first commit

README.md

@@ -0,0 +1,106 @@
+# Remote Workforce Health Index - Synthetic Dataset
+
+Dataset ini adalah data sintetis untuk analisis kesejahteraan kerja karyawan remote/hybrid, dengan fokus pada prediksi risiko burnout.
+
+File utama di folder ini:
+- `generate_remote_workforce_synthetic_data.ipynb`: notebook pembangkit data sintetis.
+- `work_wellbeing_dataset.csv`: hasil data sintetis (30.000 baris).
+
+## Tujuan Dataset
+Dataset dirancang untuk:
+- Simulasi data HR/People Analytics tanpa menggunakan data pribadi asli.
+- Eksperimen machine learning klasifikasi `Burnout_Risk`.
+- Analisis hubungan kausal antar faktor kerja remote seperti jam kerja, intensitas meeting, dan kualitas internet.
+
+## Ringkasan Dataset
+- Nama dataset: `Remote Workforce Health Index`
+- Jumlah baris: `30000`
+- Jumlah kolom: `10`
+- Target: `Burnout_Risk` (`Low`, `Medium`, `High`)
+
+## Data Dictionary
+| Kolom | Tipe Data | Skala/Atribut | Deskripsi |
+| --- | --- | --- | --- |
+| `Employee_ID` | Integer | Nominal | ID unik karyawan (inkremental). |
+| `Work_Location` | String | Nominal | Lokasi kerja utama: `Home`, `Office`, `Coworking`. |
+| `Avg_Working_Hours` | Float | Numerik | Rata-rata jam kerja per hari (rentang dibatasi 6.0-13.0). |
+| `Meeting_Intensity` | Integer | Numerik | Rata-rata jam meeting/call per hari (0-10). |
+| `Internet_Reliability` | Categorical | Ordinal | Stabilitas koneksi: `Poor`, `Fair`, `Good`, `Excellent`. |
+| `Seniority_Level` | Categorical | Ordinal | Level jabatan: `Junior`, `Mid`, `Senior`. |
+| `Work_Life_Balance` | Integer | Ordinal | Skor keseimbangan kerja-hidup (1-5). |
+| `Daily_Mood_Note` | String | Text | Catatan suasana hati harian untuk kebutuhan NLP. |
+| `Sentiment_Score` | Float | Numerik | Skor sentimen pada rentang -1.0 sampai 1.0. |
+| `Burnout_Risk` | Categorical | Target | Label risiko burnout: `Low`, `Medium`, `High`. |
+
+## Metode Sintesis Data
+Generator menggunakan pendekatan hibrida:
+- Rule-based causal generator (aturan sebab-akibat) sebagai fondasi pola utama.
+- SDV (`GaussianCopulaSynthesizer`) untuk memperkaya variasi dan hubungan multivariat.
+- Post-processing agar data tetap konsisten dengan aturan bisnis setelah sampling.
+
+### Alur Kausal (Causal-Link)
+1. Tentukan profil dasar: `Seniority_Level` dan `Work_Location`.
+2. Tentukan `Internet_Reliability` berdasarkan `Work_Location`.
+3. Bentuk `Meeting_Intensity` dari baseline senioritas + Gaussian noise.
+4. Hitung `Avg_Working_Hours` dari baseline jam kerja + pengaruh meeting + bias lokasi.
+5. Hitung `Work_Life_Balance` dari penalti jam kerja tinggi, meeting tinggi, dan kualitas internet buruk.
+6. Tentukan `Burnout_Risk` dengan kombinasi threshold rules + probabilistic scoring.
+7. Turunkan `Sentiment_Score` dari WLB + penyesuaian burnout + kualitas internet + noise.
+8. Bentuk `Daily_Mood_Note` dari template berbasis sentimen dengan variasi dari `Faker`.
+9. Finalisasi `Employee_ID` secara inkremental.
+
+## Detail Logika yang Diimplementasikan
+- Senior cenderung memiliki intensitas meeting lebih tinggi dibanding Mid/Junior.
+- `Office` cenderung memiliki internet lebih stabil (`Good`/`Excellent`) dibanding `Home`.
+- Jam kerja meningkat seiring intensitas meeting, dengan bias tambahan pada `Home`.
+- WLB turun ketika:
+  - `Avg_Working_Hours` tinggi,
+  - `Meeting_Intensity` tinggi,
+  - `Internet_Reliability` rendah (`Poor`/`Fair`).
+- Burnout cenderung tinggi pada kombinasi WLB rendah + jam kerja panjang.
+- Sentimen berkorelasi positif dengan WLB dan negatif dengan burnout.
+
+## Dependensi
+Notebook menggunakan package berikut:
+- `pandas`
+- `numpy`
+- `faker`
+- `sdv`
+- `pyarrow` (opsional jika menyimpan parquet)
+
+Instalasi sudah disiapkan dalam notebook melalui cell:
+```python
+%pip install -q pandas numpy faker sdv pyarrow
+```
+
+## Cara Menjalankan
+1. Buka `generate_remote_workforce_synthetic_data.ipynb`.
