{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 🎯 Confidence Score Calibration Demo\n",
    "\n",
    "<div style=\"background-color: rgba(200, 200, 200, 0.1); padding: 20px; border-radius: 8px; margin-bottom: 20px; border: 1px solid rgba(127, 127, 127, 0.2); max-width: 97.5%; overflow-wrap: break-word;\">\n",
    "  <p style=\"font-size: 16px; line-height: 1.6\">\n",
    "     Confidence scores from uncertainty quantification methods may not be well-calibrated probabilities. This demo demonstrates how to transform raw confidence scores into calibrated probabilities that better reflect the true likelihood of correctness using the <code>ScoreCalibrator</code> class.\n",
    "  </p>\n",
    "</div>\n",
    "\n",
    "## 📊 What You'll Do in This Demo\n",
    "\n",
    "<div style=\"display: flex; margin-bottom: 15px; align-items: center\">\n",
    "  <div style=\"background-color: #34a853; color: white; border-radius: 50%; width: 30px; height: 30px; display: flex; justify-content: center; align-items: center; margin-right: 15px; flex-shrink: 0\"><strong>1</strong></div>\n",
    "  <div>\n",
    "    <p style=\"margin: 0; font-weight: bold\"><a href=#section1>Set up LLM and prompts.</a></p>\n",
    "    <p style=\"margin: 0; color: rgba(95, 99, 104, 0.8)\">Set up LLM instance and load example data prompts.</p>\n",
    "  </div>\n",
    "</div>\n",
    "\n",
    "<div style=\"display: flex; margin-bottom: 15px; align-items: center\">\n",
    "  <div style=\"background-color: #34a853; color: white; border-radius: 50%; width: 30px; height: 30px; display: flex; justify-content: center; align-items: center; margin-right: 15px; flex-shrink: 0\"><strong>2</strong></div>\n",
    "  <div>\n",
    "    <p style=\"margin: 0; font-weight: bold\"><a href=#section2>Generate LLM Responses and Confidence Scores</a></p>\n",
    "    <p style=\"margin: 0; color: rgba(95, 99, 104, 0.8)\">Generate and score LLM responses to the example questions using the <code>WhiteBoxUQ()</code> class.</p>\n",
    "  </div>\n",
    "</div>\n",
    "\n",
    "<div style=\"display: flex; margin-bottom: 25px; align-items: center\">\n",
    "  <div style=\"background-color: #34a853; color: white; border-radius: 50%; width: 30px; height: 30px; display: flex; justify-content: center; align-items: center; margin-right: 15px; flex-shrink: 0\"><strong>3</strong></div>\n",
    "  <div>\n",
    "    <p style=\"margin: 0; font-weight: bold\"><a href=#section3>Fit Calibrators and Evaluate on Holdout Set</a></p>\n",
    "    <p style=\"margin: 0; color: rgba(95, 99, 104, 0.8)\">Train confidence score calibrators and evaluate on holdout set of prompts.</p>\n",
    "  </div>\n",
    "</div>\n",
    "\n",
    "\n",
    "## ⚖️ Calibration Methods\n",
    "\n",
    "<div style=\"display: flex; gap: 20px\">\n",
    "  <div style=\"flex: 1; background-color: rgba(0, 100, 200, 0.1); padding: 15px; border-radius: 8px; border: 1px solid rgba(0, 100, 200, 0.2)\">\n",
    "    <h3 style=\"color: #1e40af; margin-top: 0\">Platt Scaling</h3>\n",
    "    <ul style=\"margin-bottom: 0\">\n",
    "      <li><strong>Method:</strong> Logistic regression</li>\n",
    "      <li><strong>Parametric:</strong> Assumes sigmoid-shaped calibration function</li>\n",
    "      <li><strong>Best for:</strong> Small datasets, well-behaved score distributions</li>\n",
    "    </ul>\n",
    "  </div>\n",
    "  \n",
    "  <div style=\"flex: 1; background-color: rgba(0, 200, 0, 0.1); padding: 15px; border-radius: 8px; border: 1px solid rgba(0, 200, 0, 0.2)\">\n",
    "    <h3 style=\"color: #2e8b57; margin-top: 0\">Isotonic Regression</h3>\n",
    "    <ul style=\"margin-bottom: 0\">\n",
    "      <li><strong>Method:</strong> Non-parametric, monotonic</li>\n",
    "      <li><strong>Flexible:</strong> Can handle any monotonic calibration curve</li>\n",
    "      <li><strong>Best for:</strong> Larger datasets, complex score distributions</li>\n",
    "    </ul>\n",
    "  </div>\n",
    "</div>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "from uqlm import WhiteBoxUQ\n",
    "from uqlm.calibration import ScoreCalibrator, evaluate_calibration\n",
    "from uqlm.utils import load_example_dataset, LLMGrader"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Set up LLM and Prompts\n",
    "\n",
    "For this demo, we'll sample 1500 prompts from the [NQ-Open benchmark](https://aclanthology.org/Q19-1026/). The first 1000 prompts will be used to train the calibrators and remaining 500 prompts will be used as a test dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading dataset - nq_open...\n",
