{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "46b52bcd",
   "metadata": {},
   "source": [
    "# Will the patient show up?\n",
    "\n",
    "**Project 1 of 10 · Softmaxima.** In this notebook we build, step by step, a model that predicts which medical\n",
    "appointments will be missed, and turn its predictions into a call list a clinic can actually use.\n",
    "\n",
    "Everything happens in this one notebook, the way you would explore and model in practice: look at the data, clean\n",
    "it, split it, engineer features, try models, calibrate, choose a threshold from costs, test once, check fairness and\n",
    "explain. It needs nothing but the notebook itself: the data is downloaded in step 3.\n",
    "\n",
    "**How to run it**\n",
    "* **Google Colab:** open [colab.research.google.com](https://colab.research.google.com), choose *Upload* and pick this\n",
    "  notebook, then *Runtime → Run all*. Everything it needs is already installed there.\n",
    "* **On your computer:** `pip install pandas scikit-learn matplotlib kagglehub`, then open it in Jupyter or VS Code.\n",
    "\n",
    "> The outputs saved in this notebook come from **synthetic practice data with the same columns as the real dataset**,\n",
    "> so the numbers are illustrations, not results. Run it yourself to download the real data and get real results."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "141986eb",
   "metadata": {},
   "source": [
    "## Set up\n",
    "One place for every import and setting. Fixing the random seed makes every run reproducible."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "scikit-learn 1.9.1\n"
     ]
    }
   ],
   "id": "aa41d113",
   "metadata": {},
   "source": [
    "import warnings\n",
    "from pathlib import Path\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import sklearn\n",
    "\n",
    "assert tuple(map(int, sklearn.__version__.split(\".\")[:2])) >= (1, 6), (\n",
    "    \"needs scikit-learn 1.6+: pip install -U scikit-learn\"\n",
    ")\n",
    "warnings.filterwarnings(\"ignore\", category=FutureWarning)\n",
    "pd.set_option(\"display.max_columns\", 30)\n",
    "plt.rcParams.update({\"figure.dpi\": 110, \"axes.spines.top\": False, \"axes.spines.right\": False})\n",
    "\n",
    "SEED = 42\n",
    "print(\"scikit-learn\", sklearn.__version__)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "796c1586",
   "metadata": {},
   "source": [
    "## Frame the problem\n",
    "Before touching data, write down who uses the model, what decision it supports, and what a mistake costs.\n",
    "\n",
    "| | |\n",
    "|---|---|\n",
    "| **User** | Clinic reception staff, each afternoon |\n",
    "| **Decision** | Which of tomorrow's (and later) patients to phone with a reminder |\n",
    "| **Prediction time** | When the appointment is booked, using only what is known then |\n",
    "| **Not for** | Refusing, delaying or overbooking care for anyone |\n",
    "\n",
    "A call costs staff time; a missed appointment wastes a slot. With rough costs agreed with the clinic, we can work\n",
    "out when a call is worth making: a call pays off when *probability of no-show × share of no-shows a call prevents\n",
    "× value of the slot* is larger than the cost of the call."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "A call is worth it when the no-show probability is above 0.278 (27.8%)\n"
     ]
    }
   ],
   "id": "fe740331",
   "metadata": {},
   "source": [
    "COST_OF_CALL = 5.0  # AUD of staff time per reminder call\n",
    "COST_OF_NO_SHOW = 60.0  # AUD value of a wasted slot\n",
    "CALL_PREVENTS = 0.30  # share of would-be no-shows that a call turns into attendance\n",
    "\n",
    "break_even = COST_OF_CALL / (CALL_PREVENTS * COST_OF_NO_SHOW)\n",
    "print(f\"A call is worth it when the no-show probability is above {break_even:.3f} ({break_even:.1%})\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b698bbb2",
   "metadata": {},
   "source": [
    "That number is the decision threshold, set by the business before we see any model. It also tells us that\n",
    "**probabilities must be accurate** (calibrated), not just good at ranking, because we compare them to 27.8%."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f13294d7",
   "metadata": {},
   "source": [
    "## Load the data\n",
    "The real data is the Kaggle *Medical Appointment No Shows* dataset (110,527 appointments in Vitória, Brazil, 2016,\n",
    "licence CC BY-NC-SA 4.0). The next cell looks for the CSV next to the notebook, otherwise downloads it with\n",
    "`kagglehub` (preinstalled in Colab; if it asks you to sign in, add a free Kaggle API key, or download the CSV from the\n",
    "Kaggle page and drag it into Colab's Files panel).\n",
    "\n",
    "If neither works, for example with no internet, it falls back to **practice data**: a synthetic table with the same\n",
    "columns, the same quirks and similar patterns, made by the function below. Never report numbers from practice data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "outputs": [],
   "id": "2d7521ee",
   "metadata": {},
   "source": [
    "def make_practice_data(rows=110_527, seed=SEED):\n",
    "    \"\"\"Synthetic appointments with the real file's columns and quirks (for practice only).\"\"\"\n",
    "    rng = np.random.default_rng(seed)\n",
    "    hoods = [f\"BAIRRO {i:02d}\" for i in range(1, 41)]\n",
    "    n_patients = max(50, int(rows * 0.56))\n",
    "    pat_ids = rng.integers(10**11, 10**15, n_patients).astype(float)\n",
    "    pat_age = np.clip(rng.gamma(2.2, 17.0, n_patients).round(), 0, 110).astype(int)\n",
    "    pat_gender = rng.choice([\"F\", \"M\"], n_patients, p=[0.65, 0.35])\n",
    "    pat_hood = rng.choice(len(hoods), n_patients, p=np.random.default_rng(1).dirichlet(np.ones(len(hoods))))\n",
    "    pat_tendency = rng.normal(0, 0.9, n_patients)  # unobserved: some patients miss more often\n",
    "    pat_scholar = rng.random(n_patients) < 0.10\n",
    "    pat_hyper = rng.random(n_patients) < np.clip(pat_age / 140, 0, 0.6)\n",
    "    pat_diab = rng.random(n_patients) < np.clip(pat_age / 400, 0, 0.2)\n",
    "    pat_alc = rng.random(n_patients) < 0.03\n",
    "    pat_hand = rng.choice([0, 1, 2, 3, 4], n_patients, p=[0.979, 0.018, 0.0025, 0.0004, 0.0001])\n",
    "\n",
    "    who = rng.choice(n_patients, rows, p=None)\n",
    "    appt = pd.Timestamp(\"2016-04-29\") + pd.to_timedelta(rng.integers(0, 41, rows), unit=\"D\")\n",
    "    appt = appt.where(appt.dayofweek != 6, appt + pd.Timedelta(days=1))  # no Sunday appointments\n",
    "    lead = np.where(rng.random(rows) < 0.35, 0, np.ceil(rng.exponential(14, rows))).astype(int)\n",
    "    sched = (\n",
    "        appt\n",
    "        - pd.to_timedelta(lead, unit=\"D\")\n",
    "        + pd.to_timedelta(rng.integers(7 * 3600, 19 * 3600, rows), unit=\"s\")\n",
    "    )\n",
    "    sms = ((lead >= 3) & (rng.random(rows) < 0.55)).astype(int)\n",
    "    age = pat_age[who]\n",
    "    z = (\n",
    "        -1.85\n",
    "        + 0.022 * np.minimum(lead, 60)\n",
    "        - 0.012 * (age - 40)\n",
    "        + 0.35 * pat_scholar[who]\n",
    "        + 0.3 * pat_alc[who]\n",
    "        - 0.25 * sms\n",
    "        + pat_tendency[who]\n",
    "        + 0.15 * (appt.dayofweek.to_numpy() == 0)\n",
    "    )\n",
    "    no_show = rng.random(rows) < 1 / (1 + np.exp(-z))\n",
    "    df = pd.DataFrame(\n",
    "        {\n",
    "            \"PatientId\": pat_ids[who],\n",
    "            \"AppointmentID\": rng.permutation(np.arange(5_000_000, 5_000_000 + rows)),\n",
    "            \"Gender\": pat_gender[who],\n",
    "            \"ScheduledDay\": sched.strftime(\"%Y-%m-%dT%H:%M:%SZ\"),\n",
    "            \"AppointmentDay\": appt.strftime(\"%Y-%m-%dT00:00:00Z\"),\n",
    "            \"Age\": age,\n",
    "            \"Neighbourhood\": np.array(hoods)[pat_hood[who]],\n",
    "            \"Scholarship\": pat_scholar[who].astype(int),\n",
    "            \"Hipertension\": pat_hyper[who].astype(int),\n",
    "            \"Diabetes\": pat_diab[who].astype(int),\n",
    "            \"Alcoholism\": pat_alc[who].astype(int),\n",
    "            \"Handcap\": pat_hand[who],\n",
    "            \"SMS_received\": sms,\n",
    "            \"No-show\": np.where(no_show, \"Yes\", \"No\"),\n",
    "        }\n",
    "    )\n",
    "    df.loc[0, \"Age\"] = -1  # the real file's quirks, so the cleaning rules have something to do\n",
    "    late = df.index[1:3]\n",
    "    df.loc[late, \"ScheduledDay\"] = (\n",
    "        pd.to_datetime(df.loc[late, \"AppointmentDay\"]) + pd.Timedelta(days=1, hours=9)\n",
    "    ).dt.strftime(\"%Y-%m-%dT%H:%M:%SZ\")\n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "could not download the real data (ModuleNotFoundError)\nusing 110,527 rows of synthetic PRACTICE data: do not report these numbers\n"
     ]
    },
    {
     "output_type": "execute_result",
     "execution_count": 4,
     "data": {
      "text/plain": [