+2. Jalankan cell dari atas ke bawah secara berurutan.
+3. Pastikan semua dependensi terinstal.
+4. Setelah selesai, file `work_wellbeing_dataset.csv` akan terbuat/terbarui.
+
+## Quality Checks yang Disediakan di Notebook
+Notebook menampilkan cek cepat untuk memvalidasi pola data:
+- Proporsi `Burnout_Risk`.
+- Rata-rata `Work_Life_Balance` per kategori burnout.
+- Rata-rata `Avg_Working_Hours` per kategori burnout.
+- Rata-rata `Sentiment_Score` per kategori burnout.
+
+Checks ini membantu memastikan data sintetis masih masuk akal secara bisnis.
+
+## Catatan Penting
+- Dataset ini sintetis, bukan data riil karyawan.
+- Tidak boleh dianggap sebagai ground truth epidemiologis/psikologis.
+- Distribusi dapat sedikit berubah jika parameter generator diubah.
+- Reproducibility didukung dengan `SEED = 42` pada notebook.
+
+## Ide Penggunaan
+- Klasifikasi burnout (`Low/Medium/High`) dengan model ML.
+- Feature importance untuk melihat faktor paling berpengaruh terhadap burnout.
+- Eksperimen NLP pada `Daily_Mood_Note` (sentiment, topic, text classification).
+- Simulasi intervensi kebijakan kerja (contoh: mengurangi meeting intensity).
+
+## Struktur Folder
+- `README.md`
+- `generate_remote_workforce_synthetic_data.ipynb`
+- `work_wellbeing_dataset.csv`

generate_remote_workforce_synthetic_data.ipynb

@@ -0,0 +1,1033 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "a1c6ccb7",
+   "metadata": {},
+   "source": [
+    "# Remote Workforce Health Index - Synthetic Data Generator\n",
+    "\n",
+    "Notebook ini membuat **30.000 baris** data sintetis dengan pendekatan:\n",
+    "1. **Causal-link generator** (aturan sebab-akibat)\n",
+    "2. **SDV Gaussian Copula** untuk memperkaya variasi distribusi\n",
+    "3. **Post-processing** agar konsisten dengan logika bisnis burnout"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "d1f080af",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Note: you may need to restart the kernel to use updated packages.\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\n",
+      "[notice] A new release of pip is available: 25.2 -> 26.0.1\n",
+      "[notice] To update, run: python.exe -m pip install --upgrade pip\n"
+     ]
+    }
+   ],
+   "source": [
+    "# Jika package belum ada, jalankan cell ini.\n",
+    "%pip install -q pandas numpy faker sdv pyarrow"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "87ef1394",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import random\n",
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "\n",
+    "from faker import Faker\n",
+    "from sdv.metadata import SingleTableMetadata\n",
+    "from sdv.single_table import GaussianCopulaSynthesizer\n",
+    "\n",
+    "SEED = 42\n",
+    "N_ROWS = 30000\n",
+    "\n",
+    "random.seed(SEED)\n",
+    "np.random.seed(SEED)\n",
+    "fake = Faker('id_ID')\n",
+    "Faker.seed(SEED)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "eadd8e2e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "SENIORITY_LEVELS = ['Junior', 'Mid', 'Senior']\n",
+    "WORK_LOCATIONS = ['Home', 'Office', 'Coworking']\n",
+    "INTERNET_LEVELS = ['Poor', 'Fair', 'Good', 'Excellent']\n",
+    "BURNOUT_LEVELS = ['Low', 'Medium', 'High']\n",
+    "\n",
+    "POSITIVE_NOTES = [\n",
+    "    'Hari produktif, pekerjaan selesai tepat waktu dan energi masih stabil.',\n",
+    "    'Fokus bagus hari ini, meeting berjalan efisien dan tugas utama tuntas.',\n",
+    "    'Ritme kerja nyaman, bisa selesai tanpa lembur berlebihan.'\n",
+    "]\n",
+    "NEUTRAL_NOTES = [\n",
+    "    'Hari cukup padat, beberapa meeting memecah fokus tapi masih terkendali.',\n",
+    "    'Progress ada, walau ritme kerja naik turun sepanjang hari.',\n",
+    "    'Tugas selesai sebagian, perlu atur ulang prioritas untuk besok.'\n",
+    "]\n",
+    "NEGATIVE_NOTES = [\n",
+    "    'Sangat lelah dengan meeting back-to-back, butuh istirahat.',\n",
+    "    'Jam kerja terasa panjang dan fokus menurun di sore hari.',\n",
+    "    'Koneksi dan tekanan deadline membuat hari ini cukup berat.'\n",
+    "]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "e2770554",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def choose_work_location(seniority: str) -> str:\n",
+    "    probs = {\n",
+    "        'Junior': [0.45, 0.35, 0.20],\n",
+    "        'Mid': [0.50, 0.30, 0.20],\n",
+    "        'Senior': [0.55, 0.25, 0.20],\n",
+    "    }\n",
+    "    return np.random.choice(WORK_LOCATIONS, p=probs[seniority])\n",
+    "\n",
+    "\n",
+    "def choose_internet_reliability(work_location: str) -> str:\n",