      "Processing dataset...\n",
      "Dataset ready!\n"
     ]
    }
   ],
   "source": [
    "n_train, n_test = 1000, 500\n",
    "n_prompts = n_train + n_test\n",
    "\n",
    "# Load example dataset for prompts/answers (optional, for context)\n",
    "nq_open = load_example_dataset(\"nq_open\", n=n_prompts)\n",
    "\n",
    "# Define prompts\n",
    "QA_INSTRUCTION = \"You will be given a question. Return only the answer as concisely as possible without providing an explanation.\\n\"\n",
    "prompts = [QA_INSTRUCTION + prompt for prompt in nq_open.question]\n",
    "train_prompts = prompts[:n_train]\n",
    "test_prompts = prompts[-n_test:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this example, we use `ChatVertexAI` to instantiate our LLM, but any [LangChain Chat Model](https://js.langchain.com/docs/integrations/chat/) may be used. Be sure to **replace with your LLM of choice.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# import sys\n",
    "# !{sys.executable} -m pip install langchain-google-vertexai\n",
    "from langchain_google_vertexai import ChatVertexAI\n",
    "\n",
    "llm = ChatVertexAI(model=\"gemini-2.5-flash\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='section2'></a>\n",
    "## 2. Compute Confidence Scores\n",
    "We generate model responses and associated confidence scores by leveraging the `WhiteBoxUQ` class. This class generates responses to prompts, while also estimating a confidence score for each response using token probabilities."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1c159f044ece4c7a915733d3d8b382ce",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "wbuq = WhiteBoxUQ(llm=llm, scorers=[\"normalized_probability\"])\n",
    "uq_result = await wbuq.generate_and_score(prompts=train_prompts)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To obtain the labels for calibration, we 'grade' the responses against an answer key. Here, we use UQLM's out-of-the-box LLM Grader, which can be used with [LangChain Chat Model](https://js.langchain.com/docs/integrations/chat/), but you may replace this with a grading method of your choice. Some notable alternatives are [Vectara HHEM](https://huggingface.co/vectara/hallucination_evaluation_model) and [AlignScore](https://github.com/yuh-zha/AlignScore). **If you are using your own prompts/questions, be sure to update the grading method accordingly**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "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>prompt</th>\n",
       "      <th>response</th>\n",
       "      <th>logprob</th>\n",
       "      <th>normalized_probability</th>\n",
       "      <th>response_correct</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>December 14, 1972</td>\n",
       "      <td>[{'token': 'December', 'logprob': -0.044779419...</td>\n",
       "      <td>0.980881</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>Bobby Scott and Bob Russell</td>\n",
       "      <td>[{'token': 'Bobby', 'logprob': -0.065702043473...</td>\n",
       "      <td>0.979054</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>1</td>\n",
       "      <td>[{'token': '1', 'logprob': -0.0108721693977713...</td>\n",
       "      <td>0.989187</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>Super Bowl LII</td>\n",
       "      <td>[{'token': 'Super', 'logprob': -1.586603045463...</td>\n",
       "      <td>0.672557</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>South Carolina</td>\n",
       "      <td>[{'token': 'South', 'logprob': -1.502075701864...</td>\n",
       "      <td>0.999992</td>\n",
       "      <td>True</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              prompt  \\\n",
       "0  You will be given a question. Return only the ...   \n",
       "1  You will be given a question. Return only the ...   \n",
       "2  You will be given a question. Return only the ...   \n",
       "3  You will be given a question. Return only the ...   \n",
       "4  You will be given a question. Return only the ...   \n",
       "\n",
       "                      response  \\\n",
       "0            December 14, 1972   \n",
       "1  Bobby Scott and Bob Russell   \n",
       "2                            1   \n",
       "3               Super Bowl LII   \n",
       "4               South Carolina   \n",
       "\n",
       "                                             logprob  normalized_probability  \\\n",
       "0  [{'token': 'December', 'logprob': -0.044779419...                0.980881   \n",
