meta-edc 1.0.6__py3-none-any.whl → 1.1.0__py3-none-any.whl

meta_analytics/notebooks/hiv_regimens.ipynb

@@ -0,0 +1,429 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "0",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%%capture\n",
+    "import os\n",
+    "from pathlib import Path\n",
+    "import pandas as pd\n",
+    "from dj_notebook import activate\n",
+    "\n",
+    "env_file = os.environ[\"META_ENV\"]\n",
+    "reports_folder = Path(os.environ[\"META_REPORTS_FOLDER\"])\n",
+    "analysis_folder = Path(os.environ[\"META_ANALYSIS_FOLDER\"])\n",
+    "pharmacy_folder = Path(os.environ[\"META_PHARMACY_FOLDER\"])\n",
+    "plus = activate(dotenv_file=env_file)\n",
+    "pd.set_option('future.no_silent_downcasting', True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "1",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from edc_pdutils.dataframes import get_crf, get_subject_visit\n",
+    "from edc_constants.constants import YES\n",
+    "from edc_appointment.analytics import get_appointment_df\n",
+    "from datetime import datetime"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "2",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "cutoff_datetime = datetime(2026,3,1)\n",
+    "df_patienthistory = get_crf(\"meta_subject.patienthistory\", subject_visit_model=\"meta_subject.subjectvisit\")\n",
+    "df_followup_examination = get_crf(\"meta_subject.FollowupExamination\", subject_visit_model=\"meta_subject.subjectvisit\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "replacements = {\n",
+    "    \"ABC+ 3TC+ DTG\": \"ABC + 3TC + DTG\",\n",
+    "    \"ABC+3TC+DTG\": \"ABC + 3TC + DTG\",\n",
+    "    \"ABC, 3TC, DTG\":\"ABC + 3TC + DTG\",\n",
+    "    \"ABC, EFV, DTG\": \"ABC + EFV + DTG\",\n",
+    "    \"TDF+FTC+DTG\": \"TDF + FTC + DTG\",\n",
+    "    \"AZT+3TC+DTG\": \"AZT + 3TC + DTG\",\n",
+    "    \"AZT + 3 TC + DTG\":\"AZT + 3TC + DTG\",\n",
+    "    \"TDF+3TC+DTG\": \"TDF + 3TC + DTG\",\n",
+    "}\n",
+    "\n",
+    "df_patienthistory[\"other_current_arv_regimen\"] = (\n",
+    "    df_patienthistory[\"other_current_arv_regimen\"]\n",
+    "    .replace(replacements)\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "4",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_patienthistory['regimen'] = df_patienthistory[\"current_arv_regimen\"]\n",
+    "df_patienthistory.loc[df_patienthistory[\"current_arv_regimen\"]==\"Other, specify ...\", \"regimen\"] = df_patienthistory[\"other_current_arv_regimen\"]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "5",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_followup_examination[\"art_new_regimen_other\"] = (\n",
+    "    df_followup_examination[\"art_new_regimen_other\"]\n",
+    "    .replace(replacements)\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "6",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_followup_examination['regimen'] = pd.NA\n",
+    "df_followup_examination.loc[(df_followup_examination[\"art_change\"]==YES) & (df_followup_examination.art_new_regimen_other.notna()), \"regimen\"] = df_followup_examination[\"art_new_regimen_other\"]\n",
+    "df_followup_examination.loc[(df_followup_examination[\"art_change\"]==YES) & (df_followup_examination.art_new_regimen_other.isna()), \"regimen\"] = \"CHANGE_NOT_REPORTED\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "7",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_regimen = pd.concat([df_patienthistory[[\"subject_identifier\", \"visit_datetime\", \"regimen\"]], df_followup_examination[[\"subject_identifier\", \"visit_datetime\", \"regimen\"]]])\n",
+    "df_regimen[\"regimen\"] = df_regimen[\"regimen\"].replace({\"Other second line\": \"CHANGE_NOT_REPORTED\"})\n",
+    "df_regimen[\"regimen\"] = pd.Categorical(df_regimen[\"regimen\"], categories=list(df_regimen.query(\"regimen.notna()\").regimen.unique()), ordered=False)\n",
+    "df_regimen = df_regimen.sort_values([\"subject_identifier\", \"visit_datetime\"])\n",
+    "df_regimen = df_regimen.reset_index(drop=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_pivot = df_regimen.pivot_table(values=\"visit_datetime\", columns=\"regimen\", index=\"subject_identifier\", observed=True)\n",
+    "df_pivot = df_pivot.reset_index()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "subject_identifier = \"105-20-0050-0\"\n",