+    "    probs = {\n",
+    "        'Office': [0.02, 0.08, 0.35, 0.55],\n",
+    "        'Home': [0.10, 0.30, 0.40, 0.20],\n",
+    "        'Coworking': [0.04, 0.16, 0.45, 0.35],\n",
+    "    }\n",
+    "    return np.random.choice(INTERNET_LEVELS, p=probs[work_location])\n",
+    "\n",
+    "\n",
+    "def generate_meeting_intensity(seniority: str) -> int:\n",
+    "    base = {'Junior': 2.0, 'Mid': 3.8, 'Senior': 5.8}[seniority]\n",
+    "    value = np.random.normal(base, 1.15)\n",
+    "    return int(np.clip(np.round(value), 0, 10))\n",
+    "\n",
+    "\n",
+    "def generate_avg_working_hours(work_location: str, meeting_intensity: int) -> float:\n",
+    "    base = 7.6 + (0.45 * meeting_intensity)\n",
+    "    location_bias = {'Home': 1.0, 'Office': -0.2, 'Coworking': 0.2}[work_location]\n",
+    "    value = np.random.normal(base + location_bias, 0.75)\n",
+    "    return float(np.round(np.clip(value, 6.0, 13.0), 2))\n",
+    "\n",
+    "\n",
+    "def generate_wlb(avg_hours: float, meeting_intensity: int, internet: str, work_location: str) -> int:\n",
+    "    score = 5.0\n",
+    "\n",
+    "    if avg_hours > 9.0:\n",
+    "        score -= 1.0\n",
+    "    if avg_hours > 10.5:\n",
+    "        score -= 1.0\n",
+    "\n",
+    "    if meeting_intensity > 5:\n",
+    "        score -= 1.0\n",
+    "    if meeting_intensity > 7:\n",
+    "        score -= 1.0\n",
+    "\n",
+    "    if internet == 'Poor':\n",
+    "        score -= 1.0\n",
+    "    elif internet == 'Fair':\n",
+    "        score -= 0.5\n",
+    "\n",
+    "    if work_location == 'Home' and avg_hours > 10:\n",
+    "        score -= 0.5\n",
+    "\n",
+    "    score += np.random.choice([-0.5, 0.0, 0.5], p=[0.2, 0.6, 0.2])\n",
+    "    return int(np.clip(np.round(score), 1, 5))\n",
+    "\n",
+    "\n",
+    "def generate_burnout_risk(wlb: int, avg_hours: float, meeting_intensity: int, internet: str) -> str:\n",
+    "    # Rule override untuk mencerminkan skenario ekstrem yang disebutkan\n",
+    "    if wlb <= 2 and avg_hours > 10 and np.random.rand() < 0.85:\n",
+    "        return 'High'\n",
+    "    if wlb >= 4 and 7.3 <= avg_hours <= 8.8 and meeting_intensity <= 3 and np.random.rand() < 0.85:\n",
+    "        return 'Low'\n",
+    "\n",
+    "    risk_score = 0\n",
+    "    if wlb <= 2:\n",
+    "        risk_score += 2\n",
+    "    elif wlb == 3:\n",
+    "        risk_score += 1\n",
+    "\n",
+    "    if avg_hours > 10:\n",
+    "        risk_score += 2\n",
+    "    elif avg_hours > 9:\n",
+    "        risk_score += 1\n",
+    "\n",
+    "    if meeting_intensity > 6:\n",
+    "        risk_score += 1\n",
+    "\n",
+    "    if internet == 'Poor':\n",
+    "        risk_score += 1\n",
+    "\n",
+    "    if risk_score >= 5:\n",
+    "        probs = [0.02, 0.18, 0.80]\n",
+    "    elif risk_score >= 3:\n",
+    "        probs = [0.10, 0.45, 0.45]\n",
+    "    else:\n",
+    "        probs = [0.65, 0.30, 0.05]\n",
+    "\n",
+    "    return np.random.choice(BURNOUT_LEVELS, p=probs)\n",
+    "\n",
+    "\n",
+    "def generate_sentiment_score(wlb: int, burnout: str, internet: str) -> float:\n",
+    "    # Korelasi utama: WLB tinggi -> sentimen lebih positif\n",
+    "    base = -1.0 + ((wlb - 1) / 4.0) * 2.0\n",
+    "    noise = np.random.normal(0, 0.22)\n",
+    "\n",
+    "    burnout_adj = {'Low': 0.12, 'Medium': -0.05, 'High': -0.22}[burnout]\n",
+    "    internet_adj = {'Poor': -0.12, 'Fair': -0.05, 'Good': 0.03, 'Excellent': 0.08}[internet]\n",
+    "\n",
+    "    value = base + noise + burnout_adj + internet_adj\n",
+    "    return float(np.round(np.clip(value, -1.0, 1.0), 3))\n",
+    "\n",
+    "\n",
+    "def generate_daily_mood_note(sentiment_score: float) -> str:\n",
+    "    if sentiment_score > 0.5:\n",
+    "        note = random.choice(POSITIVE_NOTES)\n",
+    "    elif sentiment_score < -0.5:\n",
+    "        note = random.choice(NEGATIVE_NOTES)\n",
+    "    else:\n",
+    "        note = random.choice(NEUTRAL_NOTES)\n",
+    "\n",
+    "    # Tambahan kecil agar teks lebih beragam dan tidak terlalu templated\n",
+    "    if np.random.rand() < 0.22:\n",
+    "        note = f\"{note} Fokus tambahan: {fake.catch_phrase()}.\"\n",
+    "\n",
+    "    return note\n",
+    "\n",
+    "\n",
+    "def generate_causal_row() -> dict:\n",
+    "    seniority = np.random.choice(SENIORITY_LEVELS, p=[0.45, 0.35, 0.20])\n",
+    "    work_location = choose_work_location(seniority)\n",