       "1  [{'token': 'Bobby', 'logprob': -0.065702043473...                0.979054   \n",
       "2  [{'token': '1', 'logprob': -0.0108721693977713...                0.989187   \n",
       "3  [{'token': 'Super', 'logprob': -1.586603045463...                0.672557   \n",
       "4  [{'token': 'South', 'logprob': -1.502075701864...                0.999992   \n",
       "\n",
       "   response_correct  \n",
       "0              True  \n",
       "1              True  \n",
       "2              True  \n",
       "3             False  \n",
       "4              True  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# set up the LLM grader to grade LLM responses against the ground truth answer key (we need these grades for calibration)\n",
    "gemini_flash_lite = ChatVertexAI(model=\"gemini-2.5-flash-lite\")\n",
    "grader = LLMGrader(llm=gemini_flash_lite)\n",
    "\n",
    "# Convert to dataframe and grade responses against correct answers\n",
    "result_df = uq_result.to_df()\n",
    "result_df[\"response_correct\"] = await grader.grade_responses(prompts=nq_open[\"question\"].to_list()[:n_train], responses=result_df[\"response\"].to_list(), answers=nq_open[\"answer\"].to_list()[:n_train])\n",
    "result_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<a id='section3'></a>\n",
    "## 3. Score Calibration\n",
    "Confidence scores from uncertainty quantification methods may not be well-calibrated probabilities. You can transform raw confidence scores into calibrated probabilities that better reflect the true likelihood of correctness using the calibrate_scores method."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The first step is to train the claibrators that can done using `fit` or `fit_transform` method of `ScoreCalibrator` class. You can initiate a class object by choosing a method for training calibrators. Then call `fit_transform` method be providing `UQResult` object from training dataset and correct responses."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "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>prompt</th>\n",
       "      <th>response</th>\n",
       "      <th>logprob</th>\n",
       "      <th>normalized_probability</th>\n",
       "      <th>calibrated_normalized_probability</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>December 14, 1972</td>\n",
       "      <td>[{'token': 'December', 'logprob': -0.044779419...</td>\n",
       "      <td>0.980881</td>\n",
       "      <td>0.628571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>Bobby Scott and Bob Russell</td>\n",
       "      <td>[{'token': 'Bobby', 'logprob': -0.065702043473...</td>\n",
       "      <td>0.979054</td>\n",
       "      <td>0.628571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>1</td>\n",
       "      <td>[{'token': '1', 'logprob': -0.0108721693977713...</td>\n",
       "      <td>0.989187</td>\n",
       "      <td>0.633929</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>Super Bowl LII</td>\n",
       "      <td>[{'token': 'Super', 'logprob': -1.586603045463...</td>\n",
       "      <td>0.672557</td>\n",
       "      <td>0.490196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>South Carolina</td>\n",
       "      <td>[{'token': 'South', 'logprob': -1.502075701864...</td>\n",
       "      <td>0.999992</td>\n",
       "      <td>0.888889</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              prompt  \\\n",
       "0  You will be given a question. Return only the ...   \n",
       "1  You will be given a question. Return only the ...   \n",
       "2  You will be given a question. Return only the ...   \n",
       "3  You will be given a question. Return only the ...   \n",
       "4  You will be given a question. Return only the ...   \n",
       "\n",
       "                      response  \\\n",
       "0            December 14, 1972   \n",
       "1  Bobby Scott and Bob Russell   \n",
       "2                            1   \n",
       "3               Super Bowl LII   \n",
       "4               South Carolina   \n",
       "\n",
       "                                             logprob  normalized_probability  \\\n",
       "0  [{'token': 'December', 'logprob': -0.044779419...                0.980881   \n",
       "1  [{'token': 'Bobby', 'logprob': -0.065702043473...                0.979054   \n",
       "2  [{'token': '1', 'logprob': -0.0108721693977713...                0.989187   \n",
       "3  [{'token': 'Super', 'logprob': -1.586603045463...                0.672557   \n",
