58H80cFKR5vNeuXdOrpufHB4AjR45o1VlQwTqL66OZpxeNrfl9XtR6hij4h1JxDApFS5cuxUcffYSvv/660M/NDbFw4UJ07NgRU6dOLfT+o0ePolKlSsjJycGJEycwbtw4uLi4YPTo0YX2nzFjhta1ldLT0+Ht7Y2uXbvCycnJKDU/r1LlWCwCsLh/EOqHlOy4qpfJ9fuZmLrrPHqFtkJjnyqlGkulUiEqKgpdunSBtbU1Lj28hDX71qBN2zao61LXSBUXzTMxDVviT2JpWDD83RxMvr2yQK1WIzY2Fi1btizRnqJGC5/9e+/OLfTs2VPrPs1ZZM+Lj49HfHy8tBwTE4OePXtKx0do2gratWuX1i/AQ4cOYf/+/Rg8eLDU9sMPP0hveM+v9/zvhYiICGzevFlrvV27diElJUVrvdOnT+Po0aNaY3/xxRdax0bt2rVL641Q03bv3j2tccaOHSsdV6Lp8/zetJMnT+r02b17N0JCQgA8e2PLzs7G7t27IYSQ+kRGRmLGjBmYMWOG1tia43+sra2hUqkQGRmpNS+amubOnSu1JSQkIDk5GQDg7e2NixcvIicnRyvQarb3fAhMSEiQzmB2cnJCWloaEhIStEJRTk4OevbsiU8++USrBs0cNGjQAImJicjJycHp06e1xh45cqTO9jRja9bz8vLSWs/LywstWrSQQgzw7HnXvKd4enpK/X799dci1+vRowfat28vLbds2RJnz56VHk9Bz9epGSs+Ph6RkZHSOs/X5OXlhfPnn/3h1bFjR2nM5/to5qAgzVzWrl0bDx48eGFNz48PAA4ODli2bJlUZ0EF6yyuj2aeihq74O/zotYzhCaI68WQz+fu3bsnQkNDhZWVlXB0dBRVqlTRuulr+PDhWkfH9+zZU/Tt21evY4ref/99UaNGDb23xWOKygYeU1S+8ZgiHlPEY4p4TNHzNVS0Y4oMCkWdOnUStWvXFgsWLBAbN24UmzZt0rrpq2AoOnfunLCyshLTpk17YSgaM2aMcHFx0XtbDEVlA0NR+WbMU/KbNm2q0zZw4MBig0FRtz59+hi03vr16w1a77vvvtN685bL5dIbsubm7e2ts96cOXN02vz8/LSWlUql1lg2NjY6QUCpVGqdWq1UKkVERIRWGCvsVG47OzutPjKZTHz33Xc6bbVr19Zaz8XFRSs4eHp6iitXrohmzZpJbQ0bNhRubm5a64WFhWnVYG9vL65cuSL8/f2lNn9/fzFy5Eit9bp16ya94QMQdevWFadPnxZBQUFSm4+Pj+jWrZvWeiNHjhQBAQFaY1+5ckXrlPVXX31VCsMeHh5CJpPphGOFQiHCwsJ0Tl/funWrkMlkwsPDQ1o+fvy46N27t5DJZNJp5Jr3MA8PD/HNN9+I27dvi2+++UZab9q0aeLHH38UMplM9O7dWxw/flxMnDhRer5lMplYt26deP/996VtDx48WKSnp+tsT/PYbGxsxPTp08WVK1fE9OnTpden5j1Wn5oKU7DOwmrQp8+Lxj5y5IjYvn27OHLkyAvXKymThyI7Oztx9uxZQ1bVUjAUCSHEW2+9JWxtbQWgHYquXLkiUlJSxM2bN8UPP/wgKlWqJN555x29t8VQVDYwFJVvvE4Rb4XdXvbrFPn5+Zn1OkUFt2eJ6xQVrEGfPoaOXVomD0WNGjWSdoGXRmGh6ObNm0KpVAqg8Is3KhQK4efnJz788EPx5MkTvbfFUFQ2MBSVb6UNRadPnxY7d+7U+pneuXOnzkdpx48fF99//71W2+eff661fOjQIXHo0CGdtmvXrkkfnVhbW+vsBTh+/LhITk4WVapUEQqFQlSpUkUkJyfrrLdnzx6tj38uX74sUlJShIeHh1AqlcLDw0OkpKSIjIwM0a9fPxEcHCz69esn4uLihKOjo3R4QWJios7Y165d0xnr2rVrWuNkZGRIfaytrYvsc+/ePeHr6yscHByEr6+vuHfvnk7b5cuXRf369YWLi4uoX7++SE1NLfSxpKamavVLTEzU2V5WVpYYO3as6Nq1qxg7dqzIysoSOTk5Yvny5WLcuHFi+fLlIiUlRWd7ha1XsO3Ro0da4+Tk5OjMb0ZGhs72Hj169MKxs7KyhFqtFtHR0WLbtm0iOjpa5OTkaC2r1Wpx+fJlneddn/UKU7DOgh9PCSF0xs7KytJZp2CfwrZX2OM1tKbC6FODPn2KGjsqKkpMmTJFREVFGeUjs+eZPBT95z//Ea1btxbR0dHiwYMH4vHjx1q3soihqGxgKCrfjBGKqGQMnXMyHOfc/Ew55ya/TlH37t0BAJ06ddJqF0JAJpPpXKeAiIiIqKwzKBTxy2GJiIioojEoFJX2mgFEREREZY1BoQgAHj16hA0bNuDSpUuQyWQICgrCiBEj4OzsbMz6yo06depg6dKlOl+kSERAYGAgTp8+jcDAQEuXQkRUJKsXd9F16tQp+Pv7Y/ny5Xj48CEePHiAZcuWwd/fH2fOnDF2jeWCvb09/P39TfIVIkTlnb29PRo3bsyfDyIq0wzaUzR58mT06dMH69atky71r1arMWrUKEyaNEn6ThMiIiKi8sKgUHTq1CmtQAQACoUC06dPl749moiIiKg8MejjMycnJ9y6dUunPSkpCZUqVSp1UURERETmZlAoevPNNzFy5Ejs2LEDSUlJSE5Oxvfff49Ro0Zh0KBBxq6RiIiIyOQM+vhsyZIlkMlkGDZsGNRqNQDA2toao0ePxoIFC4xaIBEREZE5lDgU5eXlISYmBuHh4Zg/fz6uX78OIQQCAgJ4Zgm90FPVs6udX7j9uNRjqdVqJD0B4u6kQ6FQICH9CQDg+r0nyM8u/fgvEn/vicm3QURE5lPiUCSXy9GtWzdcunQJLi4uCA4ONkVdVEFd//8g8XHEeSONqMCS87EAACvb23DwAybuOIv87PtGGv/FHJQGX+6LiIjKEIN+mwcHB+PGjRvw8/Mzdj1UwXWt5wkA8Hd3hJ21vFRjqdVqHDt2DG3btoVCoUBOXjbuZIagegcfKOW2xij3hRyUCvhVdTDLtoiIyLQMCkVz587Fhx9+iH/9619o0qQJHBy03xScnJyMUhxVPC4ONhjYvKZRxlKpVEh0BOpVd4K1tTUAZzSBh1HGJiKil49Boah79+4AgD59+kAmk0ntQgjIZDLk5eUZpzoiIiIiMzEoFEVHRxu7DiIiIiKLMigUtW/f3th1EBEREVmUwafNPHr0CCdOnMC9e/eQn5+vdd+wYcNKXRgRERGRORkUin799VcMGTIEmZmZqFSpktZxRZqLOhIRERGVJwZ9zcfUqVMxYsQIZGRk4NGjR0hLS5NuDx8+NHaNRERERCZnUCi6ffs2JkyYwCtYExERUYVhUCjq1q0bTp06ZexaiIiIiCxG72OKfvnlF+n/r732GqZNm4aLFy8iODj4/y+c9z99+vQxXoVEREREZqB3KOrXr59O2+zZs3XaePFGIiIiKo/0DkUFT7snIiIiqkhKdExRx44d8ejRIxOVQkRERGQ5JQpFhw4dQm5urqlqISIiIrIYg84+IyIiIqpoSnxF64yMDNja2hbbx8nJyeCCiIiIiCyhxKHolVdeKfI+IQTPPiMiIqJyqcShaNeuXXBxcTFFLUREREQWU+JQ1KZNG7i7u5uiFiIiIiKL4YHWRERERChhKPLx8YFcLjdVLUREREQWU6KPzxISEkxVBxEREZFFlfjjs7/++gtz5szBmjVr8ODBA6370tPTMWLECKMVR0RERGQuJQpF+/fvR/PmzfH9999j4cKFqFu3LqKjo6X7nz59is2bNxu9SCIiIiJTK1EomjlzJj788ENcuHABN2/exPTp09GnTx/s27fPVPURERERmUWJjimKi4vDli1bAAAymQzTpk2Dl5cXwsLCsH37djRv3twkRRIRERGZWolCkVKpxKNHj7TaBg0aBCsrKwwcOBBLly41Zm1EREREZlOiUBQSEoLo6Gg0adJEq/3NN99Efn4+hg8fbtTiiIiIiMylRKFo9OjROHLkSKH3DRo0CACwdu3a0ldFREREZGYlCkX9+/dH//79i7x/0KBBUjgiIiIiKk9K/N1nBWVnZ2PHjh3IzMxEly5dULt2bWPURURERGRWJQpF06ZNQ25uLlauXAkAyM3NRatWrRAXFwd7e3tMnz4d+/fvR+vWrU1SLBEREZGplOg6Rb/99hs6deokLX/33XdITEzEtWvXkJaWhtdffx1z5841epFEREREplaiUHTr1i0EBQVJy/v370dYWBh8fHwgk8kwceJE/Pnnn0YvkoiIiMjUShSKrKysIISQlmNjY9GyZUtpuXLlykhLSzNedURERERmUqJQFBgYiF9//RXAs6tb37p1Cx06dJDuT0xMhIeHh3ErJCIiIjKDEh9oPWjQIOzduxdxcXHo2bMn/Pz8pPsjIyP5VR9ERERULpVoT9E//vEPREZGokGDBpg8eTJ27Nihdb+9vT3GjBlj1AKJiIiIzEHvPUXnzp1D/fr10blzZ3Tu3LnQPuHh4dL/4+LiUKdOHSgUpb4UEhEREZHJ6b2nqFGjRkhNTdV74FatWuHWrVsGFUVERERkbnrvxhFC4LPPPoO9vb1e/XNzcw0uioiIiMjc9A5F7dq1w5UrV/QeuFWrVrCzszOoKCIiIiJz0zsUHTp0yIRlEBEREVlWic4+IyIiIqqoGIqIiIiIwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgCAwtIFUMWS8CATmTlqs2xLrVYj6QkQdycdCkXhL+WcvGzcyUxEdQcfKOW2RY7loFTAr6qDqUolIqJygKGIjCbhQSY6LDlk5q0qsOR8bJH3WtnehoPfKmQmjEd+do1iR4r+MJTBiIjoJcZQREaj2UO04s0QBLg7mnx7arUax44dQ9u2bYvcU5SQfgWfnABWvhkCP6c6hfaJv/cEk3acNdseLiIiKpsYisjoAtwdUb+Gs8m3o1KpkOgI1KvuBGtr60L7WNk+C2f+7o4IcjV9TUREVH7xQGsiIiIiMBQRERERAWAoIiIiIgLAUEREREQEgKGIiIiICABD0UsvKysLZ86cQVZWlqVLISPg80lEZDiGopfc5cuX0aRJE1y+fNnSpZAR8PkkIjIcQxERERERGIqIiIiIADAUEREREQHg13wQEdFLIi8vD0ePHkVKSgqqVauGV199FXK53NJlURlSLvYUvf3225DJZFiwYIFW++7duyGTySxUFRERlRcREREICAhAhw4dMHjwYHTo0AEBAQGIiIiwdGlUhpSLUAQAtra2WLhwIdLS0ixdChERlSMREREICwtDcHAwYmJikJGRgZiYGAQHByMsLIzBiCTlJhR17twZnp6emD9/vqVLISKiciIvLw9Tp05Fr169sHv3brRs2RKOjo5o2bIldu/ejV69euHDDz9EXl6epUulMqDcHFMkl8sxb948DB48GBMmTICXl1ex/XNycpCTkyMtp6enAwBUKhVUKpXR69OMaYqxTenJ02dzdCXlEawSU0s11vX7mQCAzKc5ZpkHfeZcrVZL/xbVL1Oag8dS//LqSsojAM+eV77OKwbOeekcPnwYN2/exJYtW5CXl6cTfqZNm4Z27dohOjoa7du3B8A5twRTznlJxiw3oQgA+vfvj5CQEISHh2PDhg3F9p0/fz5mzZql075//37Y29ubqkRERUWZbGxTOH7+OgBgys7zUB59apQx9xyKwV03owyll+Lm/I76DgDg92O/I0GRUGifU/cBQIGpu86boDrzyrkbDwDYGx2DtAf3TLad8vY6rwg454Y5cuQIACA5ORmpqbp/+D19+uz33m+//YbMzEyt+zjn5meKOS/JFf7LVSgCgIULF6Jjx46YOnVqsf1mzJiBKVOmSMvp6enw9vZG165d4eTkZPS6VCoVoqKi0KVLF1hbWxt9fFOpUvUUFgFY9now6tZvWKqxrt/PxNRd59ErtBUa+1QxToHF0GfOLz28hDX71qBN2zao61K30D6eiWnYEn8SS8OC4e/mYMqSTe7SBTsM3gy81qEVWjVvavTxy+vrvDzjnJeOg4MDli1bBi8vL7Ro0ULn/tjYWABAjx49tPYUcc7Ny5RzrvmkSB/lLhS1a9cO3bp1wyeffIK33367yH5KpRJKpVKn3dra2qQvclOPb2yOds/mqE61ygjxcS3VWArFs5eTg53SrHNQ3JxralIoFEX2cZDmwBn1azibpkgzyU+tDODZ88rXecXCOTdMhw4d4Ovri0WLFmH37t2wsvrfobT5+flYvHgx/Pz80KFDB53T8znn5meKOS/JeOXmQOvnLViwAL/++iuOHz9u6VKIiKgMk8vlWLp0Kfbs2YN+/fppnX3Wr18/7NmzB0uWLOH1ighAOdxTBADBwcEYMmQIVq1aZelSiIiojBswYAB27dqFqVOnonXr1lK7n58fdu3ahQEDBliwOipLymUoAoB//etf+OGHHyxdBhERlQMDBgxA3759eUVrKla5CEWbNm3SafPx8UF2drb5iyEionJJLpcjNDTU0mVQGVYujykiIiIiMjaGIiIiIiIwFBEREREBYCh66QUGBuL06dMIDAy0dClkBHw+iYgMVy4OtCbTsbe3R+PGjS1dBhkJn08iIsNxTxERERERGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIBnn5ERPVXlAQAu3H5slu2p1WokPQHi7qRDoSj8pZyQ/gQAcP3eE+RnF15X/L0nJquRiIjKD4YiMprr/x8uPo44b8atKrDkfGyR91rZ3oaDHzBxx1nkZ98vdiQHJX8ciIheZnwXIKPpWs8TAODv7gg7a9N/87RarcaxY8fQtm3bIvcU5eRl405mCKp38IFSblvkWA5KBfyqOpiqVCIiKgcYishoXBxsMLB5TbNtT6VSIdERqFfdCdbW1kX0ckYTeJitJiIiKr94oDURERERGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAQxERERERAIYiIiIiIgAMRUREREQAGIqIiIiIADAUEREREQFgKCIiIiICwFBEREREBIChiIiIiAgAoLB0AeYihAAApKenm2R8lUqFrKwspKenw9ra2iTbIG2cc/PjnJsf59z8OOfmZ8o517zva3JAcV6aUJSRkQEA8Pb2tnAlREREZG4ZGRlwdnYuto9M6BOdKoD8/HzcuXMHlSpVgkwmM/r46enp8Pb2RlJSEpycnIw+PuninJsf59z8OOfmxzk3P1POuRACGRkZqF69Oqysij9q6KXZU2RlZQUvLy+Tb8fJyYk/RGbGOTc/zrn5cc7Nj3Nufqaa8xftIdLggdZEREREYCgiIiIiAsBQZDRKpRLh4eFQKpWWLuWlwTk3P865+XHOzY9zbn5lZc5fmgOtiYiIiIrDPUVEREREYCgiIiIiAsBQRERERASAoYiIiIgIAEORUaxZswZ+fn6wtbVFkyZNcPToUUuXVGHNnz8fzZo1Q6VKleDu7o5+/frhypUrli7rpTJ//nzIZDJMmjTJ0qVUaLdv38bQoUPh6uoKe3t7hISE4PTp05Yuq8JSq9X45z//CT8/P9jZ2aFWrVqYPXs28vPzLV1ahXLkyBH07t0b1atXh0wmw+7du7XuF0Jg5syZqF69Ouzs7BAaGoq4uDiz1cdQVEo7duzApEmT8Omnn+LPP//Eq6++ih49euDWrVuWLq1COnz4MMaOHYvY2FhERUVBrVaja9euyMzMtHRpL4WTJ09i7dq1aNCggaVLqdDS0tLQpk0bWFtb47fffsPFixexdOlSVK5c2dKlVVgLFy7E119/jdWrV+PSpUtYtGgRFi9ejFWrVlm6tAolMzMTDRs2xOrVqwu9f9GiRVi2bBlWr16NkydPwtPTE126dJG+v9TkBJVK8+bNxQcffKDVFhgYKD7++GMLVfRyuXfvngAgDh8+bOlSKryMjAxRu3ZtERUVJdq3by8mTpxo6ZIqrI8++ki0bdvW0mW8VF577TUxYsQIrbYBAwaIoUOHWqiiig+A+Omnn6Tl/Px84enpKRYsWCC1ZWdnC2dnZ/H111+bpSbuKSqF3NxcnD59Gl27dtVq79q1K44fP26hql4ujx8/BgC4uLhYuJKKb+zYsXjttdfQuXNnS5dS4f3yyy9o2rQpXn/9dbi7u6NRo0ZYt26dpcuq0Nq2bYsDBw7g6tWrAIC//voLx44dQ8+ePS1c2csjISEBd+/e1XpPVSqVaN++vdneU1+aL4Q1hQcPHiAvLw8eHh5a7R4eHrh7966Fqnp5CCEwZcoUtG3bFvXr17d0ORXa999/jzNnzuDkyZOWLuWlcOPGDXz11VeYMmUKPvnkE5w4cQITJkyAUqnEsGHDLF1ehfTRRx/h8ePHCAwMhFwuR15eHubOnYtBgwZZurSXhuZ9s7D31MTERLPUwFBkBDKZTGtZCKHTRsY3btw4nDt3DseOHbN0KRVaUlISJk6ciP3798PW1tbS5bwU8vPz0bRpU8ybNw8A0KhRI8TFxeGrr75iKDKRHTt2YOvWrdi2bRvq1auHs2fPYtKkSahevTqGDx9u6fJeKpZ8T2UoKoWqVatCLpfr7BW6d++eTtIl4xo/fjx++eUXHDlyBF5eXpYup0I7ffo07t27hyZNmkhteXl5OHLkCFavXo2cnBzI5XILVljxVKtWDUFBQVptdevWxY8//mihiiq+adOm4eOPP8bAgQMBAMHBwUhMTMT8+fMZiszE09MTwLM9RtWqVZPazfmeymOKSsHGxgZNmjRBVFSUVntUVBRat25toaoqNiEExo0bh4iICBw8eBB+fn6WLqnC69SpE86fP4+zZ89Kt6ZNm2LIkCE4e/YsA5EJtGnTRudSE1evXoWPj4+FKqr4srKyYGWl/ZYol8t5Sr4Z+fn5wdPTU+s9NTc3F4cPHzbbeyr3FJXSlClT8NZbb6Fp06Zo1aoV1q5di1u3buGDDz6wdGkV0tixY7Ft2zb8/PPPqFSpkrSXztnZGXZ2dhaurmKqVKmSzjFbDg4OcHV15bFcJjJ58mS0bt0a8+bNwxtvvIETJ05g7dq1WLt2raVLq7B69+6NuXPnombNmqhXrx7+/PNPLFu2DCNGjLB0aRXKkydPEB8fLy0nJCTg7NmzcHFxQc2aNTFp0iTMmzcPtWvXRu3atTFv3jzY29tj8ODB5inQLOe4VXBffvml8PHxETY2NqJx48Y8PdyEABR627hxo6VLe6nwlHzT+/XXX0X9+vWFUqkUgYGBYu3atZYuqUJLT08XEydOFDVr1hS2traiVq1a4tNPPxU5OTmWLq1CiY6OLvR3+PDhw4UQz07LDw8PF56enkKpVIp27dqJ8+fPm60+mRBCmCd+EREREZVdPKaIiIiICAxFRERERAAYioiIiIgAMBQRERERAWAoIiIiIgLAUEREREQEgKGIiIiICABDEREREREAhiIiIiIiAAxFRERERAAYioiIiIgAMBQRERERAQD+D8Vk67ntKSYmAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Extract INR results and whether received thrombolysis\n",
    "inr_thrombolysis = raw_data.loc[\n",
    "    ~raw_data['S2INR'].isnull(),\n",
    "    ['S2INR', 'S2Thrombolysis']]\n",
    "\n",
    "inr_thrombolysis.boxplot(by='S2Thrombolysis', vert=False)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "sam10",
   "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": "3.10.11"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}