{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "c2188cd7",
   "metadata": {},
   "source": [
    "# Read data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5b2b0060",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "52f55dde",
   "metadata": {},
   "outputs": [],
   "source": [
    "activity_cols = {\n",
    "    \"Datum\": \"date\",\n",
    "    \"Ø Pace\": \"activity_pace_average\",\n",
    "    \"Kalorien\": \"activity_calories\",\n",
    "    \"Zeit\": \"activity_duration\",\n",
    "    \"Ø Herzfrequenz\": \"activity_heart_rate_average\",\n",
    "    \"Ø Atemfrequenz\": \"activity_breathing_rate_average\",\n",
    "}\n",
    "df_activities = pd.read_csv(\n",
    "    \"data/raw/all_activities.csv\",\n",
    "    usecols=list(activity_cols.keys())\n",
    ")\n",
    "df_activities.rename(columns=activity_cols, inplace=True)\n",
    "\n",
    "sleep_cols = {\n",
    "    \"Sleep Score 7 Tage\": \"date\",\n",
    "    \"Score\": \"sleep_score\",\n",
    "    \"Dauer\": \"sleep_duration\",\n",
    "    \"Schlafenszeit\": \"sleep_bedtime\",\n",
    "    \"Ruheherzfrequenz\": \"sleep_resting_heart_rate\",\n",
    "    \"HFV-Status\": \"sleep_hrv_status\",\n",
    "    \"Atmung\": \"sleep_breathing_rate\",\n",
    "    \"Schlafbedürfnis\": \"sleep_duration_needed\",\n",
    "}\n",
    "df_sleep = pd.read_csv(\n",
    "    \"data/raw/sleep.csv\",\n",
    "    usecols=list(sleep_cols.keys())\n",
    ")\n",
    "df_sleep.rename(columns=sleep_cols, inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "91f04fe7",
   "metadata": {},
   "source": [
    "# Clean data/time features"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "97252627",
   "metadata": {},
   "source": [
    "## Activities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "826e5af0",
   "metadata": {},
   "outputs": [],
   "source": [
    "def pace_to_seconds(pace_str):\n",
    "    try:\n",
    "        if isinstance(pace_str, str) and \":\" in pace_str:\n",
    "            parts = pace_str.split(\":\")\n",
    "            if len(parts) == 2:\n",
    "                minutes, seconds = int(parts[0]), int(parts[1])\n",
    "                return minutes * 60 + seconds\n",
    "            elif len(parts) == 3:  # If format is HH:MM:SS\n",
    "                hours, minutes, seconds = int(\n",
    "                    parts[0]), int(parts[1]), int(parts[2])\n",
    "                return hours * 3600 + minutes * 60 + seconds\n",
    "        return float(pace_str)\n",
    "    except Exception:\n",
    "        return 0\n",
    "\n",
    "\n",
    "df_activities[\"date\"] = pd.to_datetime(df_activities[\"date\"], errors=\"coerce\")\n",
    "df_activities[\"activity_starting_time\"] = df_activities[\"date\"] - df_activities[\"date\"].dt.normalize()\n",
    "df_activities[\"date\"] = df_activities[\"date\"].dt.normalize()\n",
    "\n",
    "df_activities[\"activity_pace_average\"] = df_activities[\"activity_pace_average\"].apply(pace_to_seconds)\n",
    "df_activities[\"activity_duration\"] = pd.to_timedelta(df_activities[\"activity_duration\"], errors=\"coerce\")\n",
    "df_activities[\"activity_duration_seconds\"] = df_activities[\"activity_duration\"].dt.total_seconds().fillna(0).astype(int)\n",
    "\n",
    "df_activities[\"activity_ending_time\"] = df_activities[\"activity_starting_time\"] + df_activities[\"activity_duration\"]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2cbef16d",
   "metadata": {},
   "outputs": [],
   "source": [
    "numeric_columns = [\n",
    "    \"activity_calories\",\n",
    "    \"activity_heart_rate_average\",\n",
    "    \"activity_pace_average\",\n",
    "    \"activity_breathing_rate_average\",\n",
    "]\n",
    "\n",
    "for col in numeric_columns:\n",
    "    if df_activities[col].dtype == \"object\":\n",
    "        df_activities[col] = pd.to_numeric(df_activities[col].str.replace(',', '').replace('--', '0'), errors='coerce')\n",
    "    else:\n",
    "        df_activities[col] = pd.to_numeric(df_activities[col], errors='coerce')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "708414b6",