       "4  [{'token': 'South', 'logprob': -1.502075701864...                0.999992   \n",
       "\n",
       "   calibrated_normalized_probability  \n",
       "0                           0.628571  \n",
       "1                           0.628571  \n",
       "2                           0.633929  \n",
       "3                           0.490196  \n",
       "4                           0.888889  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sc = ScoreCalibrator(method=\"isotonic\")\n",
    "sc.fit_transform(uq_result=uq_result, correct_indicators=result_df.response_correct)\n",
    "\n",
    "results_df = uq_result.to_df()\n",
    "results_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can evaluate the performance of calibrated scores using `evaluate_calibration` method, which will require the correct responses."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "# Uncomment the following lines to visualize the calibrated scores\n",
    "\n",
    "# metrics = sc_object.evaluate_calibration(results, result_df.response_correct)\n",
    "# metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets generate responses and compute score on test dataset using `wbuq` object, which will return a `UQResult` object on test dataset which will contain test prompts, responses, and confidence scores."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1f7a5376b9e044f9af0795336b9d439f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Output()"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"></pre>\n"
      ],
      "text/plain": []
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "test_result = await wbuq.generate_and_score(prompts=test_prompts)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now have trained a `ScoreCalibrator` object containing fitted calibrators for each scorer (only `normalized_probability` for our example, but can be used with multiple scorers). Now, we can call `transform` method and provide a test dataset (UQResult object form test prompts), which will update the `UQResult` object to include calibrated scores.\n",
    "\n",
    "Note: `transform` method updates `UQResult` object in place, such that for every 'score', it will also contain 'calibrated_score'."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "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>prompt</th>\n",
       "      <th>response</th>\n",
       "      <th>logprob</th>\n",
       "      <th>normalized_probability</th>\n",
       "      <th>calibrated_normalized_probability</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>Games</td>\n",
       "      <td>[{'token': 'Games', 'logprob': -0.001829901477...</td>\n",
       "      <td>0.998172</td>\n",
       "      <td>0.578947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>Amir Johnson</td>\n",
       "      <td>[{'token': 'Am', 'logprob': -2.050269904430024...</td>\n",
       "      <td>0.999993</td>\n",
       "      <td>0.741935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>Frank Morris</td>\n",
       "      <td>[{'token': 'Frank', 'logprob': -8.653872646391...</td>\n",
       "      <td>0.999939</td>\n",
       "      <td>0.606061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>May 7, 1992</td>\n",
       "      <td>[{'token': 'May', 'logprob': -0.01998697593808...</td>\n",
       "      <td>0.996715</td>\n",
       "      <td>0.538462</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>You will be given a question. Return only the ...</td>\n",
       "      <td>Daisuke Ohata</td>\n",
       "      <td>[{'token': 'D', 'logprob': -8.153352973749861e...</td>\n",
       "      <td>0.999962</td>\n",
       "      <td>0.647059</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              prompt       response  \\\n",
       "0  You will be given a question. Return only the ...          Games   \n",
       "1  You will be given a question. Return only the ...   Amir Johnson   \n",
       "2  You will be given a question. Return only the ...   Frank Morris   \n",
       "3  You will be given a question. Return only the ...    May 7, 1992   \n",
       "4  You will be given a question. Return only the ...  Daisuke Ohata   \n",
       "\n",
       "                                             logprob  normalized_probability  \\\n",
       "0  [{'token': 'Games', 'logprob': -0.001829901477...                0.998172   \n",
       "1  [{'token': 'Am', 'logprob': -2.050269904430024...                0.999993   \n",
       "2  [{'token': 'Frank', 'logprob': -8.653872646391...                0.999939   \n",
