{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "1dc18eed-81fc-4f16-ab93-c188a55e0463",
   "metadata": {},
   "source": [
    "# 🎈 CleanLab之三：回归问题\n",
    "\n",
    "回归预测的是价格、收入、年龄等数值。回归数据集中的响应值可能由于以下原因而损坏：数据输入或测量错误、传感器或其他过程的噪声或数据管道损坏。为了查找数字列中的损坏值，我们将其视为目标值（即标签），由回归模型进行预测，然后使用 cleanlab 确定模型预测在偏离观察到的标签值时是否可信。[^1]\n",
    "\n",
    "[^1]: [Find Noisy Labels in Regression Datasets](https://docs.cleanlab.ai/stable/tutorials/regression.html)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "40df7579-909f-4a53-83be-dd19b7921d61",
   "metadata": {},
   "source": [
    "## 壹丨导入依赖"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "bb75d9e7d7996e83",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-07-31T16:44:41.538138Z",
     "start_time": "2024-07-31T16:44:39.991120Z"
    }
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from sklearn.ensemble import HistGradientBoostingRegressor\n",
    "from sklearn.model_selection import cross_val_predict\n",
    "from sklearn.metrics import r2_score\n",
    "\n",
    "from cleanlab.regression.learn import CleanLearning"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8464ca42-6883-4e74-a172-507cd8b6e261",
   "metadata": {},
   "source": [
    "## 贰丨加载并处理数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "2ae64b323258637d",
   "metadata": {},
   "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>exam_1</th>\n",
       "      <th>exam_2</th>\n",
       "      <th>exam_3</th>\n",
       "      <th>notes</th>\n",
       "      <th>final_score</th>\n",
       "      <th>true_final_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>72</td>\n",
       "      <td>81</td>\n",
       "      <td>80</td>\n",
       "      <td>NaN</td>\n",
       "      <td>73.3</td>\n",
       "      <td>73.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>89</td>\n",
       "      <td>62</td>\n",
       "      <td>93</td>\n",
       "      <td>NaN</td>\n",
       "      <td>83.8</td>\n",
       "      <td>83.8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>97</td>\n",
       "      <td>0</td>\n",
       "      <td>94</td>\n",
       "      <td>NaN</td>\n",
       "      <td>73.5</td>\n",
       "      <td>73.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>80</td>\n",
       "      <td>76</td>\n",
       "      <td>96</td>\n",
       "      <td>missed class frequently -10</td>\n",
       "      <td>78.6</td>\n",
       "      <td>78.6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>67</td>\n",
       "      <td>87</td>\n",
       "      <td>95</td>\n",
       "      <td>missed homework frequently -10</td>\n",
       "      <td>74.1</td>\n",
       "      <td>74.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   exam_1  exam_2  exam_3                           notes  final_score  \\\n",
       "0      72      81      80                             NaN         73.3   \n",
       "1      89      62      93                             NaN         83.8   \n",
       "2      97       0      94                             NaN         73.5   \n",
       "3      80      76      96     missed class frequently -10         78.6   \n",
       "4      67      87      95  missed homework frequently -10         74.1   \n",
       "\n",
       "   true_final_score  \n",
       "0              73.3  \n",
       "1              83.8  \n",
       "2              73.5  \n",
       "3              78.6  \n",
       "4              74.1  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "train_data = pd.read_csv(\"https://s.cleanlab.ai/student_grades_r/train.csv\")\n",
    "test_data = pd.read_csv(\"https://s.cleanlab.ai/student_grades_r/test.csv\")\n",
    "train_data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2e914557-3908-4788-96b3-a21259e0ed5f",
   "metadata": {},
   "source": [
    "`final_score`为噪声分数，`true_final_score`为真实分数。\n",
    "\n",
    "真实分数在现实中往往不可获取，这里仅做演示"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "3c8f70dbf819119e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Note: This pulldown content is for docs.cleanlab.ai, if running on local Jupyter or Colab, please ignore it.\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from mpl_toolkits.mplot3d import Axes3D\n",
    "\n",
    "def plot_data(train_data, errors_idx):\n",
    "    fig = plt.figure()\n",
    "    ax = fig.add_subplot(111, projection='3d')\n",
    "\n",
    "    x, y, z = train_data[\"exam_1\"], train_data[\"exam_2\"], train_data[\"exam_3\"]\n",
    "    labels = train_data[\"final_score\"]\n",