       "      PatientId  AppointmentID Gender          ScheduledDay        AppointmentDay  Age Neighbourhood  Scholarship  Hipertension  Diabetes  Alcoholism  Handcap  SMS_received No-show\n0  7.342097e+14        5099038      F  2016-06-04T10:21:39Z  2016-06-04T00:00:00Z   -1     BAIRRO 31            0             0         0           0        0             0     Yes\n1  2.302251e+14        5036345      F  2016-05-24T09:00:00Z  2016-05-23T00:00:00Z   27     BAIRRO 23            1             1         0           0        0             0     Yes\n2  5.406092e+14        5004843      F  2016-05-07T09:00:00Z  2016-05-06T00:00:00Z   88     BAIRRO 31            0             1         0           0        0             1      No\n3  6.937668e+13        5003094      M  2016-05-16T07:11:18Z  2016-05-16T00:00:00Z   21     BAIRRO 13            0             1         0           0        0             0      No\n4  2.957149e+14        5107907      M  2016-04-22T16:10:08Z  2016-05-17T00:00:00Z   33     BAIRRO 40            0             0         0           0        0             1      No"
      ],
      "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>PatientId</th>\n      <th>AppointmentID</th>\n      <th>Gender</th>\n      <th>ScheduledDay</th>\n      <th>AppointmentDay</th>\n      <th>Age</th>\n      <th>Neighbourhood</th>\n      <th>Scholarship</th>\n      <th>Hipertension</th>\n      <th>Diabetes</th>\n      <th>Alcoholism</th>\n      <th>Handcap</th>\n      <th>SMS_received</th>\n      <th>No-show</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>7.342097e+14</td>\n      <td>5099038</td>\n      <td>F</td>\n      <td>2016-06-04T10:21:39Z</td>\n      <td>2016-06-04T00:00:00Z</td>\n      <td>-1</td>\n      <td>BAIRRO 31</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>Yes</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>2.302251e+14</td>\n      <td>5036345</td>\n      <td>F</td>\n      <td>2016-05-24T09:00:00Z</td>\n      <td>2016-05-23T00:00:00Z</td>\n      <td>27</td>\n      <td>BAIRRO 23</td>\n      <td>1</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>Yes</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>5.406092e+14</td>\n      <td>5004843</td>\n      <td>F</td>\n      <td>2016-05-07T09:00:00Z</td>\n      <td>2016-05-06T00:00:00Z</td>\n      <td>88</td>\n      <td>BAIRRO 31</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>No</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>6.937668e+13</td>\n      <td>5003094</td>\n      <td>M</td>\n      <td>2016-05-16T07:11:18Z</td>\n      <td>2016-05-16T00:00:00Z</td>\n      <td>21</td>\n      <td>BAIRRO 13</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>No</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>2.957149e+14</td>\n      <td>5107907</td>\n      <td>M</td>\n      <td>2016-04-22T16:10:08Z</td>\n      <td>2016-05-17T00:00:00Z</td>\n      <td>33</td>\n      <td>BAIRRO 40</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>No</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "2fb5a6ed",
   "metadata": {},
   "source": [
    "CSV = \"noshowappointments-kagglev2-may-2016.csv\"\n",
    "found = next((p for p in (Path(CSV), Path(\"data\") / CSV, Path(\"../data/raw\") / CSV) if p.exists()), None)\n",
    "if found is None:\n",
    "    try:\n",
    "        import kagglehub\n",
    "\n",
    "        found = next(Path(kagglehub.dataset_download(\"joniarroba/noshowappointments\")).glob(\"*.csv\"))\n",
    "    except Exception as e:  # no internet, no Kaggle access, or kagglehub not installed\n",
    "        print(f\"could not download the real data ({type(e).__name__})\")\n",
    "if found is not None:\n",
    "    raw = pd.read_csv(found)\n",
    "    print(f\"loaded {len(raw):,} rows from {found}\")\n",
    "else:\n",
    "    raw = make_practice_data()\n",
    "    print(f\"using {len(raw):,} rows of synthetic PRACTICE data: do not report these numbers\")\n",
    "raw.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed0f750c",
   "metadata": {},
   "source": [
    "## First look\n",
    "Check the size, the types and the target before anything else. Note the misspelt column names (`Hipertension`,\n",
    "`Handcap`) in the original data: we will rename them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "(110527, 14)\n"
     ]
    },
    {
     "output_type": "execute_result",
     "execution_count": 5,
     "data": {
      "text/plain": [
       "PatientId         float64\nAppointmentID       int64\nGender                str\nScheduledDay          str\nAppointmentDay        str\nAge                 int64\nNeighbourhood         str\nScholarship         int64\nHipertension        int64\nDiabetes            int64\nAlcoholism          int64\nHandcap             int64\nSMS_received        int64\nNo-show               str\ndtype: object"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "e8ce0cae",
   "metadata": {},
   "source": [
    "print(raw.shape)\n",
    "raw.dtypes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 6,
     "data": {
      "text/plain": [
       "                  count          mean           std           min           25%           50%           75%           max\nPatientId      110527.0  4.991562e+14  2.881439e+14  1.131672e+11  2.496144e+14  5.001005e+14  7.483908e+14  9.999937e+14\nAppointmentID  110527.0  5.055263e+06  3.190654e+04  5.000000e+06  5.027632e+06  5.055263e+06  5.082894e+06  5.110526e+06\nAge            110527.0  3.718000e+01  2.411000e+01 -1.000000e+00  1.900000e+01  3.200000e+01  5.000000e+01  1.100000e+02\nScholarship    110527.0  1.000000e-01  3.000000e-01  0.000000e+00  0.000000e+00  0.000000e+00  0.000000e+00  1.000000e+00\nHipertension   110527.0  2.700000e-01  4.400000e-01  0.000000e+00  0.000000e+00  0.000000e+00  1.000000e+00  1.000000e+00\nDiabetes       110527.0  9.000000e-02  2.800000e-01  0.000000e+00  0.000000e+00  0.000000e+00  0.000000e+00  1.000000e+00\nAlcoholism     110527.0  3.000000e-02  1.700000e-01  0.000000e+00  0.000000e+00  0.000000e+00  0.000000e+00  1.000000e+00\nHandcap        110527.0  2.000000e-02  1.800000e-01  0.000000e+00  0.000000e+00  0.000000e+00  0.000000e+00  4.000000e+00\nSMS_received   110527.0  3.100000e-01  4.600000e-01  0.000000e+00  0.000000e+00  0.000000e+00  1.000000e+00  1.000000e+00"
      ],
      "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>count</th>\n      <th>mean</th>\n      <th>std</th>\n      <th>min</th>\n      <th>25%</th>\n      <th>50%</th>\n      <th>75%</th>\n      <th>max</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>PatientId</th>\n      <td>110527.0</td>\n      <td>4.991562e+14</td>\n      <td>2.881439e+14</td>\n      <td>1.131672e+11</td>\n      <td>2.496144e+14</td>\n      <td>5.001005e+14</td>\n      <td>7.483908e+14</td>\n      <td>9.999937e+14</td>\n    </tr>\n    <tr>\n      <th>AppointmentID</th>\n      <td>110527.0</td>\n      <td>5.055263e+06</td>\n      <td>3.190654e+04</td>\n      <td>5.000000e+06</td>\n      <td>5.027632e+06</td>\n      <td>5.055263e+06</td>\n      <td>5.082894e+06</td>\n      <td>5.110526e+06</td>\n    </tr>\n    <tr>\n      <th>Age</th>\n      <td>110527.0</td>\n      <td>3.718000e+01</td>\n      <td>2.411000e+01</td>\n      <td>-1.000000e+00</td>\n      <td>1.900000e+01</td>\n      <td>3.200000e+01</td>\n      <td>5.000000e+01</td>\n      <td>1.100000e+02</td>\n    </tr>\n    <tr>\n      <th>Scholarship</th>\n      <td>110527.0</td>\n      <td>1.000000e-01</td>\n      <td>3.000000e-01</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>1.000000e+00</td>\n    </tr>\n    <tr>\n      <th>Hipertension</th>\n      <td>110527.0</td>\n      <td>2.700000e-01</td>\n      <td>4.400000e-01</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>1.000000e+00</td>\n      <td>1.000000e+00</td>\n    </tr>\n    <tr>\n      <th>Diabetes</th>\n      <td>110527.0</td>\n      <td>9.000000e-02</td>\n      <td>2.800000e-01</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>1.000000e+00</td>\n    </tr>\n    <tr>\n      <th>Alcoholism</th>\n      <td>110527.0</td>\n      <td>3.000000e-02</td>\n      <td>1.700000e-01</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>1.000000e+00</td>\n    </tr>\n    <tr>\n      <th>Handcap</th>\n      <td>110527.0</td>\n      <td>2.000000e-02</td>\n      <td>1.800000e-01</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>4.000000e+00</td>\n    </tr>\n    <tr>\n      <th>SMS_received</th>\n      <td>110527.0</td>\n      <td>3.100000e-01</td>\n      <td>4.600000e-01</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>0.000000e+00</td>\n      <td>1.000000e+00</td>\n      <td>1.000000e+00</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "922a3e9a",
   "metadata": {},
   "source": [
    "raw.describe().T.round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 7,
     "data": {
      "text/plain": [
       "No-show\nNo     0.794394\nYes    0.205606\nName: share, dtype: float64"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "3f260e36",
   "metadata": {},
   "source": [
    "raw[\"No-show\"].value_counts(normalize=True).rename(\"share\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "647f38cd",
   "metadata": {},
   "source": [
    "About one appointment in five is missed. That imbalance matters: accuracy would be 80% for a model that always\n",
    "says \"will attend\", so we will judge models with **PR-AUC** (average precision) and probability quality instead."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0f22b974",
   "metadata": {},
   "source": [
    "## Check data quality\n",
    "Turn every assumption into a check. Hard problems (missing columns, unknown codes) would stop us; known quirks are\n",
    "counted so we can decide how to handle each one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 8,
     "data": {
      "text/plain": [
