iotlabkit 1.0.0__py3-none-any.whl

iotlabkit/templates/program_15.py

@@ -0,0 +1,411 @@
+"""Template generator for Program 15: Decision Tree Office Room Occupancy Classification."""
+
+import os
+from typing import Dict
+
+
+def generate(target_dir: str, mode: str = "learning") -> Dict[str, str]:
+    files = {}
+
+    files["requirements.txt"] = (
+        "pandas>=1.5.0\n"
+        "scikit-learn>=1.2.0\n"
+        "matplotlib>=3.6.0\n"
+        "joblib>=1.2.0\n"
+        "streamlit>=1.25.0\n"
+    )
+
+    files[".env.example"] = (
+        "# Model Configuration\n"
+        "MODEL_PATH=model/occupancy_tree.joblib\n"
+        "MAX_TREE_DEPTH=4\n"
+    )
+
+    files[".gitignore"] = (
+        ".env\n"
+        "__pycache__/\n"
+        "*.py[cod]\n"
+        "model/*.joblib\n"
+        "output/*.png\n"
+    )
+
+    # Embed the genuine bundled dataset
+    data_path = os.path.join(os.path.dirname(os.path.dirname(__file__)), "data", "occupancy", "datatraining.txt")
+    if os.path.exists(data_path):
+        with open(data_path, "r", encoding="utf-8") as f:
+            files["data/datatraining.txt"] = f.read()
+    else:
+        # Fallback header if not yet written
+        files["data/datatraining.txt"] = (
+            '"date","Temperature","Humidity","Light","CO2","HumidityRatio","Occupancy"\n'
+            '"1","2015-02-04 17:51:00",23.18,27.272,426,721.25,0.00479298817650529,1\n'
+            '"17","2015-02-04 18:07:00",23,27.2,0,681.5,0.00472807794966877,0\n'
+        )
+
+    # data/download_dataset.py
+    files["data/download_dataset.py"] = (
+        '"""Official Dataset Downloader for UCI Occupancy Detection (#357).\n\n'
+        "Downloads the complete dataset (8,143 training records) directly from the\n"
+        "author's official repository (Luis Candanedo, CC BY 4.0 license).\n"
+        '"""\n\n'
+        "import os\n"
+        "import sys\n"
+        "import requests\n\n"
+        "OFFICIAL_URL = 'https://raw.githubusercontent.com/LuisM78/Occupancy-detection-data/master/datatraining.txt'\n"
+        "TARGET_FILE = os.path.join(os.path.dirname(__file__), 'datatraining.txt')\n\n"
+        "def main():\n"
+        "    print(f\"Downloading complete official UCI Occupancy dataset from:\\n  {OFFICIAL_URL}...\")\n"
+        "    try:\n"
+        "        resp = requests.get(OFFICIAL_URL, timeout=15)\n"
+        "        if resp.status_code == 200:\n"
+        "            with open(TARGET_FILE, 'w', encoding='utf-8') as f:\n"
+        "                f.write(resp.text)\n"
+        "            lines = resp.text.strip().split('\\n')\n"
+        "            print(f\"✓ Downloaded {len(lines) - 1} authentic records to '{TARGET_FILE}'.\")\n"
+        "        else:\n"
+        "            print(f\"✗ Server returned HTTP {resp.status_code}\")\n"
+        "    except Exception as e:\n"
+        "        print(f\"✗ Download failed: {e}\")\n"
+        "        print(\"\\nManual Download Instructions:\")\n"
+        "        print(\"1. Open https://archive.ics.uci.edu/dataset/357/occupancy+detection\")\n"
+        "        print(\"2. Download 'occupancy+detection.zip' and extract 'datatraining.txt' into the 'data/' folder.\")\n\n"
+        "if __name__ == '__main__':\n"
+        "    main()\n"
+    )
+
+    # train.py
+    if mode == "learning":
+        train_doc = (
+            '"""Decision Tree Room Occupancy Classifier Training - Program 15.\n\n'
+            "Loads the authentic UCI Occupancy Detection dataset, inspects features,\n"
+            "trains a DecisionTreeClassifier, evaluates metrics (Accuracy, Precision, Recall, F1),\n"
+            "generates confusion matrix and decision tree structure plots, and saves the trained model.\n"
+            '"""\n'
+        )
+    else:
+        train_doc = '"""Decision Tree Classifier - Program 15 (Exam Mode)."""\n'
