inference_data_gen.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CSV dataset generated successfully!\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import random\n",
    "\n",
    "# Define constants\n",
    "student_ids = list(range(1, 37))\n",
    "assignment_id = 1\n",
    "question_ids = list(range(1, 11))\n",
    "correctness_behaviors = [\n",
    "    'Always correct',\n",
    "    'Always incorrect',\n",
    "    'Occasionally correct',\n",
    "    'Occasionally incorrect'\n",
    "]\n",
    "confidence_behaviors = [\n",
    "    'Always confident',\n",
    "    'Always partially confident',\n",
    "    'Always not confident',\n",
    "    'Always confident but occasionally partially confident',\n",
    "    'Always confident but occasionally not confident',\n",
    "    'Always partially confident but occasionally confident',\n",
    "    'Always partially confident but occasionally not confident',\n",
    "    'Always not confident but occasionally confident',\n",
    "    'Always not confident but occasionally partially confident'\n",
    "]\n",
    "grades_thresholds = [0.7, 0.9]  # thresholds for INC, PC, C\n",
    "\n",
    "# Helper function to determine grade class\n",
    "def get_grade_class(grade):\n",
    "    if grade >= 0.9:\n",
    "        return 1\n",
    "    elif grade >= 0.7:\n",
    "        return 2\n",
    "    else:\n",
    "        return 0\n",
    "\n",
    "# Generate acquired grade based on correctness behavior\n",
    "def generate_acquired_grade(correctness_behavior):\n",
    "    if correctness_behavior == 'Always correct':\n",
    "        return random.uniform(0.9, 1.0)\n",
    "    elif correctness_behavior == 'Always incorrect':\n",
    "        return random.uniform(0.0, 0.69)\n",
    "    elif correctness_behavior == 'Occasionally correct':\n",
    "        return random.choices([random.uniform(0.9, 1.0), random.uniform(0.0, 0.69)], [0.3, 0.7])[0]\n",
    "    elif correctness_behavior == 'Occasionally incorrect':\n",
    "        return random.choices([random.uniform(0.0, 0.6), random.uniform(0.9, 1.0)], [0.3, 0.7])[0]\n",
    "\n",
    "# Generate given confidence based on confidence behavior\n",
    "def generate_given_confidence(confidence_behavior):\n",
    "    if confidence_behavior == 'Always confident':\n",
    "        return 2\n",
    "    elif confidence_behavior == 'Always partially confident':\n",
    "        return 1\n",
    "    elif confidence_behavior == 'Always not confident':\n",
    "        return 0\n",
    "    elif confidence_behavior == 'Always confident but occasionally partially confident':\n",
    "        return random.choices([2, 1], [0.7, 0.3])[0]\n",
    "    elif confidence_behavior == 'Always confident but occasionally not confident':\n",
    "        return random.choices([2, 0], [0.7, 0.3])[0]\n",
    "    elif confidence_behavior == 'Always partially confident but occasionally confident':\n",
    "        return random.choices([1, 2], [0.7, 0.3])[0]\n",
    "    elif confidence_behavior == 'Always partially confident but occasionally not confident':\n",
    "        return random.choices([1, 0], [0.7, 0.3])[0]\n",
    "    elif confidence_behavior == 'Always not confident but occasionally confident':\n",
    "        return random.choices([0, 2], [0.7, 0.3])[0]\n",
    "    elif confidence_behavior == 'Always not confident but occasionally partially confident':\n",
    "        return random.choices([0, 1], [0.7, 0.3])[0]\n",
    "\n",
    "# Create dataset\n",
    "data = []\n",
    "\n",
    "for student_id in student_ids:\n",
    "    correctness_behavior = correctness_behaviors[(student_id - 1) // 9]\n",
    "    confidence_behavior = confidence_behaviors[(student_id - 1) % 9]\n",
    "    \n",
    "    for question_id in question_ids:\n",
    "        acquired_grade = generate_acquired_grade(correctness_behavior)\n",
    "        grade_class = get_grade_class(acquired_grade)\n",
    "        given_confidence = generate_given_confidence(confidence_behavior)\n",
    "        \n",
    "        data.append({\n",
    "            'student_id': student_id,\n",
    "            'assignment_id': assignment_id,\n",
    "            'question_id': question_id,\n",
    "            'correctness_trait': correctness_behavior,\n",
    "            'confidence_trait': confidence_behavior,\n",
    "            'given_confidence': given_confidence,\n",
    "            'acquired_grade': round(acquired_grade * 100, 2),  # Convert to percentage\n",
    "            'grade_class': grade_class\n",
    "        })\n",
    "\n",
    "# Convert to DataFrame\n",
    "df = pd.DataFrame(data)\n",
    "\n",
    "# Save to CSV\n",
    "df.to_csv('./Inference_Data_Sample/student_behaviors_simulated.csv', index=False)\n",
    "\n",
    "print(\"Inference dataset generated successfully!\")\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": ".venv",
   "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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}