+    "    internet = choose_internet_reliability(work_location)\n",
+    "\n",
+    "    meeting_intensity = generate_meeting_intensity(seniority)\n",
+    "    avg_working_hours = generate_avg_working_hours(work_location, meeting_intensity)\n",
+    "\n",
+    "    wlb = generate_wlb(avg_working_hours, meeting_intensity, internet, work_location)\n",
+    "    burnout = generate_burnout_risk(wlb, avg_working_hours, meeting_intensity, internet)\n",
+    "    sentiment = generate_sentiment_score(wlb, burnout, internet)\n",
+    "    mood_note = generate_daily_mood_note(sentiment)\n",
+    "\n",
+    "    return {\n",
+    "        'Work_Location': work_location,\n",
+    "        'Avg_Working_Hours': avg_working_hours,\n",
+    "        'Meeting_Intensity': meeting_intensity,\n",
+    "        'Internet_Reliability': internet,\n",
+    "        'Seniority_Level': seniority,\n",
+    "        'Work_Life_Balance': wlb,\n",
+    "        'Daily_Mood_Note': mood_note,\n",
+    "        'Sentiment_Score': sentiment,\n",
+    "        'Burnout_Risk': burnout,\n",
+    "    }"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "531f6da5",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Causal data shape: (30000, 9)\n"
+     ]
+    },
+    {
+     "data": {
+      "application/vnd.microsoft.datawrangler.viewer.v0+json": {
+       "columns": [
+        {
+         "name": "index",
+         "rawType": "int64",
+         "type": "integer"
+        },
+        {
+         "name": "Work_Location",
+         "rawType": "object",
+         "type": "string"
+        },
+        {
+         "name": "Avg_Working_Hours",
+         "rawType": "float64",
+         "type": "float"
+        },
+        {
+         "name": "Meeting_Intensity",
+         "rawType": "int64",
+         "type": "integer"
+        },
+        {
+         "name": "Internet_Reliability",
+         "rawType": "object",
+         "type": "string"
+        },
+        {
+         "name": "Seniority_Level",
+         "rawType": "object",
+         "type": "string"
+        },
+        {
+         "name": "Work_Life_Balance",
+         "rawType": "int64",
+         "type": "integer"
+        },
+        {
+         "name": "Daily_Mood_Note",
+         "rawType": "object",
+         "type": "string"
+        },
+        {
+         "name": "Sentiment_Score",
+         "rawType": "float64",
+         "type": "float"
+        },
+        {
+         "name": "Burnout_Risk",
+         "rawType": "object",
+         "type": "string"
+        }
+       ],
+       "ref": "3a919e97-f570-4990-b46c-3a5788630e58",
+       "rows": [
+        [
+         "0",
+         "Coworking",
+         "8.49",
+         "1",
+         "Excellent",
+         "Junior",
+         "4",
+         "Ritme kerja nyaman, bisa selesai tanpa lembur berlebihan.",
+         "0.761",
+         "Low"
+        ],
+        [
+         "1",
+         "Coworking",
+         "8.71",
+         "3",
+         "Excellent",
+         "Junior",
+         "4",
+         "Hari produktif, pekerjaan selesai tepat waktu dan energi masih stabil.",
+         "0.584",
+         "Low"
+        ],
+        [
+         "2",
+         "Home",
+         "8.94",
+         "2",
+         "Good",
+         "Junior",
+         "5",
+         "Hari produktif, pekerjaan selesai tepat waktu dan energi masih stabil.",
+         "1.0",
+         "Low"
+        ],
+        [
+         "3",
+         "Home",
+         "9.95",
+         "4",
+         "Good",
+         "Mid",
+         "4",
+         "Ritme kerja nyaman, bisa selesai tanpa lembur berlebihan.",
+         "0.688",
+         "Medium"
+        ],
+        [
+         "4",
+         "Office",
+         "8.59",
+         "2",
+         "Fair",
+         "Junior",
+         "4",
+         "Fokus bagus hari ini, meeting berjalan efisien dan tugas utama tuntas. Fokus tambahan: Sharable bifurcated algorithm.",
+         "1.0",
+         "Low"
+        ]
+       ],
+       "shape": {
+        "columns": 9,
+        "rows": 5
+       }
+      },
+      "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>Work_Location</th>\n",
+       "      <th>Avg_Working_Hours</th>\n",
+       "      <th>Meeting_Intensity</th>\n",
+       "      <th>Internet_Reliability</th>\n",
+       "      <th>Seniority_Level</th>\n",
+       "      <th>Work_Life_Balance</th>\n",
+       "      <th>Daily_Mood_Note</th>\n",
+       "      <th>Sentiment_Score</th>\n",
+       "      <th>Burnout_Risk</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>Coworking</td>\n",