+
+    files["train.py"] = train_doc + (
+        "import os\n"
+        "import sys\n"
+        "import joblib\n"
+        "try:\n"
+        "    import matplotlib.pyplot as plt\n"
+        "    HAS_MATPLOTLIB = True\n"
+        "except ImportError:\n"
+        "    HAS_MATPLOTLIB = False\n"
+        "import pandas as pd\n"
+        "from sklearn.metrics import accuracy_score, classification_report, confusion_matrix\n"
+        "from sklearn.model_selection import train_test_split\n"
+        "from sklearn.tree import DecisionTreeClassifier\n"
+        "try:\n"
+        "    from sklearn.tree import plot_tree\n"
+        "except ImportError:\n"
+        "    plot_tree = None\n\n"
+        "DATA_PATH = os.path.join(os.path.dirname(__file__), 'data', 'datatraining.txt')\n"
+        "MODEL_DIR = os.path.join(os.path.dirname(__file__), 'model')\n"
+        "OUTPUT_DIR = os.path.join(os.path.dirname(__file__), 'output')\n\n"
+        "def main():\n"
+        "    if not os.path.exists(DATA_PATH):\n"
+        "        print(f\"✗ Error: Dataset file '{DATA_PATH}' not found.\")\n"
+        "        print(\"  Run `python data/download_dataset.py` to obtain official UCI dataset.\")\n"
+        "        sys.exit(1)\n\n"
+        "    print('=' * 70)\n"
+        "    print('🧠 Program 15: Decision Tree Room Occupancy Classification')\n"
+        "    print('=' * 70)\n"
+        "    print(f\"Loading official UCI Occupancy dataset from: {DATA_PATH}...\")\n"
+        "    df = pd.read_csv(DATA_PATH)\n\n"
+        "    # Data Inspection\n"
+        "    print(f\"✓ Loaded {len(df)} authentic sensor records.\")\n"
+        "    print(f\"Columns present: {list(df.columns)}\")\n\n"
+        "    # Feature selection (environmental sensor measurements)\n"
+        "    feature_cols = ['Temperature', 'Humidity', 'Light', 'CO2', 'HumidityRatio']\n"
+        "    target_col = 'Occupancy'\n\n"
+        "    # Validate required columns exist\n"
+        "    for col in feature_cols + [target_col]:\n"
+        "        if col not in df.columns:\n"
+        "            raise ValueError(f\"Expected column '{col}' missing from dataset!\")\n\n"
+        "    # Clean missing values\n"
+        "    missing_count = df[feature_cols + [target_col]].isna().sum().sum()\n"
+        "    if missing_count > 0:\n"
+        "        print(f\"Handling {missing_count} missing values via median imputation...\")\n"
+        "        df = df.dropna(subset=[target_col])\n"
+        "        df[feature_cols] = df[feature_cols].fillna(df[feature_cols].median())\n"
+        "    else:\n"
+        "        print(\"✓ Zero missing values detected in dataset.\")\n\n"
+        "    X = df[feature_cols]\n"
+        "    y = df[target_col].astype(int)\n\n"
+        "    print(f\"Class Distribution: Unoccupied (0) = {(y == 0).sum()}, Occupied (1) = {(y == 1).sum()}\")\n\n"
+        "    # Train-Test Split (80% Train, 20% Test, Stratified)\n"
+        "    X_train, X_test, y_train, y_test = train_test_split(\n"
+        "        X, y, test_size=0.20, random_state=42, stratify=y\n"
+        "    )\n"
+        "    print(f\"Training set: {len(X_train)} samples | Test set: {len(X_test)} samples\")\n\n"
+        "    # Initialize and train Decision Tree\n"
+        "    # max_depth=4 prevents overfitting and ensures human-interpretable rules\n"
+        "    clf = DecisionTreeClassifier(criterion='gini', max_depth=4, random_state=42)\n"
+        "    clf.fit(X_train, y_train)\n\n"
+        "    # Model Evaluation\n"
+        "    y_pred = clf.predict(X_test)\n"
+        "    acc = accuracy_score(y_test, y_pred)\n"
+        "    cm = confusion_matrix(y_test, y_pred)\n"
+        "    report = classification_report(y_test, y_pred, target_names=['Unoccupied (0)', 'Occupied (1)'])\n\n"
+        "    print('\\n' + '=' * 70)\n"