       "missing values                                   0\nduplicate AppointmentID                          0\nGender not F/M                                   0\nnegative Age                                     1\nAge above 115                                    0\nappointment before booking day                   2\nHandcap above 1 (it is a 0-4 level)            319\nAppointmentDay with a time other than 00:00      0\nName: rows, dtype: int64"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "e6467bb4",
   "metadata": {},
   "source": [
    "sched = pd.to_datetime(raw[\"ScheduledDay\"], utc=True)\n",
    "appt = pd.to_datetime(raw[\"AppointmentDay\"], utc=True)\n",
    "checks = pd.Series(\n",
    "    {\n",
    "        \"missing values\": int(raw.isna().sum().sum()),\n",
    "        \"duplicate AppointmentID\": int(raw[\"AppointmentID\"].duplicated().sum()),\n",
    "        \"Gender not F/M\": int((~raw[\"Gender\"].isin([\"F\", \"M\"])).sum()),\n",
    "        \"negative Age\": int((raw[\"Age\"] < 0).sum()),\n",
    "        \"Age above 115\": int((raw[\"Age\"] > 115).sum()),\n",
    "        \"appointment before booking day\": int((appt.dt.normalize() < sched.dt.normalize()).sum()),\n",
    "        \"Handcap above 1 (it is a 0-4 level)\": int((raw[\"Handcap\"] > 1).sum()),\n",
    "        \"AppointmentDay with a time other than 00:00\": int((appt.dt.hour != 0).sum()),\n",
    "    },\n",
    "    name=\"rows\",\n",
    ")\n",
    "checks"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "35ce625a",
   "metadata": {},
   "source": [
    "Findings: a few impossible ages, some appointments dated before they were booked (data-entry errors), `Handcap` is\n",
    "a level from 0 to 4 rather than yes/no, and `AppointmentDay` has no time of day. Nothing stops us."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5777bfa6",
   "metadata": {},
   "source": [
    "## Clean the data\n",
    "Rename and type the columns, then remove rows with named rules, **counting** what each rule removes so the\n",
    "effect of cleaning is visible."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "110,527 rows -> 110,524 rows\n"
     ]
    },
    {
     "output_type": "execute_result",
     "execution_count": 9,
     "data": {
      "text/plain": [
       "negative_age                  1\nage_above_max                 0\nappointment_before_booking    2\nduplicate_appointment_id      0\nName: rows removed, dtype: int64"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "be1e4e0d",
   "metadata": {},
   "source": [
    "RENAME = {\n",
    "    \"PatientId\": \"patient_id\",\n",
    "    \"AppointmentID\": \"appointment_id\",\n",
    "    \"Gender\": \"gender\",\n",
    "    \"ScheduledDay\": \"scheduled_at\",\n",
    "    \"AppointmentDay\": \"appointment_date\",\n",
    "    \"Age\": \"age\",\n",
    "    \"Neighbourhood\": \"neighbourhood\",\n",
    "    \"Scholarship\": \"scholarship\",\n",
    "    \"Hipertension\": \"hypertension\",\n",
    "    \"Diabetes\": \"diabetes\",\n",
    "    \"Alcoholism\": \"alcoholism\",\n",
    "    \"Handcap\": \"handicap_level\",\n",
    "    \"SMS_received\": \"sms_received\",\n",
    "    \"No-show\": \"no_show_raw\",\n",
    "}\n",
    "\n",
    "df = raw.rename(columns=RENAME)[list(RENAME.values())].copy()\n",
    "df[\"patient_id\"] = df[\"patient_id\"].astype(\"int64\").astype(\"string\")  # stored as float in the CSV\n",
    "df[\"appointment_id\"] = df[\"appointment_id\"].astype(\"int64\")\n",
    "df[\"scheduled_at\"] = pd.to_datetime(df[\"scheduled_at\"], utc=True).dt.tz_localize(None)\n",
    "df[\"appointment_date\"] = pd.to_datetime(df[\"appointment_date\"], utc=True).dt.tz_localize(None).dt.normalize()\n",
    "df[\"gender\"] = df[\"gender\"].astype(\"category\")\n",
    "df[\"neighbourhood\"] = df[\"neighbourhood\"].astype(\"string\").str.strip().str.upper()\n",
    "for col in (\"scholarship\", \"hypertension\", \"diabetes\", \"alcoholism\", \"sms_received\"):\n",
    "    df[col] = df[col].astype(\"int8\")\n",
    "df[\"handicap_level\"] = df[\"handicap_level\"].astype(\"int8\")\n",
    "df[\"no_show\"] = (df.pop(\"no_show_raw\") == \"Yes\").astype(\"int8\")  # target: 1 = did NOT attend\n",
    "\n",
    "rules = {\n",
    "    \"negative_age\": df[\"age\"] < 0,\n",
    "    \"age_above_max\": df[\"age\"] > 115,\n",
    "    \"appointment_before_booking\": df[\"appointment_date\"] < df[\"scheduled_at\"].dt.normalize(),\n",
    "    \"duplicate_appointment_id\": df[\"appointment_id\"].duplicated(keep=\"first\"),\n",
    "}\n",
    "keep = pd.Series(True, index=df.index)\n",
    "removed = {}\n",
    "for name, mask in rules.items():\n",
    "    removed[name] = int((mask & keep).sum())  # each row counted once, under the first rule that removes it\n",
    "    keep &= ~mask\n",
    "clean = (\n",
    "    df.loc[keep].sort_values([\"appointment_date\", \"scheduled_at\", \"appointment_id\"]).reset_index(drop=True)\n",
    ")\n",
    "print(f\"{len(df):,} rows -> {len(clean):,} rows\")\n",
    "pd.Series(removed, name=\"rows removed\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d7d9e3c",
   "metadata": {},
   "source": [
    "## Split by time, before exploring\n",
    "The model will always be trained on the past and used on the future, so we split the same way: the earliest 70%\n",
    "of appointment dates for training, the next 15% for validation (choosing and calibrating the model) and the last\n",
    "15% for a final test, used **once**. We split *before* exploring so that what we see in the test period cannot\n",
    "influence our choices."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 10,
     "data": {
      "text/plain": [
       "             rows  no_show_rate      first       last\ntrain       75179      0.204472 2016-04-29 2016-05-26\nvalidation  16229      0.206174 2016-05-27 2016-06-01\ntest        19116      0.209510 2016-06-02 2016-06-08"
      ],
      "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>rows</th>\n      <th>no_show_rate</th>\n      <th>first</th>\n      <th>last</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>train</th>\n      <td>75179</td>\n      <td>0.204472</td>\n      <td>2016-04-29</td>\n      <td>2016-05-26</td>\n    </tr>\n    <tr>\n      <th>validation</th>\n      <td>16229</td>\n      <td>0.206174</td>\n      <td>2016-05-27</td>\n      <td>2016-06-01</td>\n    </tr>\n    <tr>\n      <th>test</th>\n      <td>19116</td>\n      <td>0.209510</td>\n      <td>2016-06-02</td>\n      <td>2016-06-08</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "8614e2e1",
   "metadata": {},
   "source": [
    "counts = clean.groupby(\"appointment_date\").size().sort_index()\n",
    "share = counts.cumsum() / counts.sum()\n",
    "validation_start = share.index[(share > 0.70).to_numpy().argmax()]\n",
    "test_start = share.index[(share > 0.85).to_numpy().argmax()]\n",
    "period = np.where(\n",
    "    clean[\"appointment_date\"] < validation_start,\n",
    "    \"train\",\n",
    "    np.where(clean[\"appointment_date\"] < test_start, \"validation\", \"test\"),\n",
    ")\n",
    "clean.groupby(period, sort=False).agg(\n",
    "    rows=(\"no_show\", \"size\"),\n",
    "    no_show_rate=(\"no_show\", \"mean\"),\n",
    "    first=(\"appointment_date\", \"min\"),\n",
    "    last=(\"appointment_date\", \"max\"),\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa7f13ab",
   "metadata": {},
   "source": [
    "## Explore the training period\n",
    "Questions first, then charts. Does waiting longer for an appointment raise the risk? Who misses more? Do SMS\n",
    "reminders help? Everything here uses the training period only."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "10100e8a",
   "metadata": {},
   "source": [
    "train_raw = clean[period == \"train\"].copy()\n",
    "train_raw[\"lead_days\"] = (train_raw[\"appointment_date\"] - train_raw[\"scheduled_at\"].dt.normalize()).dt.days\n",
    "overall = train_raw[\"no_show\"].mean()\n",
    "\n",
    "\n",
    "def rate_by(col, title, ax):\n",
    "    t = train_raw.groupby(col, observed=True)[\"no_show\"].agg([\"mean\", \"size\"])\n",
    "    ax.bar(t.index.astype(str), t[\"mean\"], color=\"#2E6FF2\")\n",
    "    ax.axhline(overall, color=\"#E07B00\", ls=\"--\", lw=1)\n",
    "    ax.set(title=title, ylabel=\"no-show rate\")\n",
    "    ax.tick_params(axis=\"x\", rotation=45, labelsize=8)\n",
    "    return t\n",
    "\n",
    "\n",
    "fig, axes = plt.subplots(2, 2, figsize=(11, 7))\n",
    "lead_bins = pd.cut(\n",
    "    train_raw[\"lead_days\"],\n",
    "    [-1, 0, 2, 7, 14, 30, 60, 400],\n",
    "    labels=[\"same day\", \"1-2\", \"3-7\", \"8-14\", \"15-30\", \"31-60\", \"60+\"],\n",
    ")\n",
    "rate_by(lead_bins, \"By days between booking and appointment\", axes[0, 0])\n",
    "age_band = pd.cut(\n",
    "    train_raw[\"age\"],\n",
    "    [-1, 12, 17, 34, 49, 64, 200],\n",
    "    labels=[\"0-12\", \"13-17\", \"18-34\", \"35-49\", \"50-64\", \"65+\"],\n",
    ")\n",
    "rate_by(age_band, \"By age band\", axes[0, 1])\n",
    "rate_by(\"sms_received\", \"By SMS reminder received\", axes[1, 0])\n",
    "weekday = pd.Categorical(\n",
    "    train_raw[\"appointment_date\"].dt.day_name().str[:3], categories=[\"Mon\", \"Tue\", \"Wed\", \"Thu\", \"Fri\", \"Sat\"]\n",
    ")\n",
    "rate_by(weekday, \"By weekday\", axes[1, 1])\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "19f04c08",
   "metadata": {},