    "\n",
    "    img = ax.scatter(x, y, z, c=labels, cmap=\"jet\")\n",
    "    fig.colorbar(img)\n",
    "\n",
    "    ax.plot(\n",
    "        x.iloc[errors_idx],\n",
    "        y.iloc[errors_idx],\n",
    "        z.iloc[errors_idx],\n",
    "        \"x\",\n",
    "        markeredgecolor=\"black\",\n",
    "        markersize=10,\n",
    "        markeredgewidth=2.5,\n",
    "        alpha=0.8,\n",
    "        label=\"Label Errors\"\n",
    "    )\n",
    "    ax.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "147bc30b46bc1c36",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "errors_mask = train_data[\"final_score\"] != train_data[\"true_final_score\"]\n",
    "errors_idx = np.where(errors_mask == 1)\n",
    "\n",
    "plot_data(train_data, errors_idx)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "b2dda65d4deed37d",
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_columns = [\"exam_1\", \"exam_2\", \"exam_3\", \"notes\"]\n",
    "predicted_column = \"final_score\"\n",
    "\n",
    "X_train_raw, y_train = train_data[feature_columns], train_data[predicted_column]\n",
    "X_test_raw, y_test = test_data[feature_columns], test_data[predicted_column]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1087127d-a4b2-452f-b3da-1c7881f3f8a5",
   "metadata": {},
   "source": [
    "使用独热编码预处理数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "76e102aa3990207d",
   "metadata": {},
   "outputs": [],
   "source": [
    "categorical_features = [\"notes\"]\n",
    "X_train = pd.get_dummies(X_train_raw, columns=categorical_features)\n",
    "X_test = pd.get_dummies(X_test_raw, columns=categorical_features)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dd69f65c-4856-4600-8c69-7369bb8b13bb",
   "metadata": {},
   "source": [
    "## 叁丨定义回归模型并使用CleanLab查找潜在标签错误\n",
    "\n",
    "定义一个直方图梯度提升模型，并使用`find_label_issues`方法查找数据集的数字标签列中的潜在错误。\n",
    "\n",
    "可以替换成任何其他与sklearn兼容的回归模型\n",
    "\n",
    "`CleanLearning`通过交叉验证和引导法在内部拟合回归模型的多个副本，以计算数据集的预测和不确定性，用于识别标签问题。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "3983ccda-5a9a-41a6-b3e2-3f072d590d4e",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = HistGradientBoostingRegressor()\n",
    "cl = CleanLearning(model)\n",
    "\n",
    "label_issues = cl.find_label_issues(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cb0d279f-f656-40fa-b086-bc3edf9af53c",
   "metadata": {},
   "source": [
    "此方法返回一个DataFrame，包含每个样本的标签质量分数（0-1之间）。分数越低，被错误标记的可能性就越大。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "22da614c-dbfb-4740-853b-2faef9ceea0e",
   "metadata": {},
   "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>is_label_issue</th>\n",
       "      <th>label_quality</th>\n",
       "      <th>given_label</th>\n",
       "      <th>predicted_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>0.230476</td>\n",
       "      <td>73.3</td>\n",
       "      <td>77.058670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>False</td>\n",
       "      <td>0.804244</td>\n",
       "      <td>83.8</td>\n",
       "      <td>83.212808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>True</td>\n",
       "      <td>0.050572</td>\n",
       "      <td>73.5</td>\n",
       "      <td>62.947544</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>False</td>\n",
       "      <td>0.280991</td>\n",
       "      <td>78.6</td>\n",
       "      <td>75.172543</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>False</td>\n",
       "      <td>0.466165</td>\n",
       "      <td>74.1</td>\n",
       "      <td>76.343107</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   is_label_issue  label_quality  given_label  predicted_label\n",
       "0           False       0.230476         73.3        77.058670\n",
       "1           False       0.804244         83.8        83.212808\n",
       "2            True       0.050572         73.5        62.947544\n",
       "3           False       0.280991         78.6        75.172543\n",
       "4           False       0.466165         74.1        76.343107"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label_issues.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "2c56d076-d158-47a7-9b74-a7ed7ae351e0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cleanlab found 141 potential label errors in the dataset.\n",
      "Here are indices of the top 10 most likely errors: \n",
      " [659  56 318 367 688 560 657 685 305 289]\n"
     ]
    }
   ],
   "source": [