+        "    print('📊 Model Evaluation Results (Test Set):')\n"
+        "    print('=' * 70)\n"
+        "    print(f\"Accuracy Score: {acc * 100:.2f}%\")\n"
+        "    print('\\nClassification Report:')\n"
+        "    print(report)\n"
+        "    print('Confusion Matrix:')\n"
+        "    print(f\"  True Negative  (TN): {cm[0][0]:>4} | False Positive (FP): {cm[0][1]:>4}\")\n"
+        "    print(f\"  False Negative (FN): {cm[1][0]:>4} | True Positive  (TP): {cm[1][1]:>4}\")\n"
+        "    print('=' * 70 + '\\n')\n\n"
+        "    # Feature Importances\n"
+        "    importances = clf.feature_importances_\n"
+        "    print('Feature Importances (Gini Impurity Reduction):')\n"
+        "    for feat, imp in sorted(zip(feature_cols, importances), key=lambda x: x[1], reverse=True):\n"
+        "        print(f\"  - {feat:<16}: {imp * 100:>5.2f}%\")\n\n"
+        "    # Save Model\n"
+        "    os.makedirs(MODEL_DIR, exist_ok=True)\n"
+        "    os.makedirs(OUTPUT_DIR, exist_ok=True)\n"
+        "    model_path = os.path.join(MODEL_DIR, 'occupancy_tree.joblib')\n"
+        "    joblib.dump(clf, model_path)\n"
+        "    print(f\"\\n✓ Saved trained model to: {model_path}\")\n\n"
+        "    if HAS_MATPLOTLIB:\n"
+        "        # Visualization 1: Confusion Matrix Plot\n"
+        "        plt.figure(figsize=(5, 4))\n"
+        "        plt.imshow(cm, interpolation='nearest', cmap=plt.cm.Blues)\n"
+        "        plt.title(f'Decision Tree Confusion Matrix (Acc: {acc*100:.1f}%)')\n"
+        "        plt.colorbar()\n"
+        "        tick_marks = [0, 1]\n"
+        "        plt.xticks(tick_marks, ['Unoccupied', 'Occupied'])\n"
+        "        plt.yticks(tick_marks, ['Unoccupied', 'Occupied'])\n"
+        "        thresh = cm.max() / 2.\n"
+        "        for i in range(2):\n"
+        "            for j in range(2):\n"
+        "                plt.text(j, i, format(cm[i, j], 'd'), ha='center', va='center',\n"
+        "                         color='white' if cm[i, j] > thresh else 'black')\n"
+        "        plt.ylabel('True Label')\n"
+        "        plt.xlabel('Predicted Label')\n"
+        "        p1 = os.path.join(OUTPUT_DIR, 'confusion_matrix.png')\n"
+        "        plt.savefig(p1, dpi=150, bbox_inches='tight')\n"
+        "        plt.close()\n"
+        "        print(f\"✓ Saved Confusion Matrix Plot: {p1}\")\n\n"
+        "        # Visualization 2: Feature Importances Bar Chart\n"
+        "        plt.figure(figsize=(7, 4))\n"
+        "        plt.barh(feature_cols, importances, color='teal')\n"
+        "        plt.title('Sensor Feature Importances for Room Occupancy')\n"
+        "        plt.xlabel('Relative Importance')\n"
+        "        plt.grid(axis='x', linestyle='--', alpha=0.6)\n"
+        "        p2 = os.path.join(OUTPUT_DIR, 'feature_importances.png')\n"
+        "        plt.savefig(p2, dpi=150, bbox_inches='tight')\n"
+        "        plt.close()\n"
+        "        print(f\"✓ Saved Feature Importance Chart: {p2}\")\n\n"
+        "        # Visualization 3: Decision Tree Structure Graph\n"
+        "        if plot_tree is not None:\n"
+        "            plt.figure(figsize=(16, 8))\n"
+        "            plot_tree(clf, feature_names=feature_cols, class_names=['Unoccupied', 'Occupied'],\n"
+        "                      filled=True, rounded=True, fontsize=9)\n"
+        "            plt.title('Trained Decision Tree Logic Architecture (Max Depth = 4)')\n"
+        "            p3 = os.path.join(OUTPUT_DIR, 'decision_tree_structure.png')\n"
+        "            plt.savefig(p3, dpi=150, bbox_inches='tight')\n"
+        "            plt.close()\n"
+        "            print(f\"✓ Saved Decision Tree Architecture: {p3}\")\n"
+        "    else:\n"