+    "df_pivot[df_pivot.subject_identifier==subject_identifier].melt().query(\"value.notna() and regimen!='subject_identifier'\").sort_values(\"value\", ascending=False)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "10",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_melt = df_pivot.melt(id_vars=[\"subject_identifier\"]).query(\"value.notna()\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "11",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_melt[\"max_date\"] = df_melt.groupby(\"subject_identifier\")[\"value\"].transform(\"max\")\n",
+    "df_melt[\"current_regimen\"] = df_melt[df_melt.value==df_melt.max_date][\"regimen\"]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "12",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_current_regimens = df_melt.query(\"current_regimen.notna()\")[[\"subject_identifier\", \"max_date\", \"current_regimen\"]].copy()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "13",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_visit = get_subject_visit(model=\"meta_subject.subjectvisit\")\n",
+    "df_visit = df_visit[df_visit.visit_code==1000.0].copy()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "14",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_appointment = get_appointment_df()\n",
+    "df_appointment_next = (\n",
+    "    df_appointment\n",
+    "    .groupby(by=[\"subject_identifier\", \"next_appt_datetime\", \"next_visit_code\"])\n",
+    "    .size()\n",
+    "    .copy()\n",
+    "    .reset_index()\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "15",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_appointment_last = (\n",
+    "    df_appointment[df_appointment.appt_datetime<cutoff_datetime][[\"subject_identifier\", \"appt_datetime\", \"visit_code\"]]\n",
+    "    .sort_values([\"subject_identifier\", \"appt_datetime\", \"visit_code\"])\n",
+    "    .groupby(by=[\"subject_identifier\"])\n",
+    "    .agg([\"last\"])\n",
+    "    .reset_index()                       )\n",
+    "df_appointment_last.columns = [\"_\".join(col).strip() for col in df_appointment_last.columns.values]\n",
+    "df_appointment_last = (\n",
+    "    df_appointment_last\n",
+    "        .rename(columns={\n",
+    "            \"subject_identifier_\":\"subject_identifier\",\n",
+    "            \"appt_datetime_last\":\"last_appt_datetime\",\n",
+    "            \"visit_code_last\":\"last_visit_code\"\n",
+    "        }\n",
+    "    )\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "16",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df = df_current_regimens.merge(df_visit[[ \"subject_identifier\", \"baseline_datetime\", \"endline_visit_datetime\", \"endline_visit_code\"]], on=\"subject_identifier\", how=\"left\")\n",
+    "df = df.reset_index(drop=True)\n",
+    "df[\"changed\"] = False\n",
+    "df.loc[df.max_date != df.baseline_datetime, \"changed\"] = True"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "17",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df = df.merge(df_appointment_next[[\"subject_identifier\", \"next_appt_datetime\", \"next_visit_code\"]], on=\"subject_identifier\", how=\"left\")\n",
+    "df = df.merge(df_appointment_last[[\"subject_identifier\", \"last_appt_datetime\", \"last_visit_code\"]], on=\"subject_identifier\", how=\"left\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "18",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# from last seen to final scheduled appt\n",
+    "df[\"remaining_delta_from_last_seen\"] = df.last_appt_datetime - df.endline_visit_datetime\n",
+    "df[\"remaining_delta_from_last_seen\"] = df[\"remaining_delta_from_last_seen\"].apply(lambda x: 0 if x.total_seconds()<0 else x)\n",
+    "df[\"remaining_delta_from_last_seen\"] = pd.to_timedelta(df[\"remaining_delta_from_last_seen\"])\n",
+    "df[\"remaining_days_last_seen_to_final\"] = df[\"remaining_delta_from_last_seen\"].dt.days\n",
+    "\n",
+    "# from now to final scheduled appt\n",
+    "df[\"remaining_delta_from_now\"] = 0.0\n",
+    "df[\"remaining_delta_from_now\"] = df[df.remaining_days_last_seen_to_final>0].last_appt_datetime - datetime.now()\n",
+    "df[\"remaining_delta_from_now\"] = df[\"remaining_delta_from_now\"].apply(lambda x: 0 if x.total_seconds()<0 else x)\n",