+       "      <td>8.49</td>\n",
+       "      <td>1</td>\n",
+       "      <td>Excellent</td>\n",
+       "      <td>Junior</td>\n",
+       "      <td>4</td>\n",
+       "      <td>Ritme kerja nyaman, bisa selesai tanpa lembur ...</td>\n",
+       "      <td>0.761</td>\n",
+       "      <td>Low</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>Coworking</td>\n",
+       "      <td>8.71</td>\n",
+       "      <td>3</td>\n",
+       "      <td>Excellent</td>\n",
+       "      <td>Junior</td>\n",
+       "      <td>4</td>\n",
+       "      <td>Hari produktif, pekerjaan selesai tepat waktu ...</td>\n",
+       "      <td>0.584</td>\n",
+       "      <td>Low</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>Home</td>\n",
+       "      <td>8.94</td>\n",
+       "      <td>2</td>\n",
+       "      <td>Good</td>\n",
+       "      <td>Junior</td>\n",
+       "      <td>5</td>\n",
+       "      <td>Hari produktif, pekerjaan selesai tepat waktu ...</td>\n",
+       "      <td>1.000</td>\n",
+       "      <td>Low</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>Home</td>\n",
+       "      <td>9.95</td>\n",
+       "      <td>4</td>\n",
+       "      <td>Good</td>\n",
+       "      <td>Mid</td>\n",
+       "      <td>4</td>\n",
+       "      <td>Ritme kerja nyaman, bisa selesai tanpa lembur ...</td>\n",
+       "      <td>0.688</td>\n",
+       "      <td>Medium</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>Office</td>\n",
+       "      <td>8.59</td>\n",
+       "      <td>2</td>\n",
+       "      <td>Fair</td>\n",
+       "      <td>Junior</td>\n",
+       "      <td>4</td>\n",
+       "      <td>Fokus bagus hari ini, meeting berjalan efisien...</td>\n",
+       "      <td>1.000</td>\n",
+       "      <td>Low</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "  Work_Location  Avg_Working_Hours  Meeting_Intensity Internet_Reliability  \\\n",
+       "0     Coworking               8.49                  1            Excellent   \n",
+       "1     Coworking               8.71                  3            Excellent   \n",
+       "2          Home               8.94                  2                 Good   \n",
+       "3          Home               9.95                  4                 Good   \n",
+       "4        Office               8.59                  2                 Fair   \n",
+       "\n",
+       "  Seniority_Level  Work_Life_Balance  \\\n",
+       "0          Junior                  4   \n",
+       "1          Junior                  4   \n",
+       "2          Junior                  5   \n",
+       "3             Mid                  4   \n",
+       "4          Junior                  4   \n",
+       "\n",
+       "                                     Daily_Mood_Note  Sentiment_Score  \\\n",
+       "0  Ritme kerja nyaman, bisa selesai tanpa lembur ...            0.761   \n",
+       "1  Hari produktif, pekerjaan selesai tepat waktu ...            0.584   \n",
+       "2  Hari produktif, pekerjaan selesai tepat waktu ...            1.000   \n",
+       "3  Ritme kerja nyaman, bisa selesai tanpa lembur ...            0.688   \n",
+       "4  Fokus bagus hari ini, meeting berjalan efisien...            1.000   \n",
+       "\n",
+       "  Burnout_Risk  \n",
+       "0          Low  \n",
+       "1          Low  \n",
+       "2          Low  \n",
+       "3       Medium  \n",
+       "4          Low  "
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# 1) Generate data kausal mentah\n",
+    "causal_rows = [generate_causal_row() for _ in range(N_ROWS)]\n",
+    "causal_df = pd.DataFrame(causal_rows)\n",
+    "\n",
+    "print('Causal data shape:', causal_df.shape)\n",
+    "causal_df.head()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "54c02d55",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\zakyf\\AppData\\Roaming\\Python\\Python311\\site-packages\\sdv\\single_table\\base.py:168: FutureWarning:\n",
+      "\n",
+      "The 'SingleTableMetadata' is deprecated. Please use the new 'Metadata' class for synthesizers.\n",
+      "\n",
+      "C:\\Users\\zakyf\\AppData\\Roaming\\Python\\Python311\\site-packages\\sdv\\single_table\\base.py:134: UserWarning:\n",
+      "\n",
+      "We strongly recommend saving the metadata using 'save_to_json' for replicability in future SDV versions.\n",
+      "\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "SDV refinement success: (30000, 9)\n"
+     ]
+    }
+   ],
+   "source": [
+    "# 2) SDV refinement untuk meningkatkan kemiripan distribusi multi-fitur\n",