+        "        print(\"\\nℹ️  Notice: matplotlib not installed. Evaluation metrics and confusion matrix are printed above. Install matplotlib to save PNG charts.\")\n\n"
+        "    print('\\nReady for prediction! Run: python predict.py --temp 23.2 --humidity 27.2 --light 426 --co2 721')\n\n"
+        "if __name__ == '__main__':\n"
+        "    main()\n"
+    )
+
+    # predict.py
+    if mode == "learning":
+        pred_doc = (
+            '"""Real-Time Sensor Reading Prediction Script.\n\n'
+            "Loads the trained Decision Tree model and evaluates whether a room is occupied\n"
+            "given new live or user-supplied environmental sensor readings.\n"
+            '"""\n'
+        )
+    else:
+        pred_doc = '"""Occupancy Inference - Program 15 (Exam Mode)."""\n'
+
+    files["predict.py"] = pred_doc + (
+        "import argparse\n"
+        "import os\n"
+        "import sys\n"
+        "import joblib\n"
+        "import pandas as pd\n\n"
+        "MODEL_PATH = os.path.join(os.path.dirname(__file__), 'model', 'occupancy_tree.joblib')\n\n"
+        "def predict_occupancy(temp: float, humidity: float, light: float, co2: float, hratio: float = None) -> tuple:\n"
+        "    if not os.path.exists(MODEL_PATH):\n"
+        "        raise FileNotFoundError(f\"Model '{MODEL_PATH}' not found. Run `python train.py` first.\")\n\n"
+        "    clf = joblib.load(MODEL_PATH)\n\n"
+        "    # If humidity ratio is not provided, compute it using Candanedo's approximation:\n"
+        "    if hratio is None:\n"
+        "        # Simplified saturation vapor pressure & humidity ratio approximation\n"
+        "        p_sat = 6.112 * (10 ** (7.5 * temp / (237.7 + temp)))\n"
+        "        p_v = (humidity / 100.0) * p_sat\n"
+        "        hratio = round(0.62198 * p_v / (1013.25 - p_v), 6)\n\n"
+        "    features = pd.DataFrame([{\n"
+        "        'Temperature': temp,\n"
+        "        'Humidity': humidity,\n"
+        "        'Light': light,\n"
+        "        'CO2': co2,\n"
+        "        'HumidityRatio': hratio\n"
+        "    }])\n\n"
+        "    pred_class = int(clf.predict(features)[0])\n"
+        "    probs = clf.predict_proba(features)[0]\n"
+        "    return pred_class, probs, hratio\n\n"
+        "def main():\n"
+        "    parser = argparse.ArgumentParser(description='Classify room occupancy from sensor values')\n"
+        "    parser.add_argument('--temp', type=float, default=23.18, help='Temperature in °C (default: 23.18)')\n"
+        "    parser.add_argument('--humidity', type=float, default=27.27, help='Humidity in % (default: 27.27)')\n"
+        "    parser.add_argument('--light', type=float, default=426.0, help='Light in Lux (default: 426.0)')\n"
+        "    parser.add_argument('--co2', type=float, default=721.25, help='CO2 in ppm (default: 721.25)')\n"
+        "    parser.add_argument('--hratio', type=float, default=None, help='Humidity Ratio (optional)')\n"
+        "    args = parser.parse_args()\n\n"
+        "    try:\n"
+        "        pred, probs, hr = predict_occupancy(args.temp, args.humidity, args.light, args.co2, args.hratio)\n"
+        "        status_str = 'OCCUPIED (1)' if pred == 1 else 'NOT OCCUPIED (0)'\n"
+        "        confidence = probs[pred] * 100\n\n"
+        "        print('=' * 60)\n"
+        "        print('   Decision Tree Inference for Room Occupancy')\n"
+        "        print('=' * 60)\n"
+        "        print(f\"Input Readings:\")\n"
+        "        print(f\"  - Temperature    : {args.temp} °C\")\n"
+        "        print(f\"  - Humidity       : {args.humidity} %\")\n"
+        "        print(f\"  - Light          : {args.light} Lux\")\n"
+        "        print(f\"  - CO2            : {args.co2} ppm\")\n"
+        "        print(f\"  - Humidity Ratio : {hr}\")\n"
+        "        print('-' * 60)\n"