   "source": [
    "**What to look for.** The saved charts come from synthetic data; the real data shows these patterns, some more strongly:\n",
    "* **Lead time is the strongest pattern:** risk climbs the longer patients wait. In the real data, same-day\n",
    "  appointments are almost always kept.\n",
    "* **Age:** children, teenagers and young adults miss more often than older patients.\n",
    "* **The SMS paradox (real data):** patients who received an SMS miss *more* (about 28% against 17%). SMS were only\n",
    "  sent for appointments booked days ahead, so SMS is tangled up with lead time. This is confounding, not proof that\n",
    "  SMS backfires, and a reason never to read a model's weights as causes. (The synthetic generator does not\n",
    "  reproduce it, so the bar chart above is flat.)\n",
    "\n",
    "Does past behaviour predict future behaviour? We need each patient's history, but only the part known at booking."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "49% of patients have more than one appointment in the training period\n"
     ]
    },
    {
     "output_type": "execute_result",
     "execution_count": 12,
     "data": {
      "text/plain": [
       "0    0.203976\n1    0.205288\n2    0.205501\n3    0.203218\nName: no-show rate by visit number (3 = 3rd+), dtype: float64"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "da914c5b",
   "metadata": {},
   "source": [
    "visits = train_raw.groupby(\"patient_id\").size()\n",
    "print(f\"{(visits > 1).mean():.0%} of patients have more than one appointment in the training period\")\n",
    "train_raw.groupby(train_raw.groupby(\"patient_id\").cumcount().clip(upper=3))[\"no_show\"].mean().rename(\n",
    "    \"no-show rate by visit number (3 = 3rd+)\"\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3cb19cf5",
   "metadata": {},
   "source": [
    "## Engineer features without leaking the future\n",
    "Two kinds of features:\n",
    "* **Row features** use only the appointment itself (lead time, age, weekday, conditions...).\n",
    "* **History features** count the patient's earlier appointments and no-shows, using **only appointments whose\n",
    "  outcome was known when this one was booked** (at least one day before the booking date).\n",
    "\n",
    "Counting all of a patient's no-shows, including later ones, is a classic leak: it scores brilliantly offline and\n",
    "fails in production."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 13,
     "data": {
      "text/plain": [
       "   lead_days  age  scheduled_hour  n_conditions  handicap_level  prior_appointments  prior_no_shows  prior_no_show_rate  days_since_last_appointment  same_day  gender_f  scholarship  hypertension  diabetes  alcoholism  sms_received appointment_weekday neighbourhood\n0        135   54              15             0               0                   0               0                 NaN                          NaN         0         1            0             0         0           0             1                 Fri     BAIRRO 29\n1        134  110              14             1               0                   0               0                 NaN                          NaN         0         1            0             1         0           0             0                 Fri     BAIRRO 24\n2         88   39               7             1               0                   0               0                 NaN                          NaN         0         1            0             0         1           0             0                 Fri     BAIRRO 19\n3         80   45              18             2               1                   0               0                 NaN                          NaN         0         0            0             1         0           0             0                 Fri     BAIRRO 03\n4         74    8              16             0               0                   0               0                 NaN                          NaN         0         1            0             0         0           0             1                 Fri     BAIRRO 36"
      ],
      "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>lead_days</th>\n      <th>age</th>\n      <th>scheduled_hour</th>\n      <th>n_conditions</th>\n      <th>handicap_level</th>\n      <th>prior_appointments</th>\n      <th>prior_no_shows</th>\n      <th>prior_no_show_rate</th>\n      <th>days_since_last_appointment</th>\n      <th>same_day</th>\n      <th>gender_f</th>\n      <th>scholarship</th>\n      <th>hypertension</th>\n      <th>diabetes</th>\n      <th>alcoholism</th>\n      <th>sms_received</th>\n      <th>appointment_weekday</th>\n      <th>neighbourhood</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>135</td>\n      <td>54</td>\n      <td>15</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>Fri</td>\n      <td>BAIRRO 29</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>134</td>\n      <td>110</td>\n      <td>14</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>Fri</td>\n      <td>BAIRRO 24</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>88</td>\n      <td>39</td>\n      <td>7</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>Fri</td>\n      <td>BAIRRO 19</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>80</td>\n      <td>45</td>\n      <td>18</td>\n      <td>2</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>Fri</td>\n      <td>BAIRRO 03</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>74</td>\n      <td>8</td>\n      <td>16</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>NaN</td>\n      <td>NaN</td>\n      <td>0</td>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>1</td>\n      <td>Fri</td>\n      <td>BAIRRO 36</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "7daa1ee3",
   "metadata": {},
   "source": [
    "WEEKDAYS = [\"Mon\", \"Tue\", \"Wed\", \"Thu\", \"Fri\", \"Sat\", \"Sun\"]\n",
    "\n",
    "\n",
    "def row_features(d):\n",
    "    out = pd.DataFrame(index=d.index)\n",
    "    booked_day = d[\"scheduled_at\"].dt.normalize()\n",
    "    out[\"lead_days\"] = (d[\"appointment_date\"] - booked_day).dt.days.astype(\"int32\")\n",
    "    out[\"same_day\"] = (out[\"lead_days\"] == 0).astype(\"int8\")\n",
    "    out[\"scheduled_hour\"] = d[\"scheduled_at\"].dt.hour.astype(\"int8\")\n",
    "    out[\"appointment_weekday\"] = pd.Categorical(\n",
    "        d[\"appointment_date\"].dt.dayofweek.map(dict(enumerate(WEEKDAYS))), categories=WEEKDAYS\n",
    "    )\n",
    "    out[\"age\"] = d[\"age\"].astype(\"int16\")\n",
    "    out[\"age_band\"] = pd.cut(\n",
    "        d[\"age\"],\n",
    "        bins=[-1, 12, 17, 34, 49, 64, 200],\n",
    "        labels=[\"0-12\", \"13-17\", \"18-34\", \"35-49\", \"50-64\", \"65+\"],\n",
    "    )\n",
    "    out[\"gender_f\"] = (d[\"gender\"].astype(\"string\") == \"F\").astype(\"int8\")\n",
    "    out[\"handicap_level\"] = d[\"handicap_level\"].astype(\"int8\")\n",
    "    for col in (\"scholarship\", \"hypertension\", \"diabetes\", \"alcoholism\", \"sms_received\"):\n",
    "        out[col] = d[col].astype(\"int8\")\n",
    "    out[\"n_conditions\"] = (\n",
    "        d[\"hypertension\"] + d[\"diabetes\"] + d[\"alcoholism\"] + (d[\"handicap_level\"] > 0)\n",
    "    ).astype(\"int8\")\n",
    "    out[\"neighbourhood\"] = d[\"neighbourhood\"].astype(\"string\")\n",
    "    return out\n",
    "\n",
    "\n",
    "def history_features(d, min_days_gap=1):\n",
    "    \"\"\"Earlier appointments and no-shows per patient, known at booking time (merge_asof: strictly before).\"\"\"\n",
    "    events = (\n",
    "        d.groupby([\"patient_id\", \"appointment_date\"], observed=True)[\"no_show\"]\n",
    "        .agg(n=\"size\", ns=\"sum\")\n",
    "        .reset_index()\n",
    "        .sort_values(\"appointment_date\")\n",
    "    )\n",
    "    events[\"cum_n\"] = events.groupby(\"patient_id\")[\"n\"].cumsum()\n",
    "    events[\"cum_ns\"] = events.groupby(\"patient_id\")[\"ns\"].cumsum()\n",
    "    q = pd.DataFrame(\n",
    "        {\n",
    "            \"row\": d.index,\n",
    "            \"patient_id\": d[\"patient_id\"],\n",
    "            \"cutoff\": d[\"scheduled_at\"].dt.normalize() - pd.Timedelta(days=min_days_gap - 1),\n",
    "        }\n",
    "    )\n",
    "    m = pd.merge_asof(\n",
    "        q.sort_values(\"cutoff\"),\n",
    "        events[[\"patient_id\", \"appointment_date\", \"cum_n\", \"cum_ns\"]],\n",
    "        left_on=\"cutoff\",\n",
    "        right_on=\"appointment_date\",\n",
    "        by=\"patient_id\",\n",
    "        allow_exact_matches=False,\n",
    "        direction=\"backward\",\n",
    "    )\n",
    "    m = m.set_index(\"row\").reindex(d.index)\n",
    "    out = pd.DataFrame(index=d.index)\n",
    "    out[\"prior_appointments\"] = m[\"cum_n\"].fillna(0).astype(\"int32\")\n",
    "    out[\"prior_no_shows\"] = m[\"cum_ns\"].fillna(0).astype(\"int32\")\n",
    "    with np.errstate(divide=\"ignore\", invalid=\"ignore\"):\n",
    "        out[\"prior_no_show_rate\"] = (\n",
    "            out[\"prior_no_shows\"] / out[\"prior_appointments\"].replace(0, np.nan)\n",
    "        ).astype(\"float64\")\n",
    "    out[\"days_since_last_appointment\"] = (m[\"cutoff\"] - m[\"appointment_date\"]).dt.days.astype(\"float64\")\n",
    "    return out\n",
    "\n",
    "\n",
    "NUMERIC = [\n",