+    "#    terhadap data kausal yang sudah realistis.\n",
+    "train_cols = [\n",
+    "    'Work_Location',\n",
+    "    'Avg_Working_Hours',\n",
+    "    'Meeting_Intensity',\n",
+    "    'Internet_Reliability',\n",
+    "    'Seniority_Level',\n",
+    "    'Work_Life_Balance',\n",
+    "    'Daily_Mood_Note',\n",
+    "    'Sentiment_Score',\n",
+    "    'Burnout_Risk',\n",
+    "]\n",
+    "\n",
+    "try:\n",
+    "    metadata = SingleTableMetadata()\n",
+    "    metadata.detect_from_dataframe(causal_df[train_cols])\n",
+    "\n",
+    "    synthesizer = GaussianCopulaSynthesizer(\n",
+    "        metadata=metadata,\n",
+    "        enforce_min_max_values=True,\n",
+    "        enforce_rounding=False\n",
+    "    )\n",
+    "    synthesizer.fit(causal_df[train_cols])\n",
+    "    refined_df = synthesizer.sample(num_rows=N_ROWS)\n",
+    "    print('SDV refinement success:', refined_df.shape)\n",
+    "except Exception as e:\n",
+    "    print('SDV refinement skipped, fallback to causal data. Reason:', str(e))\n",
+    "    refined_df = causal_df.copy()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "af3fbbf3",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Final data shape: (30000, 10)\n"
+     ]
+    },
+    {
+     "data": {
+      "application/vnd.microsoft.datawrangler.viewer.v0+json": {
+       "columns": [
+        {
+         "name": "index",
+         "rawType": "int64",
+         "type": "integer"
+        },
+        {
+         "name": "Employee_ID",
+         "rawType": "int32",
+         "type": "integer"
+        },
+        {
+         "name": "Work_Location",
+         "rawType": "object",
+         "type": "string"
+        },
+        {
+         "name": "Avg_Working_Hours",
+         "rawType": "float64",
+         "type": "float"
+        },
+        {
+         "name": "Meeting_Intensity",
+         "rawType": "int32",
+         "type": "integer"
+        },
+        {
+         "name": "Internet_Reliability",
+         "rawType": "object",
+         "type": "string"
+        },
+        {
+         "name": "Seniority_Level",
+         "rawType": "object",
+         "type": "string"
+        },
+        {
+         "name": "Work_Life_Balance",
+         "rawType": "int32",
+         "type": "integer"
+        },
+        {
+         "name": "Daily_Mood_Note",
+         "rawType": "object",
+         "type": "string"
+        },
+        {
+         "name": "Sentiment_Score",
+         "rawType": "float64",
+         "type": "float"
+        },
+        {
+         "name": "Burnout_Risk",
+         "rawType": "object",
+         "type": "string"
+        }
+       ],
+       "ref": "9ff6614f-a61f-4a49-8d19-95153af3a999",
+       "rows": [
+        [
+         "0",
+         "1",
+         "Home",
+         "8.14",
+         "2",
+         "Fair",
+         "Junior",
+         "4",
+         "Fokus bagus hari ini, meeting berjalan efisien dan tugas utama tuntas.",
+         "0.685",
+         "Low"
+        ],
+        [
+         "1",
+         "2",
+         "Office",
+         "9.66",
+         "4",
+         "Good",
+         "Mid",
+         "2",
+         "Hari cukup padat, beberapa meeting memecah fokus tapi masih terkendali. Fokus tambahan: Organized leadingedge info-mediaries.",
+         "-0.463",
+         "Medium"
+        ],
+        [
+         "2",
+         "3",
+         "Home",
+         "8.21",
+         "1",
+         "Excellent",
+         "Senior",
+         "2",
+         "Progress ada, walau ritme kerja naik turun sepanjang hari.",
+         "-0.226",
+         "Low"
+        ],
+        [
+         "3",
+         "4",
+         "Home",
+         "9.71",
+         "4",
+         "Good",
+         "Mid",
+         "5",
+         "Hari produktif, pekerjaan selesai tepat waktu dan energi masih stabil.",
+         "0.897",
+         "Low"
+        ],
+        [
+         "4",
+         "5",
+         "Home",
+         "11.26",
+         "5",
+         "Excellent",
+         "Mid",
+         "4",
+         "Progress ada, walau ritme kerja naik turun sepanjang hari. Fokus tambahan: Persistent high-level Graphical User Interface.",
+         "0.365",
+         "Medium"
+        ]
+       ],
+       "shape": {
+        "columns": 10,
+        "rows": 5
+       }
+      },
+      "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>Employee_ID</th>\n",
+       "      <th>Work_Location</th>\n",
+       "      <th>Avg_Working_Hours</th>\n",
+       "      <th>Meeting_Intensity</th>\n",