+    "df[\"remaining_delta_from_now\"] = pd.to_timedelta(df[\"remaining_delta_from_now\"])\n",
+    "df[\"remaining_days_now_to_final\"] = df[\"remaining_delta_from_now\"].dt.days\n",
+    "\n",
+    "# from next appointment to final scheduled appt\n",
+    "df[\"remaining_delta_from_next\"] = 0.0\n",
+    "df[\"remaining_delta_from_next\"] = df[df.remaining_days_last_seen_to_final>0].last_appt_datetime - df[df.remaining_days_last_seen_to_final>0].next_appt_datetime\n",
+    "df[\"remaining_delta_from_next\"] = df[\"remaining_delta_from_next\"].apply(lambda x: 0 if x.total_seconds()<0 else x)\n",
+    "df[\"remaining_delta_from_next\"] = pd.to_timedelta(df[\"remaining_delta_from_next\"])\n",
+    "df[\"remaining_days_next_to_final\"] = df[\"remaining_delta_from_next\"].dt.days"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "19",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_final = (\n",
+    "    df\n",
+    "    .rename(columns={\n",
+    "        \"max_date\": \"current_regimen_date\",\n",
+    "        \"endline_visit_code\": \"last_attended_visit_code\",\n",
+    "        \"endline_visit_datetime\": \"last_attended_visit_datetime\",\n",
+    "    })\n",
+    "    .copy()\n",
+    ")\n",
+    "df_final = df_final[[\n",
+    "    \"subject_identifier\",\n",
+    "    \"current_regimen\",\n",
+    "    \"current_regimen_date\",\n",
+    "    \"changed\",\n",
+    "    \"baseline_datetime\",\n",
+    "    \"last_attended_visit_code\",\n",
+    "    \"last_attended_visit_datetime\",\n",
+    "    \"next_visit_code\",\n",
+    "    \"next_appt_datetime\",\n",
+    "    \"last_visit_code\",\n",
+    "    \"last_appt_datetime\",\n",
+    "    \"remaining_days_last_seen_to_final\",\n",
+    "    \"remaining_days_now_to_final\",\n",
+    "    \"remaining_days_next_to_final\",\n",
+    "]].copy()\n",
+    "\n",
+    "df_final = (\n",
+    "    df_final\n",
+    "    .sort_values(\"subject_identifier\")\n",
+    "    .reset_index(drop=True)\n",
+    ")\n",
+    "df_final[\"remaining_days_last_seen_to_final\"] = df_final[\"remaining_days_last_seen_to_final\"].astype(\"float64\").fillna(0)\n",
+    "df_final[\"remaining_days_now_to_final\"] = df_final[\"remaining_days_now_to_final\"].astype(\"float64\").fillna(0)\n",
+    "df_final[\"remaining_days_next_to_final\"] = df_final[\"remaining_days_next_to_final\"].astype(\"float64\").fillna(0)\n",
+    "df_final"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "20",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# need from now until end of study\n",
+    "df_summary1 = (pd.merge(\n",
+    "    df_final.groupby(by=[\"current_regimen\"]).remaining_days_now_to_final.sum(),\n",
+    "    df_final.groupby(by=[\"current_regimen\"]).subject_identifier.count(), on=\"current_regimen\")\n",
+    "    .rename(columns={\n",
+    "        \"remaining_days_now_to_final\": \"days_medication_needed\",\n",
+    "        \"subject_identifier\": \"subjects\"\n",
+    "    })\n",
+    "    .sort_values(\"days_medication_needed\", ascending=False)\n",
+    "    .reset_index()\n",
+    ")\n",
+    "df_summary1"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "21",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# need from last seen to end of study\n",
+    "df_summary2 = (pd.merge(\n",
+    "    df_final.groupby(by=[\"current_regimen\"]).remaining_days_last_seen_to_final.sum(),\n",
+    "    df_final.groupby(by=[\"current_regimen\"]).subject_identifier.count(), on=\"current_regimen\")\n",
+    "    .rename(columns={\n",
+    "        \"remaining_days_last_seen_to_final\": \"days_medication_needed\",\n",
+    "        \"subject_identifier\": \"subjects\"\n",
+    "    })\n",
+    "    .sort_values(\"days_medication_needed\", ascending=False)\n",
+    "    .reset_index()\n",
+    ")\n",
+    "df_summary2"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "22",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# need from next to end of study\n",
+    "df_summary3 = (pd.merge(\n",
+    "    df_final.groupby(by=[\"current_regimen\"]).remaining_days_next_to_final.sum(),\n",
+    "    df_final.groupby(by=[\"current_regimen\"]).subject_identifier.count(), on=\"current_regimen\")\n",
+    "    .rename(columns={\n",
+    "        \"remaining_days_next_to_final\": \"days_medication_needed\",\n",
+    "        \"subject_identifier\": \"subjects\"\n",