+        "        print(f\"Predicted State    : {status_str}\")\n"
+        "        print(f\"Model Confidence   : {confidence:.2f}%\")\n"
+        "        print(f\"Probabilities      : [Unoccupied: {probs[0]*100:.1f}%, Occupied: {probs[1]*100:.1f}%]\")\n"
+        "        print('=' * 60)\n"
+        "    except Exception as err:\n"
+        "        print(f\"✗ Prediction error: {err}\")\n"
+        "        sys.exit(1)\n\n"
+        "if __name__ == '__main__':\n"
+        "    main()\n"
+    )
+
+    # dashboard.py
+    files["dashboard.py"] = (
+        "import os\n"
+        "import streamlit as st\n"
+        "from predict import predict_occupancy\n\n"
+        "st.set_page_config(page_title='Program 15: Occupancy Predictor', layout='wide')\n"
+        "st.title('🏢 Decision Tree Room Occupancy Classifier')\n"
+        "st.caption('MSc/MCA IoT Practical — Program 15 (UCI Occupancy Detection #357)')\n\n"
+        "col1, col2 = st.columns([1, 1])\n\n"
+        "with col1:\n"
+        "    st.subheader('🎛️ Input Sensor Readings')\n"
+        "    temp = st.slider('Temperature (°C)', min_value=15.0, max_value=35.0, value=23.18, step=0.1)\n"
+        "    humidity = st.slider('Relative Humidity (%)', min_value=10.0, max_value=90.0, value=27.27, step=0.1)\n"
+        "    light = st.slider('Ambient Light (Lux)', min_value=0.0, max_value=1000.0, value=426.0, step=5.0)\n"
+        "    co2 = st.slider('Carbon Dioxide (ppm)', min_value=350.0, max_value=2500.0, value=721.25, step=10.0)\n\n"
+        "with col2:\n"
+        "    st.subheader('🎯 Real-Time Decision Tree Prediction')\n"
+        "    try:\n"
+        "        pred, probs, hr = predict_occupancy(temp, humidity, light, co2)\n"
+        "        if pred == 1:\n"
+        "            st.success(f'### 🟢 ROOM OCCUPIED (Class 1)\\n**Confidence**: {probs[1]*100:.1f}%')\n"
+        "        else:\n"
+        "            st.info(f'### ⚪ ROOM UNOCCUPIED (Class 0)\\n**Confidence**: {probs[0]*100:.1f}%')\n\n"
+        "        st.write(f'**Derived Humidity Ratio**: `{hr}` kg/kg')\n"
+        "        st.progress(float(probs[1]))\n"
+        "        st.caption(f'Occupancy Probability: {probs[1]*100:.1f}% | Unoccupied Probability: {probs[0]*100:.1f}%')\n"
+        "    except FileNotFoundError:\n"
+        "        st.warning('Trained model not found. Run `python train.py` in your terminal to train the Decision Tree.')\n\n"
+        "st.divider()\n"
+        "st.subheader('🖼️ Visual Model Artifacts')\n"
+        "c_img1, c_img2 = st.columns(2)\n"
+        "output_dir = os.path.join(os.path.dirname(__file__), 'output')\n"
+        "cm_path = os.path.join(output_dir, 'confusion_matrix.png')\n"
+        "fi_path = os.path.join(output_dir, 'feature_importances.png')\n\n"
+        "if os.path.exists(cm_path):\n"
+        "    c_img1.image(cm_path, caption='Confusion Matrix')\n"
+        "if os.path.exists(fi_path):\n"
+        "    c_img2.image(fi_path, caption='Feature Importances')\n"
+    )
+
+    # README.md
+    if mode == "learning":
+        readme_content = (
+            "# Program 15 — Decision Tree Office Room Occupancy Classification\n\n"
+            "## 1. Overview\n"
+            "This practical builds a supervised machine-learning classifier to predict office room occupancy:\n"
+            "- Uses the genuine **UCI Occupancy Detection Dataset (#357)** created by Luis Candanedo (CC BY 4.0).\n"
+            "- Features: `Temperature` (°C), `Humidity` (%), `Light` (Lux), `CO2` (ppm), and derived `HumidityRatio`.\n"
+            "- Target: Binary `Occupancy` (0 for unoccupied, 1 for occupied).\n"
+            "- Evaluates classification performance via **Accuracy**, **Precision**, **Recall**, **F1-score**, and **Confusion Matrix**.\n"
+            "- Visualizes decision boundaries, feature importances, and tree rules.\n"