+       "      <th>Internet_Reliability</th>\n",
+       "      <th>Seniority_Level</th>\n",
+       "      <th>Work_Life_Balance</th>\n",
+       "      <th>Daily_Mood_Note</th>\n",
+       "      <th>Sentiment_Score</th>\n",
+       "      <th>Burnout_Risk</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>1</td>\n",
+       "      <td>Home</td>\n",
+       "      <td>8.14</td>\n",
+       "      <td>2</td>\n",
+       "      <td>Fair</td>\n",
+       "      <td>Junior</td>\n",
+       "      <td>4</td>\n",
+       "      <td>Fokus bagus hari ini, meeting berjalan efisien...</td>\n",
+       "      <td>0.685</td>\n",
+       "      <td>Low</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>2</td>\n",
+       "      <td>Office</td>\n",
+       "      <td>9.66</td>\n",
+       "      <td>4</td>\n",
+       "      <td>Good</td>\n",
+       "      <td>Mid</td>\n",
+       "      <td>2</td>\n",
+       "      <td>Hari cukup padat, beberapa meeting memecah fok...</td>\n",
+       "      <td>-0.463</td>\n",
+       "      <td>Medium</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>3</td>\n",
+       "      <td>Home</td>\n",
+       "      <td>8.21</td>\n",
+       "      <td>1</td>\n",
+       "      <td>Excellent</td>\n",
+       "      <td>Senior</td>\n",
+       "      <td>2</td>\n",
+       "      <td>Progress ada, walau ritme kerja naik turun sep...</td>\n",
+       "      <td>-0.226</td>\n",
+       "      <td>Low</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>4</td>\n",
+       "      <td>Home</td>\n",
+       "      <td>9.71</td>\n",
+       "      <td>4</td>\n",
+       "      <td>Good</td>\n",
+       "      <td>Mid</td>\n",
+       "      <td>5</td>\n",
+       "      <td>Hari produktif, pekerjaan selesai tepat waktu ...</td>\n",
+       "      <td>0.897</td>\n",
+       "      <td>Low</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>5</td>\n",
+       "      <td>Home</td>\n",
+       "      <td>11.26</td>\n",
+       "      <td>5</td>\n",
+       "      <td>Excellent</td>\n",
+       "      <td>Mid</td>\n",
+       "      <td>4</td>\n",
+       "      <td>Progress ada, walau ritme kerja naik turun sep...</td>\n",
+       "      <td>0.365</td>\n",
+       "      <td>Medium</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   Employee_ID Work_Location  Avg_Working_Hours  Meeting_Intensity  \\\n",
+       "0            1          Home               8.14                  2   \n",
+       "1            2        Office               9.66                  4   \n",
+       "2            3          Home               8.21                  1   \n",
+       "3            4          Home               9.71                  4   \n",
+       "4            5          Home              11.26                  5   \n",
+       "\n",
+       "  Internet_Reliability Seniority_Level  Work_Life_Balance  \\\n",
+       "0                 Fair          Junior                  4   \n",
+       "1                 Good             Mid                  2   \n",
+       "2            Excellent          Senior                  2   \n",
+       "3                 Good             Mid                  5   \n",
+       "4            Excellent             Mid                  4   \n",
+       "\n",
+       "                                     Daily_Mood_Note  Sentiment_Score  \\\n",
+       "0  Fokus bagus hari ini, meeting berjalan efisien...            0.685   \n",
+       "1  Hari cukup padat, beberapa meeting memecah fok...           -0.463   \n",
+       "2  Progress ada, walau ritme kerja naik turun sep...           -0.226   \n",
+       "3  Hari produktif, pekerjaan selesai tepat waktu ...            0.897   \n",
+       "4  Progress ada, walau ritme kerja naik turun sep...            0.365   \n",
+       "\n",
+       "  Burnout_Risk  \n",
+       "0          Low  \n",
+       "1       Medium  \n",
+       "2          Low  \n",
+       "3          Low  \n",
+       "4       Medium  "
+      ]
+     },
+     "execution_count": 7,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# 3) Post-processing untuk jaga konsistensi aturan bisnis\n",
+    "final_df = refined_df.copy()\n",
+    "\n",
+    "final_df['Work_Location'] = final_df['Work_Location'].where(\n",
+    "    final_df['Work_Location'].isin(WORK_LOCATIONS),\n",
+    "    np.random.choice(WORK_LOCATIONS, size=len(final_df), p=[0.5, 0.3, 0.2])\n",
+    ")\n",
+    "\n",
+    "final_df['Seniority_Level'] = final_df['Seniority_Level'].where(\n",
+    "    final_df['Seniority_Level'].isin(SENIORITY_LEVELS),\n",