    "    \"lead_days\",\n",
    "    \"age\",\n",
    "    \"scheduled_hour\",\n",
    "    \"n_conditions\",\n",
    "    \"handicap_level\",\n",
    "    \"prior_appointments\",\n",
    "    \"prior_no_shows\",\n",
    "    \"prior_no_show_rate\",\n",
    "    \"days_since_last_appointment\",\n",
    "]\n",
    "BINARY = [\"same_day\", \"gender_f\", \"scholarship\", \"hypertension\", \"diabetes\", \"alcoholism\", \"sms_received\"]\n",
    "CATEGORICAL = [\"appointment_weekday\", \"neighbourhood\"]\n",
    "FEATURES = NUMERIC + BINARY + CATEGORICAL\n",
    "META = [\"appointment_id\", \"patient_id\", \"appointment_date\", \"scheduled_at\", \"gender\", \"age_band\"]\n",
    "\n",
    "feats = pd.concat(\n",
    "    [\n",
    "        clean[[\"appointment_id\", \"patient_id\", \"appointment_date\", \"scheduled_at\", \"gender\"]],\n",
    "        row_features(clean),\n",
    "        history_features(clean),\n",
    "        clean[[\"no_show\"]],\n",
    "    ],\n",
    "    axis=1,\n",
    ")[META + FEATURES + [\"no_show\"]]\n",
    "feats[FEATURES].head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e320c5e9",
   "metadata": {},
   "source": [
    "**Prove there is no leakage.** Pick a patient with several visits and check by hand that each history count only\n",
    "includes appointments that ended at least a day before the booking."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 14,
     "data": {
      "text/plain": [
       "             scheduled_at appointment_date  no_show  prior_appointments  prior_no_shows\n3753  2016-04-21 12:38:07       2016-04-30        0                   0               0\n4733  2016-04-30 09:14:30       2016-04-30        0                   0               0\n8761  2016-05-01 09:47:53       2016-05-02        0                   2               0\n44519 2016-05-04 10:00:36       2016-05-16        0                   3               0\n63344 2016-05-12 16:00:36       2016-05-23        0                   3               0"
      ],
      "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>scheduled_at</th>\n      <th>appointment_date</th>\n      <th>no_show</th>\n      <th>prior_appointments</th>\n      <th>prior_no_shows</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>3753</th>\n      <td>2016-04-21 12:38:07</td>\n      <td>2016-04-30</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>4733</th>\n      <td>2016-04-30 09:14:30</td>\n      <td>2016-04-30</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>8761</th>\n      <td>2016-05-01 09:47:53</td>\n      <td>2016-05-02</td>\n      <td>0</td>\n      <td>2</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>44519</th>\n      <td>2016-05-04 10:00:36</td>\n      <td>2016-05-16</td>\n      <td>0</td>\n      <td>3</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>63344</th>\n      <td>2016-05-12 16:00:36</td>\n      <td>2016-05-23</td>\n      <td>0</td>\n      <td>3</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "eb22787a",
   "metadata": {},
   "source": [
    "pid = feats.loc[feats[\"prior_appointments\"] >= 2, \"patient_id\"].iloc[0]\n",
    "one = feats[feats[\"patient_id\"] == pid].sort_values(\"scheduled_at\")\n",
    "for _, r in one.iterrows():\n",
    "    earlier = one[one[\"appointment_date\"] < r[\"scheduled_at\"].normalize()]\n",
    "    assert r[\"prior_appointments\"] == len(earlier) and r[\"prior_no_shows\"] == earlier[\"no_show\"].sum()\n",
    "one[[\"scheduled_at\", \"appointment_date\", \"no_show\", \"prior_appointments\", \"prior_no_shows\"]]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "745b6906",
   "metadata": {},
   "source": [
    "Now split the feature table with the same dates as before."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "train 75,179 | validation 16,229 | test 19,116 (test untouched until the end)\n"
     ]
    }
   ],
   "id": "cad4def7",
   "metadata": {},
   "source": [
    "d = feats[\"appointment_date\"]\n",
    "train, val, test = (\n",
    "    feats[d < validation_start],\n",
    "    feats[(d >= validation_start) & (d < test_start)],\n",
    "    feats[d >= test_start],\n",
    ")\n",
    "X_train, y_train = train[FEATURES], train[\"no_show\"]\n",
    "X_val, y_val = val[FEATURES], val[\"no_show\"]\n",
    "print(f\"train {len(train):,} | validation {len(val):,} | test {len(test):,} (test untouched until the end)\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bbb3d42a",
   "metadata": {},
   "source": [
    "## Start with a baseline\n",
    "The simplest possible model predicts the training no-show rate for everyone. Its PR-AUC equals the share of\n",
    "no-shows. Any real model must clearly beat it, or it is not worth the complexity.\n",
    "\n",
    "Cross-validation folds are **grouped by patient**: the same patient never appears in both the training and the\n",
    "scoring part of a fold, so a model cannot score well by memorising individuals."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 16,
     "data": {
      "text/plain": [
       "          ROC-AUC  PR-AUC   Brier\nbaseline      0.5  0.2062  0.1637"
      ],
      "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>ROC-AUC</th>\n      <th>PR-AUC</th>\n      <th>Brier</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>baseline</th>\n      <td>0.5</td>\n      <td>0.2062</td>\n      <td>0.1637</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "671c9e6c",
   "metadata": {},
   "source": [
    "from sklearn.dummy import DummyClassifier\n",
    "from sklearn.metrics import average_precision_score, brier_score_loss, roc_auc_score\n",
    "from sklearn.model_selection import GroupKFold\n",
    "\n",
    "folds = list(GroupKFold(n_splits=5).split(X_train, y_train, groups=train[\"patient_id\"]))\n",
    "\n",
    "\n",
    "def scores(y, p):\n",
    "    return {\n",
    "        \"ROC-AUC\": roc_auc_score(y, p),\n",
    "        \"PR-AUC\": average_precision_score(y, p),\n",
    "        \"Brier\": brier_score_loss(y, p),\n",
    "    }\n",
    "\n",
    "\n",
    "dummy = DummyClassifier(strategy=\"prior\").fit(X_train, y_train)\n",
    "results = {\"baseline\": scores(y_val, dummy.predict_proba(X_val)[:, 1])}\n",
    "pd.DataFrame(results).T.round(4)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "468237f4",
   "metadata": {},
   "source": [
    "## Model 1: logistic regression\n",
    "A linear model is fast, explainable and often hard to beat. Preprocessing goes **inside the pipeline** so it is\n",
    "learned from training folds only and saved with the model:\n",
    "* numbers: fill missing history with the median, add a \"was missing\" flag, standardise;\n",
    "* weekday: one-hot; neighbourhood: one-hot with rare ones grouped as \"infrequent\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "best: {'clf__C': 0.005, 'clf__class_weight': 'balanced'} CV PR-AUC 0.2821\n"
     ]
    }
   ],
   "id": "b64fdfab",
   "metadata": {},
   "source": [
    "from scipy.stats import loguniform, randint, uniform\n",
    "from sklearn.compose import ColumnTransformer\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from sklearn.pipeline import Pipeline, make_pipeline\n",
    "from sklearn.preprocessing import OneHotEncoder, StandardScaler\n",
    "\n",
    "logistic = Pipeline(\n",
    "    [\n",
    "        (\n",
    "            \"pre\",\n",
    "            ColumnTransformer(\n",
    "                [\n",
    "                    (\n",
    "                        \"num\",\n",
    "                        make_pipeline(SimpleImputer(strategy=\"median\", add_indicator=True), StandardScaler()),\n",
    "                        NUMERIC,\n",
    "                    ),\n",
    "                    (\"bin\", \"passthrough\", BINARY),\n",
    "                    (\n",
    "                        \"wday\",\n",
    "                        OneHotEncoder(categories=[WEEKDAYS], drop=\"first\", handle_unknown=\"ignore\"),\n",
    "                        [\"appointment_weekday\"],\n",
    "                    ),\n",
    "                    (\n",
    "                        \"hood\",\n",
    "                        OneHotEncoder(min_frequency=50, handle_unknown=\"infrequent_if_exist\"),\n",
    "                        [\"neighbourhood\"],\n",
    "                    ),\n",
    "                ],\n",
    "                verbose_feature_names_out=True,\n",
    "            ),\n",
    "        ),\n",
    "        (\"clf\", LogisticRegression(max_iter=3000, random_state=SEED)),\n",
    "    ]\n",
    ")\n",
    "search_lr = RandomizedSearchCV(\n",
    "    logistic,\n",
    "    {\"clf__C\": loguniform(1e-3, 1e2), \"clf__class_weight\": [None, \"balanced\"]},\n",
    "    n_iter=20,\n",
    "    scoring=\"average_precision\",\n",
    "    cv=folds,\n",
    "    random_state=SEED,\n",
    "    n_jobs=-1,\n",
    ")\n",
    "search_lr.fit(X_train, y_train)\n",
    "print(\n",
    "    \"best:\",\n",
    "    {k: (round(float(v), 4) if isinstance(v, float) else v) for k, v in search_lr.best_params_.items()},\n",
    "    f\"CV PR-AUC {search_lr.best_score_:.4f}\",\n",
    ")\n",
    "results[\"logistic\"] = scores(y_val, search_lr.predict_proba(X_val)[:, 1])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fe534a46",
   "metadata": {},
   "source": [
    "## Model 2: gradient boosting\n",
    "Trees capture interactions (lead time matters more for young adults) and handle missing history natively.\n",
    "Two details worth copying:\n",