+            "- Exports trained model to `model/occupancy_tree.joblib` and provides CLI + Streamlit inference.\n\n"
+            "## 2. Decision Tree Algorithm Primer\n"
+            "The Decision Tree splits nodes by maximizing information gain or minimizing the **Gini Impurity**:\n"
+            "$$Gini(D) = 1 - \\sum_{i=1}^{C} p_i^2$$\n"
+            "Light intensity (Lux) is predominantly the top splitting decision node because indoor office lighting is toggled when persons enter or leave a room.\n\n"
+            "## 3. Step-by-Step Execution Guide\n\n"
+            "### Step A: Setup Environment\n"
+            "```bash\n"
+            "pip install -r requirements.txt\n"
+            "```\n\n"
+            "### Step B: (Optional) Download Full Official Dataset\n"
+            "A balanced verified subset of authentic records is already pre-bundled in `data/datatraining.txt`. To fetch the entire 8,143 records from the official repository:\n"
+            "```bash\n"
+            "python data/download_dataset.py\n"
+            "```\n\n"
+            "### Step C: Train and Evaluate Decision Tree\n"
+            "```bash\n"
+            "python train.py\n"
+            "```\n"
+            "Inspect the printed Accuracy score, Confusion Matrix, and the generated charts in `output/`:\n"
+            "- `confusion_matrix.png`\n"
+            "- `feature_importances.png`\n"
+            "- `decision_tree_structure.png`\n\n"
+            "### Step D: Predict on New Sensor Readings\n"
+            "```bash\n"
+            "# Office room during work hours with lights on:\n"
+            "python predict.py --temp 23.18 --humidity 27.27 --light 426 --co2 721\n\n"
+            "# Empty office room at night:\n"
+            "python predict.py --temp 21.0 --humidity 28.0 --light 0 --co2 420\n"
+            "```\n\n"
+            "### Step E: Interactive Streamlit Dashboard\n"
+            "```bash\n"
+            "streamlit run dashboard.py\n"
+            "```\n\n"
+            "## 4. Viva Questions and Answers\n"
+            "**Q1: What are True Positives, False Positives, True Negatives, and False Negatives?**  \n"
+            "A: \n"
+            "- TP: Room was occupied, model predicted occupied.\n"
+            "- TN: Room was empty, model predicted empty.\n"
+            "- FP (Type I Error): Room was empty, model predicted occupied.\n"
+            "- FN (Type II Error): Room was occupied, model predicted empty.\n\n"
+            "**Q2: Why do we constrain `max_depth` in Decision Trees?**  \n"
+            "A: Unconstrained decision trees will continue splitting until every leaf is pure, memorizing training noise and severely overfitting. Limiting depth acts as regularization.\n"
+        )
+    else:
+        readme_content = (
+            "# Program 15 — Decision Tree Occupancy Classification (Exam Mode)\n\n"
+            "## Execution\n"
+            "1. Train Decision Tree & save model:\n"
+            "   ```bash\n"
+            "   python train.py\n"
+            "   ```\n"
+            "2. Run prediction on new reading:\n"
+            "   ```bash\n"
+            "   python predict.py --temp 23.2 --humidity 27.2 --light 426 --co2 721\n"
+            "   ```\n"
+            "3. Interactive Dashboard:\n"
+            "   ```bash\n"
+            "   streamlit run dashboard.py\n"
+            "   ```\n\n"
+            "## Viva Q&A\n"
+            "- **Q: Which sensor feature has the highest predictive importance?**  \n"
+            "  A: Light (Lux), followed by CO2 concentration.\n"
+            "- **Q: What metric evaluates class-imbalanced classification?**  \n"
+            "  A: F1-score (harmonic mean of Precision and Recall), as simple Accuracy can be misleading under severe class imbalance.\n"
+        )
+
+    files["README.md"] = readme_content
+    return files