   "metadata": {},
   "source": [
    "## Sleep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d3ec20f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_sleep[\"date\"] = pd.to_datetime(df_sleep[\"date\"], errors=\"coerce\")\n",
    "\n",
    "df_sleep = df_sleep[df_sleep[\"sleep_bedtime\"] != \"--\"]\n",
    "\n",
    "df_sleep[\"sleep_bedtime\"] = pd.to_timedelta(\n",
    "    pd.to_datetime(df_sleep[\"sleep_bedtime\"].astype(str), format=\"%I:%M %p\").dt.hour * 3600 +\n",
    "    pd.to_datetime(df_sleep[\"sleep_bedtime\"].astype(str), format=\"%I:%M %p\").dt.minute * 60,\n",
    "    unit=\"s\"\n",
    ")\n",
    "\n",
    "# Handle AM times by adding 24 hours to times before noon\n",
    "mask = df_sleep[\"sleep_bedtime\"].dt.components['hours'] < 12\n",
    "df_sleep.loc[mask, \"sleep_bedtime\"] = df_sleep.loc[mask, \"sleep_bedtime\"] + pd.Timedelta(days=1)\n",
    "\n",
    "df_sleep[\"sleep_duration\"] = pd.to_timedelta(df_sleep[\"sleep_duration\"], errors=\"coerce\")\n",
    "df_sleep[\"sleep_duration_seconds\"] = (df_sleep[\"sleep_duration\"]).dt.total_seconds()\n",
    "df_sleep[\"sleep_duration_needed\"] = pd.to_timedelta(df_sleep[\"sleep_duration_needed\"], errors=\"coerce\")\n",
    "df_sleep[\"sleep_duration_needed_seconds\"] = (df_sleep[\"sleep_duration_needed\"]).dt.total_seconds() \n",
    "\n",
    "df_sleep[\"sleep_duration_needed_delta\"] = df_sleep[\"sleep_duration\"] - df_sleep[\"sleep_duration_needed\"]\n",
    "df_sleep[\"sleep_duration_needed_delta_seconds\"] = (df_sleep[\"sleep_duration_needed_delta\"]).dt.total_seconds()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "845cc713",
   "metadata": {},
   "outputs": [],
   "source": [
    "numeric_columns = [\n",
    "    \"sleep_score\",\n",
    "    \"sleep_resting_heart_rate\",\n",
    "    \"sleep_hrv_status\",\n",
    "    \"sleep_breathing_rate\",\n",
    "]\n",
    "\n",
    "for col in numeric_columns:\n",
    "    if df_sleep[col].dtype == \"object\":\n",
    "        df_sleep[col] = pd.to_numeric(df_sleep[col].str.replace(',', '').replace('--', '0'), errors='coerce')\n",
    "    else:\n",
    "        df_sleep[col] = pd.to_numeric(df_sleep[col], errors='coerce')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fabceba5",
   "metadata": {},
   "source": [
    "## Combined"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "05da5fe7",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_latest_activity = df_activities.sort_values('activity_ending_time').groupby('date', as_index=False).last()\n",
    "\n",
    "df_combined = pd.merge(df_latest_activity, df_sleep, on='date', how='right')\n",
    "\n",
    "if \"activity_ending_time\" in df_combined.columns and \"sleep_bedtime\" in df_combined.columns:\n",
    "    df_combined[\"bedtime_activity_ending_delta\"] = df_combined[\"sleep_bedtime\"] - df_combined[\"activity_ending_time\"]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "427af06c",
   "metadata": {},
   "source": [
    "## Save cleaned data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6a8888ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_combined.to_csv(\"data/cleaned/combined_activities_sleep.csv\", index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "730ab7c4",
   "metadata": {},
   "source": [
    "## Data overview"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "47ffa998",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_combined.head(30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ce818c76",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_combined.dtypes"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4c848335",
   "metadata": {},
   "source": [
    "# Visualizations"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e228914c",
   "metadata": {},
   "source": [
    "## Corelation Matrix sleep after all activities"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "41080d47",
   "metadata": {},
   "outputs": [],
   "source": [
    "filtered_all_activities = df_combined[(df_combined[\"activity_calories\"].notna())]\n",