    "* **Target encoding** for neighbourhood, cross-fitted so a row's code never uses its own label.\n",
    "* A **monotonic constraint**: more past no-shows must never *lower* the predicted risk. Domain knowledge as a rule\n",
    "  makes the model more trustworthy and harder to fool with noise."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "best: {'clf__l2_regularization': 0.0001, 'clf__learning_rate': 0.0629, 'clf__max_features': 0.6974, 'clf__max_leaf_nodes': 15, 'clf__min_samples_leaf': 82} CV PR-AUC 0.2799\n"
     ]
    }
   ],
   "id": "bcb2c231",
   "metadata": {},
   "source": [
    "from sklearn.ensemble import HistGradientBoostingClassifier\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.preprocessing import OrdinalEncoder, TargetEncoder\n",
    "\n",
    "hgb = Pipeline(\n",
    "    [\n",
    "        (\n",
    "            \"pre\",\n",
    "            ColumnTransformer(\n",
    "                [\n",
    "                    (\"num\", \"passthrough\", NUMERIC),\n",
    "                    (\"bin\", \"passthrough\", BINARY),\n",
    "                    (\n",
    "                        \"wday\",\n",
    "                        OrdinalEncoder(\n",
    "                            categories=[WEEKDAYS], handle_unknown=\"use_encoded_value\", unknown_value=-1\n",
    "                        ),\n",
    "                        [\"appointment_weekday\"],\n",
    "                    ),\n",
    "                    (\n",
    "                        \"hood\",\n",
    "                        TargetEncoder(target_type=\"binary\", cv=KFold(5, shuffle=True, random_state=SEED)),\n",
    "                        [\"neighbourhood\"],\n",
    "                    ),\n",
    "                ],\n",
    "                verbose_feature_names_out=True,\n",
    "            ).set_output(transform=\"pandas\"),\n",
    "        ),\n",
    "        (\n",
    "            \"clf\",\n",
    "            HistGradientBoostingClassifier(\n",
    "                random_state=SEED,\n",
    "                early_stopping=True,\n",
    "                validation_fraction=0.1,\n",
    "                n_iter_no_change=20,\n",
    "                monotonic_cst={\"num__prior_no_shows\": 1, \"num__prior_no_show_rate\": 1},\n",
    "                categorical_features=[\"wday__appointment_weekday\"],\n",
    "            ),\n",
    "        ),\n",
    "    ]\n",
    ")\n",
    "space = {\n",
    "    \"clf__learning_rate\": loguniform(0.02, 0.3),\n",
    "    \"clf__max_leaf_nodes\": randint(8, 64),\n",
    "    \"clf__min_samples_leaf\": randint(20, 300),\n",
    "    \"clf__l2_regularization\": loguniform(1e-4, 10),\n",
    "    \"clf__max_features\": uniform(0.5, 0.5),\n",
    "}\n",
    "search_hgb = RandomizedSearchCV(\n",
    "    hgb, space, n_iter=20, scoring=\"average_precision\", cv=folds, random_state=SEED, n_jobs=-1\n",
    ")\n",
    "search_hgb.fit(X_train, y_train)\n",
    "print(\n",
    "    \"best:\",\n",
    "    {k: (round(float(v), 4) if isinstance(v, float) else v) for k, v in search_hgb.best_params_.items()},\n",
    "    f\"CV PR-AUC {search_hgb.best_score_:.4f}\",\n",
    ")\n",
    "results[\"gradient boosting\"] = scores(y_val, search_hgb.predict_proba(X_val)[:, 1])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "73cb1677",
   "metadata": {},
   "source": [
    "## Compare and choose on validation\n",
    "The validation period decides. The test period is still untouched."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "chosen: logistic\n"
     ]
    },
    {
     "output_type": "execute_result",
     "execution_count": 19,
     "data": {
      "text/plain": [
       "                   ROC-AUC  PR-AUC   Brier  PR-AUC gain over baseline\nbaseline            0.5000  0.2062  0.1637                     0.0000\nlogistic            0.6239  0.2950  0.2386                     0.0888\ngradient boosting   0.6183  0.2877  0.1589                     0.0815"
      ],
      "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>ROC-AUC</th>\n      <th>PR-AUC</th>\n      <th>Brier</th>\n      <th>PR-AUC gain over baseline</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>baseline</th>\n      <td>0.5000</td>\n      <td>0.2062</td>\n      <td>0.1637</td>\n      <td>0.0000</td>\n    </tr>\n    <tr>\n      <th>logistic</th>\n      <td>0.6239</td>\n      <td>0.2950</td>\n      <td>0.2386</td>\n      <td>0.0888</td>\n    </tr>\n    <tr>\n      <th>gradient boosting</th>\n      <td>0.6183</td>\n      <td>0.2877</td>\n      <td>0.1589</td>\n      <td>0.0815</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "9ebb23e2",
   "metadata": {},
   "source": [
    "comparison = pd.DataFrame(results).T\n",
    "comparison[\"PR-AUC gain over baseline\"] = comparison[\"PR-AUC\"] - comparison.loc[\"baseline\", \"PR-AUC\"]\n",
    "best_name = comparison.drop(index=\"baseline\")[\"PR-AUC\"].idxmax()\n",
    "best = {\"logistic\": search_lr, \"gradient boosting\": search_hgb}[best_name].best_estimator_\n",
    "assert comparison.loc[best_name, \"PR-AUC gain over baseline\"] >= 0.02, (\n",
    "    \"not clearly better than the baseline: stop here\"\n",
    ")\n",
    "print(\"chosen:\", best_name)\n",
    "comparison.round(4)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "49dc26dd",
   "metadata": {},
   "source": [
    "If the two models are close, the simpler one is often the better choice in practice. Here we follow the rule we\n",
    "set: the higher validation PR-AUC wins, and it must beat the baseline by at least 0.02."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d5ae2330",
   "metadata": {},
   "source": [
    "## Calibrate the probabilities\n",
    "We compare probabilities with a 27.8% threshold, so a predicted 30% must really mean about 30%. A reliability\n",
    "chart shows it: points on the diagonal are well calibrated. **Isotonic regression**, fitted on the validation\n",
    "period while the model stays frozen, corrects the mapping."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "2b151e57",
   "metadata": {},
   "source": [
    "from sklearn.calibration import CalibratedClassifierCV, calibration_curve\n",
    "from sklearn.frozen import FrozenEstimator\n",
    "\n",
    "calibrated = CalibratedClassifierCV(FrozenEstimator(best), method=\"isotonic\").fit(X_val, y_val)\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(5, 4.5))\n",
    "for model, label in ((best, \"before\"), (calibrated, \"after (isotonic)\")):\n",
    "    frac, mean_p = calibration_curve(y_val, model.predict_proba(X_val)[:, 1], n_bins=10, strategy=\"quantile\")\n",
    "    ax.plot(mean_p, frac, marker=\"o\", label=label)\n",
    "ax.plot([0, 1], [0, 1], ls=\"--\", color=\"grey\", lw=1)\n",
    "ax.set(xlabel=\"predicted probability\", ylabel=\"observed no-show rate\", title=\"Reliability (validation)\")\n",
    "ax.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "86a35455",
   "metadata": {},
   "source": [
    "Calibrating on the same data we then plot is optimistic; the honest check comes on the test period below."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c9ed086e",
   "metadata": {},
   "source": [
    "## Choose the threshold from costs\n",
    "For each possible threshold: how many patients would we call, how many no-shows would we reach, and what is the\n",
    "**net benefit** per 100 appointments? The break-even threshold from the business case should sit near the top of\n",
    "the curve."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure>"
      ]
     },
     "metadata": {}
    },
    {
     "output_type": "execute_result",
     "execution_count": 21,
     "data": {
      "text/plain": [
       "    threshold  call share  no-shows reached  net benefit per 100\n0        0.05       0.991             0.999             -124.955\n5        0.10       0.953             0.984             -111.325\n10       0.15       0.743             0.852              -55.487\n15       0.20       0.530             0.669              -16.791\n20       0.25       0.220             0.343               17.432\n25       0.30       0.143             0.236               16.107\n30       0.35       0.042             0.082                9.594\n35       0.40       0.017             0.036                4.806\n40       0.45       0.002             0.004                0.832\n45       0.50       0.002             0.004                0.832\n50       0.55       0.002             0.004                0.832\n55       0.60       0.000             0.000                0.000\n60       0.65       0.000             0.000                0.000\n65       0.70       0.000             0.000                0.000\n70       0.75       0.000             0.000                0.000\n75       0.80       0.000             0.000                0.000"
      ],