    "# 可以获取标记有标签问题的样本子集按照标签质量分数排序以找到最有可能标记错误的10个样本的索引\n",
    "identified_issues = label_issues[label_issues[\"is_label_issue\"] == True]\n",
    "lowest_quality_labels = label_issues[\"label_quality\"].argsort()[:10].to_numpy()\n",
    "print(\n",
    "    f\"cleanlab found {len(identified_issues)} potential label errors in the dataset.\\n\"\n",
    "    f\"Here are indices of the top 10 most likely errors: \\n {lowest_quality_labels}\"\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "f4f11f77-191c-4260-baae-f648fd25346d",
   "metadata": {},
   "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>exam_1</th>\n",
       "      <th>exam_2</th>\n",
       "      <th>exam_3</th>\n",
       "      <th>notes</th>\n",
       "      <th>given_label</th>\n",
       "      <th>predicted_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>659</th>\n",
       "      <td>67</td>\n",
       "      <td>93</td>\n",
       "      <td>93</td>\n",
       "      <td>NaN</td>\n",
       "      <td>17.4</td>\n",
       "      <td>84.7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>75</td>\n",
       "      <td>83</td>\n",
       "      <td>69</td>\n",
       "      <td>NaN</td>\n",
       "      <td>8.9</td>\n",
       "      <td>72.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>318</th>\n",
       "      <td>41</td>\n",
       "      <td>88</td>\n",
       "      <td>98</td>\n",
       "      <td>missed class frequently -10</td>\n",
       "      <td>0.0</td>\n",
       "      <td>67.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>367</th>\n",
       "      <td>78</td>\n",
       "      <td>0</td>\n",
       "      <td>86</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>57.9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>688</th>\n",
       "      <td>91</td>\n",
       "      <td>68</td>\n",
       "      <td>88</td>\n",
       "      <td>NaN</td>\n",
       "      <td>36.3</td>\n",
       "      <td>80.6</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     exam_1  exam_2  exam_3                        notes  given_label  \\\n",
       "659      67      93      93                          NaN         17.4   \n",
       "56       75      83      69                          NaN          8.9   \n",
       "318      41      88      98  missed class frequently -10          0.0   \n",
       "367      78       0      86                          NaN          0.0   \n",
       "688      91      68      88                          NaN         36.3   \n",
       "\n",
       "     predicted_label  \n",
       "659             84.7  \n",
       "56              72.5  \n",
       "318             67.0  \n",
       "367             57.9  \n",
       "688             80.6  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def view_datapoint(index):\n",
    "    given_labels = label_issues[\"given_label\"]\n",
    "    predicted_labels = label_issues[\"predicted_label\"].round(1)\n",
    "    return pd.concat(\n",
    "        [X_train_raw, given_labels, predicted_labels], axis=1\n",
    "    ).iloc[index]\n",
    "\n",
    "view_datapoint(lowest_quality_labels[:5])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ada75f99-e629-4234-bc62-b2be54b85899",
   "metadata": {},
   "source": [
    "## 肆丨从噪声标签中训练出更鲁棒的模型\n",
    "\n",
    "cleanlab可以过滤嘈杂的样本并在剩余干净的数据上训练模型"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ee10de07-8a22-4026-8b5b-e757cf076b02",
   "metadata": {},
   "source": [
    "### 1. 建立基准"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f2717d51-0aef-46ac-82d5-833bac7d06d7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "r-squared score of original model: 0.838\n"
     ]
    }
   ],
   "source": [
    "baseline_model = HistGradientBoostingRegressor()\n",
    "baseline_model.fit(X_train, y_train)\n",
    "\n",
    "preds_og = baseline_model.predict(X_test)\n",
    "r2_og = r2_score(y_test, preds_og)\n",
    "print(f\"r-squared score of original model: {r2_og:.3f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "851e3b2a-5171-4dd8-975e-40654497d455",
   "metadata": {},
   "source": [
    "通过`label_issues`参数传递已经计算的标签问题。\n",
    "\n",
    "不传递`label_issues`时，将通过交叉验证重新估计"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8e873638-5d35-490c-94cd-a8be02814c9a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "r-squared score of cleanlab's model: 0.916\n"
     ]
    }
   ],
   "source": [
    "found_label_issues = cl.get_label_issues()\n",
    "cl.fit(X_train, y_train, label_issues=found_label_issues)\n",