    "correlation_matrix = df_combined.corr(numeric_only=True)\n",
    "\n",
    "plt.figure(figsize=(12, 10))\n",
    "sns.heatmap(correlation_matrix, annot=True, cmap=\"coolwarm\", center=0)\n",
    "plt.title(\"Korrelationsmatrix - Alle Aktivitären\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d5ba27c4",
   "metadata": {},
   "source": [
    "## Corelation Matrix sleep after activities < 4h before sleep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cf54e6c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "df_combine_activities_4_hours_before_sleep = df_combined[(df_combined[\"activity_calories\"].notna()) & (\n",
    "    df_combined[\"bedtime_activity_ending_delta\"] < pd.Timedelta(hours=4))]\n",
    "\n",
    "correlation_matrix = df_combine_activities_4_hours_before_sleep.corr(\n",
    "    numeric_only=True)\n",
    "\n",
    "plt.figure(figsize=(12, 10))\n",
    "sns.heatmap(correlation_matrix, annot=True, cmap=\"coolwarm\", center=0)\n",
    "plt.title(\"Korrelationsmatrix- Trainings 4 Stunden vor dem Schlafen\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "634ee858",
   "metadata": {},
   "source": [
    "## Scatter plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e2246df2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "filtered_df_combined_4h_before_sleep = df_combine_activities_4_hours_before_sleep[\n",
    "    (df_combine_activities_4_hours_before_sleep[\"activity_calories\"].notna())\n",
    "]\n",
    "\n",
    "blue_count = len(filtered_df_combined_4h_before_sleep)\n",
    "\n",
    "plt.figure(figsize=(8, 6))\n",
    "plt.scatter(\n",
    "    filtered_df_combined_4h_before_sleep['activity_heart_rate_average'],\n",
    "    filtered_df_combined_4h_before_sleep['sleep_score'],\n",
    "    alpha=0.7,\n",
    "    color='blue',\n",
    "    label=f'Training <4h vor Schlaf (n={blue_count})'\n",
    ")\n",
    "filtered_df_combined_more_than_4h_before_sleep = df_combined[\n",
    "    (df_combined[\"activity_calories\"].notna()) &\n",
    "    (df_combined[\"bedtime_activity_ending_delta\"] > pd.Timedelta(hours=4))  # &\n",
    "]\n",
    "\n",
    "red_count = len(filtered_df_combined_more_than_4h_before_sleep)\n",
    "plt.scatter(\n",
    "    filtered_df_combined_more_than_4h_before_sleep['activity_heart_rate_average'],\n",
    "    filtered_df_combined_more_than_4h_before_sleep['sleep_score'],\n",
    "    alpha=0.7,\n",
    "    color='red',\n",
    "    label=f'Training ≥4h vor Schlaf (n={red_count})'\n",
    ")\n",
    "\n",
    "plt.xlabel('Activity Heart Rate Average')\n",
    "plt.ylabel('HRV Status That Night')\n",
    "plt.ylim(60, 110)\n",
    "plt.grid(True)\n",
    "plt.legend(title='Gruppe', loc='best')\n",
    "plt.show()\n",
    "\n",
    "print(f'Number of blue points: {blue_count}')\n",
    "print(f'Number of red points: {red_count}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "52f93b1e",
   "metadata": {},
   "outputs": [],
   "source": [
    "counts = {\n",
    "    \"Kein Training\": len(df_combined[(df_combined[\"activity_calories\"].isna())]),\n",
    "    \"Training weniger als 4h vor Schlaf\": len(df_combined[(df_combined[\"activity_calories\"].notna()) & (\n",
    "        df_combined[\"bedtime_activity_ending_delta\"] < pd.Timedelta(hours=4))]),\n",
    "    \"Training mehr als 4h vor Schlaf\": len(df_combined[(df_combined[\"activity_calories\"].notna()) & (\n",
    "        df_combined[\"bedtime_activity_ending_delta\"] >= pd.Timedelta(hours=4))]),\n",
    "}\n",
    "\n",
    "labels = list(counts.keys())\n",
    "sizes = list(counts.values())\n",
    "\n",
    "plt.figure(figsize=(6, 6))\n",
    "plt.pie(\n",
    "    sizes,\n",
    "    labels=[f\"{lab} ({cnt})\" for lab, cnt in zip(labels, sizes)],\n",
    "    autopct=\"%1.1f%%\",\n",
    "    startangle=90,\n",
    ")\n",
    "plt.title(\n",
    "    \"Verteilung: Kein Training / Training <4h vor Schlaf / Training ≥4h vor Schlaf\")\n",
    "plt.axis(\"equal\")\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.13.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}