      "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>threshold</th>\n      <th>call share</th>\n      <th>no-shows reached</th>\n      <th>net benefit per 100</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0.05</td>\n      <td>0.991</td>\n      <td>0.999</td>\n      <td>-124.955</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>0.10</td>\n      <td>0.953</td>\n      <td>0.984</td>\n      <td>-111.325</td>\n    </tr>\n    <tr>\n      <th>10</th>\n      <td>0.15</td>\n      <td>0.743</td>\n      <td>0.852</td>\n      <td>-55.487</td>\n    </tr>\n    <tr>\n      <th>15</th>\n      <td>0.20</td>\n      <td>0.530</td>\n      <td>0.669</td>\n      <td>-16.791</td>\n    </tr>\n    <tr>\n      <th>20</th>\n      <td>0.25</td>\n      <td>0.220</td>\n      <td>0.343</td>\n      <td>17.432</td>\n    </tr>\n    <tr>\n      <th>25</th>\n      <td>0.30</td>\n      <td>0.143</td>\n      <td>0.236</td>\n      <td>16.107</td>\n    </tr>\n    <tr>\n      <th>30</th>\n      <td>0.35</td>\n      <td>0.042</td>\n      <td>0.082</td>\n      <td>9.594</td>\n    </tr>\n    <tr>\n      <th>35</th>\n      <td>0.40</td>\n      <td>0.017</td>\n      <td>0.036</td>\n      <td>4.806</td>\n    </tr>\n    <tr>\n      <th>40</th>\n      <td>0.45</td>\n      <td>0.002</td>\n      <td>0.004</td>\n      <td>0.832</td>\n    </tr>\n    <tr>\n      <th>45</th>\n      <td>0.50</td>\n      <td>0.002</td>\n      <td>0.004</td>\n      <td>0.832</td>\n    </tr>\n    <tr>\n      <th>50</th>\n      <td>0.55</td>\n      <td>0.002</td>\n      <td>0.004</td>\n      <td>0.832</td>\n    </tr>\n    <tr>\n      <th>55</th>\n      <td>0.60</td>\n      <td>0.000</td>\n      <td>0.000</td>\n      <td>0.000</td>\n    </tr>\n    <tr>\n      <th>60</th>\n      <td>0.65</td>\n      <td>0.000</td>\n      <td>0.000</td>\n      <td>0.000</td>\n    </tr>\n    <tr>\n      <th>65</th>\n      <td>0.70</td>\n      <td>0.000</td>\n      <td>0.000</td>\n      <td>0.000</td>\n    </tr>\n    <tr>\n      <th>70</th>\n      <td>0.75</td>\n      <td>0.000</td>\n      <td>0.000</td>\n      <td>0.000</td>\n    </tr>\n    <tr>\n      <th>75</th>\n      <td>0.80</td>\n      <td>0.000</td>\n      <td>0.000</td>\n      <td>0.000</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "05e465b7",
   "metadata": {},
   "source": [
    "def net_benefit_per_100(y, p, t):\n",
    "    call = p >= t\n",
    "    gain = y[call].sum() * CALL_PREVENTS * COST_OF_NO_SHOW - call.sum() * COST_OF_CALL\n",
    "    return 100 * gain / len(y)\n",
    "\n",
    "\n",
    "p_val = calibrated.predict_proba(X_val)[:, 1]\n",
    "yv = y_val.to_numpy()\n",
    "ts = np.round(np.arange(0.05, 0.81, 0.01), 2)\n",
    "curve = pd.DataFrame(\n",
    "    {\n",
    "        \"threshold\": ts,\n",
    "        \"call share\": [(p_val >= t).mean() for t in ts],\n",
    "        \"no-shows reached\": [(p_val[yv == 1] >= t).mean() for t in ts],\n",
    "        \"net benefit per 100\": [net_benefit_per_100(yv, p_val, t) for t in ts],\n",
    "    }\n",
    ")\n",
    "fig, ax = plt.subplots(figsize=(7, 3.6))\n",
    "ax.plot(curve[\"threshold\"], curve[\"net benefit per 100\"], color=\"#15803D\")\n",
    "ax.axvline(break_even, ls=\"--\", color=\"#E07B00\", label=f\"break-even {break_even:.3f}\")\n",
    "ax.set(xlabel=\"call if probability ≥\", ylabel=\"AUD per 100 appointments\", title=\"Net benefit on validation\")\n",
    "ax.legend()\n",
    "plt.show()\n",
    "threshold = break_even\n",
    "curve.iloc[::5].round(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "74621a5f",
   "metadata": {},
   "source": [
    "## Test once\n",
    "Everything is now fixed: features, model, calibration, threshold. We score the test period a single time and\n",
    "report each metric with a **bootstrap 95% confidence interval**, plus what the call list would achieve."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 22,
     "data": {
      "text/plain": [
       "                      test  95% CI low  95% CI high\nROC-AUC             0.6181      0.6080       0.6279\nPR-AUC              0.2941      0.2826       0.3054\nBrier               0.1604         NaN          NaN\nPR-AUC of baseline  0.2095         NaN          NaN"
      ],
      "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>test</th>\n      <th>95% CI low</th>\n      <th>95% CI high</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>ROC-AUC</th>\n      <td>0.6181</td>\n      <td>0.6080</td>\n      <td>0.6279</td>\n    </tr>\n    <tr>\n      <th>PR-AUC</th>\n      <td>0.2941</td>\n      <td>0.2826</td>\n      <td>0.3054</td>\n    </tr>\n    <tr>\n      <th>Brier</th>\n      <td>0.1604</td>\n      <td>NaN</td>\n      <td>NaN</td>\n    </tr>\n    <tr>\n      <th>PR-AUC of baseline</th>\n      <td>0.2095</td>\n      <td>NaN</td>\n      <td>NaN</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "76fa2216",
   "metadata": {},
   "source": [
    "X_test, y_test = test[FEATURES], test[\"no_show\"].to_numpy()\n",
    "p_test = calibrated.predict_proba(X_test)[:, 1]\n",
    "rng = np.random.default_rng(SEED)\n",
    "boot = {\"ROC-AUC\": [], \"PR-AUC\": []}\n",
    "for _ in range(1000):\n",
    "    i = rng.integers(0, len(y_test), len(y_test))\n",
    "    boot[\"ROC-AUC\"].append(roc_auc_score(y_test[i], p_test[i]))\n",
    "    boot[\"PR-AUC\"].append(average_precision_score(y_test[i], p_test[i]))\n",
    "test_scores = scores(y_test, p_test)\n",
    "report = pd.DataFrame(\n",
    "    {\n",
    "        \"test\": test_scores,\n",
    "        \"95% CI low\": {k: np.percentile(v, 2.5) for k, v in boot.items()},\n",
    "        \"95% CI high\": {k: np.percentile(v, 97.5) for k, v in boot.items()},\n",
    "    }\n",
    ")\n",
    "report.loc[\"PR-AUC of baseline\", \"test\"] = y_test.mean()\n",
    "report.round(4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 23,
     "data": {
      "text/plain": [
       "share of patients called                             0.211\nshare of no-shows reached (recall)                   0.332\nshare of calls that reach a no-show (precision)      0.329\nnet benefit per 100 appointments, AUD               19.392\nnet benefit if we call everyone, AUD              -122.881\ndtype: float64"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "648f1940",
   "metadata": {},
   "source": [
    "call = p_test >= threshold\n",
    "summary = pd.Series(\n",
    "    {\n",
    "        \"share of patients called\": call.mean(),\n",
    "        \"share of no-shows reached (recall)\": call[y_test == 1].mean(),\n",
    "        \"share of calls that reach a no-show (precision)\": y_test[call].mean(),\n",
    "        \"net benefit per 100 appointments, AUD\": net_benefit_per_100(y_test, p_test, threshold),\n",
    "        \"net benefit if we call everyone, AUD\": net_benefit_per_100(y_test, np.ones_like(p_test), 0.5),\n",
    "    }\n",
    ").round(3)\n",
    "summary"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0af69bad",
   "metadata": {},
   "source": [
    "Compare the model with the obvious alternative policy, \"call everyone\": if that is better, the model is not\n",
    "needed. A positive net benefit that beats calling everyone is the real success measure."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc4dfe17",
   "metadata": {},
   "source": [
    "## Check fairness\n",
    "The model must not flag some groups far more often, or miss their no-shows far more, without a good reason. We\n",
    "compare the flag rate and recall of each group with the overall figures and mark any group that differs by more\n",
    "than 25% for review."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "outputs": [
    {
     "output_type": "execute_result",
     "execution_count": 24,
     "data": {
      "text/plain": [
       "         group  value      n  no-show rate  flag rate  recall  review\n0       gender      F  12448         0.211      0.215   0.335   False\n1       gender      M   6668         0.207      0.205   0.326   False\n2     age_band   0-12   2426         0.254      0.443   0.585    True\n3     age_band  13-17   1853         0.262      0.341   0.458    True\n4     age_band  18-34   6052         0.227      0.258   0.376   False\n5     age_band  35-49   3812         0.190      0.141   0.217    True\n6     age_band  50-64   2363         0.182      0.070   0.118    True\n7     age_band    65+   2610         0.145      0.027   0.058    True\n8  scholarship      0  17212         0.204      0.189   0.302   False\n9  scholarship      1   1904         0.258      0.417   0.545    True"
      ],
      "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>group</th>\n      <th>value</th>\n      <th>n</th>\n      <th>no-show rate</th>\n      <th>flag rate</th>\n      <th>recall</th>\n      <th>review</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>gender</td>\n      <td>F</td>\n      <td>12448</td>\n      <td>0.211</td>\n      <td>0.215</td>\n      <td>0.335</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>gender</td>\n      <td>M</td>\n      <td>6668</td>\n      <td>0.207</td>\n      <td>0.205</td>\n      <td>0.326</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>age_band</td>\n      <td>0-12</td>\n      <td>2426</td>\n      <td>0.254</td>\n      <td>0.443</td>\n      <td>0.585</td>\n      <td>True</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>age_band</td>\n      <td>13-17</td>\n      <td>1853</td>\n      <td>0.262</td>\n      <td>0.341</td>\n      <td>0.458</td>\n      <td>True</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>age_band</td>\n      <td>18-34</td>\n      <td>6052</td>\n      <td>0.227</td>\n      <td>0.258</td>\n      <td>0.376</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>5</th>\n      <td>age_band</td>\n      <td>35-49</td>\n      <td>3812</td>\n      <td>0.190</td>\n      <td>0.141</td>\n      <td>0.217</td>\n      <td>True</td>\n    </tr>\n    <tr>\n      <th>6</th>\n      <td>age_band</td>\n      <td>50-64</td>\n      <td>2363</td>\n      <td>0.182</td>\n      <td>0.070</td>\n      <td>0.118</td>\n      <td>True</td>\n    </tr>\n    <tr>\n      <th>7</th>\n      <td>age_band</td>\n      <td>65+</td>\n      <td>2610</td>\n      <td>0.145</td>\n      <td>0.027</td>\n      <td>0.058</td>\n      <td>True</td>\n    </tr>\n    <tr>\n      <th>8</th>\n      <td>scholarship</td>\n      <td>0</td>\n      <td>17212</td>\n      <td>0.204</td>\n      <td>0.189</td>\n      <td>0.302</td>\n      <td>False</td>\n    </tr>\n    <tr>\n      <th>9</th>\n      <td>scholarship</td>\n      <td>1</td>\n      <td>1904</td>\n      <td>0.258</td>\n      <td>0.417</td>\n      <td>0.545</td>\n      <td>True</td>\n    </tr>\n  </tbody>\n</table>\n</div>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "ce845672",