    "\n",
    "preds_cl = cl.predict(X_test)\n",
    "r2_cl = r2_score(y_test, preds_cl)\n",
    "print(f\"r-squared score of cleanlab's model: {r2_cl:.3f}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb64304f-1da0-40dd-9469-74991807624c",
   "metadata": {},
   "source": [
    "## 伍丨在回归数据集中查找噪声标签的其他方法\n",
    "\n",
    "如果模型或数据与CleanLearning不兼容，可以对回归模型运行交叉验证以获取样本外预测，然后使用Datalab评估数据集中每个样本的质量分数"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "018b3345-f904-48fc-bc31-03f452439346",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Finding label issues ...\n",
      "\n",
      "Audit complete. 50 issues found in the dataset.\n"
     ]
    }
   ],
   "source": [
    "# Get out-of-sample predictions using cross-validation:\n",
    "model = HistGradientBoostingRegressor()\n",
    "predictions = cross_val_predict(estimator=model, X=X_train, y=y_train)\n",
    "\n",
    "\n",
    "from cleanlab import Datalab\n",
    "\n",
    "lab = Datalab(\n",
    "    data=train_data.drop(columns=[\"true_final_score\"]),\n",
    "    label_name=\"final_score\",\n",
    "    task=\"regression\",\n",
    ")\n",
    "\n",
    "lab.find_issues(\n",
    "    pred_probs=predictions,\n",
    "    issue_types={\"label\": {}},  # specify we're only interested in label issues here\n",
    ")\n",
    "\n",
    "\n",
    "label_issues = lab.get_issues(\"label\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "e21a0d8e-c2cc-43f0-8581-945ca8c46c1d",
   "metadata": {},
   "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>is_label_issue</th>\n",
       "      <th>label_score</th>\n",
       "      <th>given_label</th>\n",
       "      <th>predicted_label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>318</th>\n",
       "      <td>True</td>\n",
       "      <td>1.968627e-09</td>\n",
       "      <td>0.0</td>\n",
       "      <td>78.228799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>659</th>\n",
       "      <td>True</td>\n",
       "      <td>2.646674e-08</td>\n",
       "      <td>17.4</td>\n",
       "      <td>86.402962</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>True</td>\n",
       "      <td>4.323818e-08</td>\n",
       "      <td>8.9</td>\n",
       "      <td>75.952758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>160</th>\n",
       "      <td>True</td>\n",
       "      <td>2.422144e-07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>60.456908</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>367</th>\n",
       "      <td>True</td>\n",
       "      <td>8.465815e-07</td>\n",
       "      <td>0.0</td>\n",
       "      <td>55.753968</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     is_label_issue   label_score  given_label  predicted_label\n",
       "318            True  1.968627e-09          0.0        78.228799\n",
       "659            True  2.646674e-08         17.4        86.402962\n",
       "56             True  4.323818e-08          8.9        75.952758\n",
       "160            True  2.422144e-07          0.0        60.456908\n",
       "367            True  8.465815e-07          0.0        55.753968"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "label_issues.sort_values(\"label_score\").head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fa9c6a5f-28a6-481a-888e-d32adb3e96d1",
   "metadata": {},
   "source": [
    "建议使用`CleanLearning`包装回归模型，而不是直接提供预测结果。\n",
    "\n",
    "这样做可以更好的考虑回归模型中的不确定性（包括随机不确定性和模型不确定性）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "bbe5f3d0-8e28-4243-8366-d1a29633eddc",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Finding label issues ...\n",
      "\n",
      "Audit complete. 150 issues found in the dataset.\n"
     ]
    }
   ],
   "source": [
    "lab = Datalab(\n",
    "    data=train_data.drop(columns=[\"true_final_score\"]),\n",
    "    label_name=\"final_score\",\n",
    "    task=\"regression\",\n",
    ")\n",
    "\n",
    "lab.find_issues(\n",
    "    features=X_train,\n",
    "    issue_types={  # Optional drop this to simultaneously detect many types of data/label issues\n",
    "        \"label\": {\n",
    "            # Optional: Specify which type of 🔟 scikit-learn-compatible regression model is used to find label errors\n",
    "            \"clean_learning_kwargs\": {\"model\": HistGradientBoostingRegressor()}\n",
    "        }\n",
    "    },\n",
    ")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2ebaa7fe-39c4-406d-8943-3b267a55877e",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}