+    "    np.random.choice(SENIORITY_LEVELS, size=len(final_df), p=[0.45, 0.35, 0.2])\n",
+    ")\n",
+    "\n",
+    "final_df['Internet_Reliability'] = final_df['Internet_Reliability'].where(\n",
+    "    final_df['Internet_Reliability'].isin(INTERNET_LEVELS),\n",
+    "    np.random.choice(INTERNET_LEVELS, size=len(final_df), p=[0.07, 0.22, 0.42, 0.29])\n",
+    ")\n",
+    "\n",
+    "final_df['Meeting_Intensity'] = pd.to_numeric(final_df['Meeting_Intensity'], errors='coerce').fillna(3)\n",
+    "final_df['Meeting_Intensity'] = final_df['Meeting_Intensity'].round().clip(0, 10).astype(int)\n",
+    "\n",
+    "final_df['Avg_Working_Hours'] = pd.to_numeric(final_df['Avg_Working_Hours'], errors='coerce').fillna(8.0)\n",
+    "final_df['Avg_Working_Hours'] = final_df['Avg_Working_Hours'].clip(6.0, 13.0).round(2)\n",
+    "\n",
+    "final_df['Work_Life_Balance'] = pd.to_numeric(final_df['Work_Life_Balance'], errors='coerce').fillna(3)\n",
+    "final_df['Work_Life_Balance'] = final_df['Work_Life_Balance'].round().clip(1, 5).astype(int)\n",
+    "\n",
+    "# Recompute burnout, sentiment, mood agar hubungan antar fitur tetap masuk akal\n",
+    "final_df['Burnout_Risk'] = final_df.apply(\n",
+    "    lambda r: generate_burnout_risk(\n",
+    "        int(r['Work_Life_Balance']),\n",
+    "        float(r['Avg_Working_Hours']),\n",
+    "        int(r['Meeting_Intensity']),\n",
+    "        str(r['Internet_Reliability'])\n",
+    "    ), axis=1\n",
+    ")\n",
+    "\n",
+    "final_df['Sentiment_Score'] = final_df.apply(\n",
+    "    lambda r: generate_sentiment_score(\n",
+    "        int(r['Work_Life_Balance']),\n",
+    "        str(r['Burnout_Risk']),\n",
+    "        str(r['Internet_Reliability'])\n",
+    "    ), axis=1\n",
+    ")\n",
+    "\n",
+    "final_df['Daily_Mood_Note'] = final_df['Sentiment_Score'].apply(generate_daily_mood_note)\n",
+    "\n",
+    "# Employee_ID final, incremental\n",
+    "final_df.insert(0, 'Employee_ID', np.arange(1, len(final_df) + 1, dtype=int))\n",
+    "\n",
+    "final_df = final_df[[\n",
+    "    'Employee_ID',\n",
+    "    'Work_Location',\n",
+    "    'Avg_Working_Hours',\n",
+    "    'Meeting_Intensity',\n",
+    "    'Internet_Reliability',\n",
+    "    'Seniority_Level',\n",
+    "    'Work_Life_Balance',\n",
+    "    'Daily_Mood_Note',\n",
+    "    'Sentiment_Score',\n",
+    "    'Burnout_Risk',\n",
+    "]]\n",
+    "\n",
+    "print('Final data shape:', final_df.shape)\n",
+    "final_df.head()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "664656b1",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "\n",
+      "Burnout distribution (proporsi):\n",
+      "Burnout_Risk\n",
+      "Low       0.519\n",
+      "Medium    0.253\n",
+      "High      0.228\n",
+      "Name: proportion, dtype: float64\n",
+      "\n",
+      "WLB mean by burnout:\n",
+      "Burnout_Risk\n",
+      "High      2.46\n",
+      "Low       4.08\n",
+      "Medium    3.68\n",
+      "Name: Work_Life_Balance, dtype: float64\n",
+      "\n",
+      "Avg working hours by burnout:\n",
+      "Burnout_Risk\n",
+      "High      10.60\n",
+      "Low        9.07\n",
+      "Medium     9.73\n",
+      "Name: Avg_Working_Hours, dtype: float64\n",
+      "\n",
+      "Sentiment mean by burnout:\n",
+      "Burnout_Risk\n",
+      "High     -0.442\n",
+      "Low       0.635\n",
+      "Medium    0.298\n",
+      "Name: Sentiment_Score, dtype: float64\n"
+     ]
+    }
+   ],
+   "source": [
+    "# 4) Quick quality checks\n",
+    "print('\\nBurnout distribution (proporsi):')\n",
+    "print(final_df['Burnout_Risk'].value_counts(normalize=True).round(3))\n",
+    "\n",
+    "print('\\nWLB mean by burnout:')\n",
+    "print(final_df.groupby('Burnout_Risk')['Work_Life_Balance'].mean().round(2))\n",
+    "\n",
+    "print('\\nAvg working hours by burnout:')\n",
+    "print(final_df.groupby('Burnout_Risk')['Avg_Working_Hours'].mean().round(2))\n",
+    "\n",
+    "print('\\nSentiment mean by burnout:')\n",
+    "print(final_df.groupby('Burnout_Risk')['Sentiment_Score'].mean().round(3))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "id": "8a8b09f0",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Saved: work_wellbeing_dataset.csv\n"
+     ]
+    }
+   ],
+   "source": [
+    "# 5) Save output\n",
+    "csv_path = 'work_wellbeing_dataset.csv'\n",
+    "\n",
+    "final_df.to_csv(csv_path, index=False)\n",
+    "\n",
+    "print(f'Saved: {csv_path}')\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}