   "metadata": {},
   "source": [
    "rows = []\n",
    "overall_flag, overall_recall = call.mean(), call[y_test == 1].mean()\n",
    "for g in [\"gender\", \"age_band\", \"scholarship\"]:\n",
    "    for value, idx in test.reset_index(drop=True).groupby(g, observed=True).indices.items():\n",
    "        if len(idx) < 100:\n",
    "            continue\n",
    "        yy, ff = y_test[idx], call[idx]\n",
    "        flag, rec = ff.mean(), ff[yy == 1].mean()\n",
    "        rows.append(\n",
    "            {\n",
    "                \"group\": g,\n",
    "                \"value\": str(value),\n",
    "                \"n\": len(idx),\n",
    "                \"no-show rate\": yy.mean(),\n",
    "                \"flag rate\": flag,\n",
    "                \"recall\": rec,\n",
    "                \"review\": not (0.8 <= flag / overall_flag <= 1.25 and 0.8 <= rec / overall_recall <= 1.25),\n",
    "            }\n",
    "        )\n",
    "fairness = pd.DataFrame(rows)\n",
    "fairness.round(3)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "199b11c4",
   "metadata": {},
   "source": [
    "Groups marked for review need a human decision, not an automatic fix: a group with a genuinely higher no-show\n",
    "rate will be flagged more often. What matters is that the **extra calls help** (a reminder is not a penalty) and\n",
    "that the model is never used to deny care."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b899feb2",
   "metadata": {},
   "source": [
    "## Explain the model\n",
    "**Permutation importance** (on validation, not test): how much PR-AUC drops when a feature is shuffled.\n",
    "**Partial dependence**: how the average prediction changes with one feature."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "09ebe8bb",
   "metadata": {},
   "source": [
    "from sklearn.inspection import PartialDependenceDisplay, permutation_importance\n",
    "\n",
    "sample = val.sample(5000, random_state=SEED)\n",
    "imp = permutation_importance(\n",
    "    calibrated,\n",
    "    sample[FEATURES],\n",
    "    sample[\"no_show\"],\n",
    "    scoring=\"average_precision\",\n",
    "    n_repeats=5,\n",
    "    random_state=SEED,\n",
    "    n_jobs=-1,\n",
    ")\n",
    "importance = pd.Series(imp.importances_mean, index=FEATURES).sort_values()\n",
    "fig, ax = plt.subplots(figsize=(6, 5))\n",
    "importance.plot.barh(ax=ax, color=\"#2E6FF2\")\n",
    "ax.set(title=\"Permutation importance (drop in PR-AUC)\", xlabel=\"\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure>"
      ]
     },
     "metadata": {}
    }
   ],
   "id": "fa4f1b2e",
   "metadata": {},
   "source": [
    "fig, axes = plt.subplots(1, 3, figsize=(12, 3.4))\n",
    "PartialDependenceDisplay.from_estimator(\n",
    "    calibrated,\n",
    "    sample[FEATURES].astype({c: \"float64\" for c in NUMERIC}),\n",
    "    [\"lead_days\", \"age\", \"prior_no_show_rate\"],\n",
    "    ax=axes,\n",
    "    grid_resolution=30,\n",
    ")\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "81243916",
   "metadata": {},
   "source": [
    "Check the story against common sense: risk should rise with lead time and with past no-shows. A surprising top\n",
    "feature is often a sign of leakage, so this step doubles as a final leak check."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "01060c6c",
   "metadata": {},
   "source": [
    "## Save the model\n",
    "Save the model **with** everything needed to use it correctly: the threshold, the feature list and versions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "saved models/noshow_model.joblib\n"
     ]
    }
   ],
   "id": "7dc51805",
   "metadata": {},
   "source": [
    "import joblib\n",
    "import sklearn\n",
    "\n",
    "bundle = {\n",
    "    \"model\": calibrated,\n",
    "    \"threshold\": float(threshold),\n",
    "    \"features\": FEATURES,\n",
    "    \"model_name\": best_name,\n",
    "    \"sklearn\": sklearn.__version__,\n",
    "    \"validation_start\": str(validation_start.date()),\n",
    "    \"test_start\": str(test_start.date()),\n",
    "}\n",
    "Path(\"models\").mkdir(exist_ok=True)\n",
    "joblib.dump(bundle, \"models/noshow_model.joblib\")\n",
    "print(\"saved models/noshow_model.joblib\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e388bace",
   "metadata": {},
   "source": [
    "## From notebook to production\n",
    "This notebook is the project: it is where the answer was found. If the clinic wants the call list **every day**,\n",
    "a notebook is not enough: it has to run unattended, be tested on every change, and answer requests from the booking\n",
    "system. So the next step is to move this notebook's code, unchanged, into a Python package with tests, an API and\n",
    "monitoring. The reference repository for this project (`p01-noshow.zip`) shows the result:\n",
    "\n",
    "| notebook section | moves to | tested by |\n",
    "|---|---|---|\n",
    "| Check data quality | `src/noshow/data/schema.py` | `tests/test_data.py` |\n",
    "| Clean the data | `src/noshow/data/clean.py` | `tests/test_data.py` |\n",
    "| Split by time | `src/noshow/features/split.py` | `tests/test_features.py` |\n",
    "| Engineer features | `src/noshow/features/build.py` | `tests/test_features.py` (including leakage) |\n",
    "| Models | `src/noshow/models/pipelines.py`, `train.py` | `tests/test_model_and_api.py` |\n",
    "| Test, fairness, explain | `src/noshow/evaluation/report.py` → `MODEL_CARD.md` | `tests/test_metrics.py` |\n",
    "| Use the model | `src/noshow/serving/app.py` (API), `Dockerfile` | `tests/test_model_and_api.py` |\n",
    "\n",
    "Moving code must not change results. When this notebook sits inside that repository, the next cell proves it: the\n",
    "package's cleaning and features give the same table, and its model with our tuned settings gives the same predictions.\n",
    "On its own (for example in Colab), the cell simply skips."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "notebook and production package agree: same clean data, same features, same predictions\n"
     ]
    }
   ],
   "id": "043c4336",
   "metadata": {},
   "source": [
    "try:\n",
    "    from noshow.data.clean import clean as pkg_clean\n",
    "    from noshow.features.build import FEATURES as PKG_FEATURES\n",
    "    from noshow.features.build import build_features as pkg_build_features\n",
    "    from noshow.models.pipelines import build as pkg_build\n",
    "except ImportError:\n",
    "    print(\"the production package is not installed here, so the comparison is skipped\")\n",
    "else:\n",
    "    pkg_clean_df, _ = pkg_clean(raw, max_age=115)\n",
    "    pd.testing.assert_frame_equal(pkg_clean_df, clean)\n",
    "    pd.testing.assert_frame_equal(\n",
    "        pkg_build_features(pkg_clean_df).reset_index(drop=True), feats.reset_index(drop=True)\n",
    "    )\n",
    "    assert PKG_FEATURES == FEATURES\n",
    "    name = {\"logistic\": \"logistic\", \"gradient boosting\": \"hist_gradient_boosting\"}[best_name]\n",
    "    best_params = {\"logistic\": search_lr, \"gradient boosting\": search_hgb}[best_name].best_params_\n",
    "    pkg_model = pkg_build(name, SEED).set_params(**best_params).fit(X_train, y_train)\n",
    "    np.testing.assert_allclose(pkg_model.predict_proba(X_val)[:, 1], best.predict_proba(X_val)[:, 1])\n",
    "    print(\"notebook and production package agree: same clean data, same features, same predictions\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "26b6226a",
   "metadata": {},
   "source": [
    "In the repository, the whole pipeline then runs from a terminal:\n",
    "\n",
    "```bash\n",
    "make all      # validate -> prepare -> eda -> train -> evaluate, writes reports/MODEL_CARD.md\n",
    "make test     # 42 tests\n",
    "make serve    # the prediction API on http://127.0.0.1:8000\n",
    "```\n",
    "\n",
    "## What you built\n",
    "* A problem framed by its decision and costs, with a threshold set before modelling.\n",
    "* Clean, validated data, split by time, with history features that cannot leak the future.\n",
    "* A baseline, two tuned models, calibration, one honest test with confidence intervals.\n",
    "* A fairness check, explanations and a saved model, and proof that the production code matches the notebook.\n",
    "\n",
    "**Try next:** add the number of days until the appointment's weekday holiday; change the costs and watch the\n",
    "threshold move; compare isotonic with sigmoid calibration."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.12"
  },
  "colab": {
   "provenance": [],
   "toc_visible": true
  },
  "title": "Will the patient show up?"
 },
 "nbformat": 4,
 "nbformat_minor": 5
}