       "3  [{'token': 'May', 'logprob': -0.01998697593808...                0.996715   \n",
       "4  [{'token': 'D', 'logprob': -8.153352973749861e...                0.999962   \n",
       "\n",
       "   calibrated_normalized_probability  \n",
       "0                           0.578947  \n",
       "1                           0.741935  \n",
       "2                           0.606061  \n",
       "3                           0.538462  \n",
       "4                           0.647059  "
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Calibrate scores\n",
    "sc.transform(test_result)\n",
    "\n",
    "test_result_df = test_result.to_df()\n",
    "test_result_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lets evaluate calibrated score from test dataset (since we also have correct response on test dataset)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x500 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "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>average_confidence</th>\n",
       "      <th>average_accuracy</th>\n",
       "      <th>calibration_gap</th>\n",
       "      <th>brier_score</th>\n",
       "      <th>log_loss</th>\n",
       "      <th>ece</th>\n",
       "      <th>mce</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>normalized_probability</th>\n",
       "      <td>0.904642</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.424642</td>\n",
       "      <td>0.421297</td>\n",
       "      <td>2.586512</td>\n",
       "      <td>0.428037</td>\n",
       "      <td>0.511129</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>calibrated_normalized_probability</th>\n",
       "      <td>0.492642</td>\n",
       "      <td>0.48</td>\n",
       "      <td>0.012642</td>\n",
       "      <td>0.233129</td>\n",
       "      <td>0.793354</td>\n",
       "      <td>0.030675</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                   average_confidence  average_accuracy  \\\n",
       "normalized_probability                       0.904642              0.48   \n",
       "calibrated_normalized_probability            0.492642              0.48   \n",
       "\n",
       "                                   calibration_gap  brier_score  log_loss  \\\n",
       "normalized_probability                    0.424642     0.421297  2.586512   \n",
       "calibrated_normalized_probability         0.012642     0.233129  0.793354   \n",
       "\n",
       "                                        ece       mce  \n",
       "normalized_probability             0.428037  0.511129  \n",
       "calibrated_normalized_probability  0.030675  0.500000  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Grade responses against correct answers for test set\n",
    "test_result_df[\"response_correct\"] = await grader.grade_responses(prompts=nq_open[\"question\"].to_list()[-n_test:], responses=test_result_df[\"response\"].to_list(), answers=nq_open[\"answer\"].to_list()[-n_test:])\n",
    "\n",
    "test_metrics = evaluate_calibration(test_result, test_result_df.response_correct)\n",
    "test_metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Note the substantial improvement in calibration quality before and after transforming confidence scores with the fitted `ScoreCalibrator` object."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Summary\n",
    "\n",
    "This calibration analysis demonstrates:\n",
    "\n",
    "### 🎯 **Key Findings**\n",
    "\n",
    "1. **Calibration Quality**: Use reliability diagrams and metrics like ECE and MCE score to assess how well confidence scores reflect true probabilities\n",
    "\n",
    "2. **Method Selection**: \n",
    "   - **Platt Scaling** works well for smaller datasets and when the calibration curve is roughly sigmoid-shaped\n",
    "   - **Isotonic Regression** is more flexible and can handle complex, non-parametric calibration curves\n",
    "\n",
    "3. **Practical Impact**: Calibration can significantly improve:\n",
    "   - Reliability of confidence scores for decision-making\n",
    "   - User trust in model predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "© 2025 CVS Health and/or one of its affiliates. All rights reserved."
   ]
  }
 ],
 "metadata": {
  "environment": {
   "kernel": "uqlm_my_test",
   "name": "workbench-notebooks.m126",
   "type": "gcloud",
   "uri": "us-docker.pkg.dev/deeplearning-platform-release/gcr.io/workbench-notebooks:m126"
  },
  "kernelspec": {
   "display_name": "uqlm_my_test",
   "language": "python",
   "name": "uqlm_my_test"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}