+    "    })\n",
+    "    .sort_values(\"days_medication_needed\", ascending=False)\n",
+    "    .reset_index()\n",
+    ")\n",
+    "\n",
+    "df_summary3"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "23",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "with pd.ExcelWriter(\n",
+    "        analysis_folder / \"hiv_medication.xlsx\",\n",
+    "        date_format=\"YYYY-MM-DD\",\n",
+    "        datetime_format=\"YYYY-MM-DD HH:MM:SS\"\n",
+    ") as writer:\n",
+    "    df_final.to_excel(writer, sheet_name=\"subjects\", index=False)\n",
+    "    df_summary1.to_excel(writer, sheet_name=\"now to final\", index=False)\n",
+    "    df_summary2.to_excel(writer, sheet_name=\"last seen to final\", index=False)\n",
+    "    df_summary3.to_excel(writer, sheet_name=\"next to final\", index=False)\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 2
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython2",
+   "version": "2.7.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

meta_analytics/notebooks/incidence.ipynb

@@ -0,0 +1,232 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "0",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%%capture\n",
+    "import os\n",
+    "from pathlib import Path\n",
+    "import pandas as pd\n",
+    "from dj_notebook import activate\n",
+    "\n",
+    "env_file = os.environ[\"META_ENV\"]\n",
+    "reports_folder = Path(os.environ[\"META_REPORTS_FOLDER\"])\n",
+    "analysis_folder = Path(os.environ[\"META_ANALYSIS_FOLDER\"])\n",
+    "pharmacy_folder = Path(os.environ[\"META_PHARMACY_FOLDER\"])\n",
+    "plus = activate(dotenv_file=env_file)\n",
+    "pd.set_option('future.no_silent_downcasting', True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "1",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "\n",
+    "import pdfkit\n",
+    "from datetime import date\n",
+    "from edc_pdutils.dataframes import get_subject_visit\n",
+    "from meta_analytics.dataframes import get_eos_df\n",
+    "from meta_analytics.dataframes import GlucoseEndpointsByDate\n",
+    "from scipy.stats import chi2\n",
+    "from meta_analytics.utils import df_as_great_table\n",
+    "from great_tables import md\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "2",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "html_data = []\n",
+    "cutoff_date = date(2025,3, 31)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "3",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_visit = get_subject_visit(\"meta_subject.subjectvisit\")\n",
+    "df_visit = df_visit[df_visit.appt_datetime.dt.date<=cutoff_date]\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "4",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "cls = GlucoseEndpointsByDate()\n",
+    "cls.run()\n",
+    "df_endpoint = cls.endpoint_only_df.copy()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "5",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def get_df_main(df_visit:pd.DataFrame, lower_days:float|None=None, upper_days:float|None=None):\n",
+    "    if not lower_days:\n",
+    "        lower_days = -1\n",
+    "    cutoff_datetime = df_visit.query(\"@lower_days<followup_days<=@upper_days\").visit_datetime.max()\n",
+    "    # exclude subjects for this reason\n",
+    "    offstudy_reasons = ['Patient fulfilled late exclusion criteria (due to abnormal blood values or raised blood pressure at enrolment']\n",
+    "\n",
+    "    df_eos = get_eos_df()\n",
+    "    df_eos_excluded = (\n",
+    "        df_eos\n",
+    "        .query(\"followup_days<@lower_days and followup_days<=@upper_days and offstudy_reason.isin(@offstudy_reasons)\")\n",
+    "        .copy()\n",
+    "        .reset_index()\n",
+    "    )\n",
+    "    df_visit_final = (\n",
+    "        df_visit.query(\"@lower_days<followup_days<=@upper_days and reason!='missed'\")\n",
+    "        .merge(df_eos_excluded[[\"subject_identifier\"]], on=\"subject_identifier\", how=\"left\", suffixes=(\"\", \"_y\"), indicator=True)\n",
+    "        .query(\"_merge=='left_only'\")\n",
+    "        .drop(columns=[\"_merge\"])\n",
+    "    )\n",
+    "    df_main = (\n",
+    "        df_visit_final\n",
+    "        .groupby(by=[\"subject_identifier\"])[[\"baseline_datetime\", \"visit_datetime\", \"followup_days\"]]\n",
+    "        .max()\n",
+    "        .reset_index()\n",
+    "    )\n",
+    "\n",
+    "    df_main = (\n",
+    "        df_main\n",
+    "        .merge(\n",
+    "            df_endpoint.query(\"days_to_endpoint>@lower_days\")[[\"subject_identifier\", \"endpoint_label\", \"endpoint_type\", \"days_to_endpoint\"]],\n",
+    "            how=\"left\",\n",
+    "            on=[\"subject_identifier\"])\n",
+    "        .reset_index(drop=True)\n",
+    "    )\n",
+    "    if lower_days>=365.25:\n",
+    "        df_main[\"followup_days\"] = df_main[\"followup_days\"] - lower_days\n",
+    "    df_main[\"followup_years\"] = df_main[\"followup_days\"]/365.25\n",
+    "    return df_main, len(df_main), len(df_main.query(\"@lower_days<days_to_endpoint<=@upper_days and endpoint_label.notna()\"))\n",
+    "\n",
+    "def get_rate_and_ci(events, person_years_total):\n",
+    "    lower_ci = (chi2.ppf(0.025, 2 * events) / (2 * person_years_total)) * 1000\n",
+    "    upper_ci = (chi2.ppf(0.975, 2 * (events + 1)) / (2 * person_years_total)) * 1000\n",
+    "    return events/person_years_total*1000, lower_ci, upper_ci\n",
+    "\n",
+    "def get_incidence_data(term:str, lower_days:float, upper_days:float):\n",
+    "    data = {}\n",
+    "    df_main, subjects, events = get_df_main(df_visit, lower_days=lower_days, upper_days=upper_days)\n",
+    "    person_years_total = df_main.followup_years.sum()\n",
+    "    data.update({term:[person_years_total, subjects, events, *get_rate_and_ci(events, person_years_total)]})\n",
+    "    return data"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "6",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "incidence_data = {}\n",
+    "incidence_data.update(get_incidence_data(\"total\", lower_days=0, upper_days=10000))\n",
+    "incidence_data.update(get_incidence_data(\"0-1 years\", lower_days=0, upper_days=365.25))\n",
+    "incidence_data.update(get_incidence_data(\"1-2 years\", lower_days=365.25, upper_days=2*365.25))\n",
+    "incidence_data.update(get_incidence_data(\"2-3 years\", lower_days=2*365.25, upper_days=3*365.25))\n",
+    "incidence_data.update(get_incidence_data(\"3+ years\", lower_days=3*365.25, upper_days=10*365.25))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "7",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "data = dict(label=[], person_years=[], failures=[], rate=[], lower_ci=[], upper_ci=[])\n",
+    "for k in incidence_data:\n",
+    "    data[\"label\"].append(k)\n",
+    "\n",
+    "for v in incidence_data.values():\n",
+    "    data[\"person_years\"].append(v[0])\n",
+    "    data[\"failures\"].append(v[2])\n",
+    "    data[\"rate\"].append(v[3])\n",
+    "    data[\"lower_ci\"].append(v[4])\n",
+    "    data[\"upper_ci\"].append(v[5])\n",
+    "\n",
+    "df = pd.DataFrame(data=data)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "gt = df_as_great_table(\n",
+    "    df,\n",
+    "    title=\"Table 9: Incident Rate per 1000 person years\",\n",
+    "    subtitle=md(\"using randomisation to diabetes/last seen\"),\n",
+    ")\n",
+    "gt = gt.fmt_number(columns=[\"person_years\", \"failures\", \"rate\", \"lower_ci\", \"upper_ci\"], decimals=2)\n",
+    "gt = (gt\n",
+    "    .cols_label({\"label\": \"Label\", \"person_years\": \"Person years\", \"failures\": \"Failures\", \"rate\": \"Rate\", \"lower_ci\": \"Lower\", \"upper_ci\": \"Upper\"})\n",
+    "    .cols_align(align=\"left\", columns=[\"label\"])\n",
+    "    .cols_align(align=\"center\", columns=[\"person_years\", \"failures\", \"rate\", \"lower_ci\", \"upper_ci\"])\n",
+    "    .tab_spanner(\n",
+    "        label=\"95%CI\",\n",
+    "        columns=[\"lower_ci\", \"upper_ci\"],\n",
+    "    )\n",
+    ")\n",
+    "gt.show()\n",
+    "html_data.append(gt.as_raw_html())\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "9",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "raw_html = \"</BR>\".join(html_data)\n",
+    "raw_html = '<!DOCTYPE html>\\n<html lang=\"en\">\\n<head>\\n<meta charset=\"utf-8\"/>\\n</head>\\n<body>\\n' + raw_html + '\\n</body>\\n</html>\\n'\n",
+    "pdfkit.from_string(raw_html, str(analysis_folder / \"incident_rate.pdf\"))\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "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": "3.12.4"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}