find_label_errors_iris/find_label_errors_iris.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "fca30380",
   "metadata": {
    "papermill": {
     "duration": 0.007628,
     "end_time": "2022-09-08T14:39:23.128057",
     "exception": false,
     "start_time": "2022-09-08T14:39:23.120429",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Find Label Errors Introduced in the Iris Classification Dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6ee6e7b1",
   "metadata": {},
   "source": [
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/cleanlab/examples/blob/master/find_label_errors_iris/find_label_errors_iris.ipynb)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "91188b55",
   "metadata": {
    "papermill": {
     "duration": 0.003554,
     "end_time": "2022-09-08T14:39:23.136307",
     "exception": false,
     "start_time": "2022-09-08T14:39:23.132753",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "In this example we show how `cleanlab` can be used to find label errors and improve the performance of a classifier model in classification tasks. We will use a modified version of the Iris dataset that has artificially introduced label errors to demonstrate this task.\n",
    "\n",
    "Here we show the performance with LogisticRegression classifier versus LogisticRegression *without* cleanlab on the modified Iris dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "73c34481",
   "metadata": {
    "papermill": {
     "duration": 0.002874,
     "end_time": "2022-09-08T14:39:23.142780",
     "exception": false,
     "start_time": "2022-09-08T14:39:23.139906",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Note: `cleanlab` can be used with any classifier and dataset for multiclass learning with noisy labels. Its comprised of components from the theory and algorithms of **confident learning**. It's a Python class that wraps around\n",
    "any classifier as long as .fit(X, y, sample_weight), .predict(X), .predict_proba(X) are defined.\n",
    "See https://docs.cleanlab.ai/ for docs."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb73f045",
   "metadata": {},
   "source": [
    "Please install the dependencies specified in this [requirements.txt](https://github.com/cleanlab/examples/blob/master/find_label_errors_iris/requirements.txt) file before running the notebook."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a0fe6946",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-08T14:39:23.149217Z",
     "iopub.status.busy": "2022-09-08T14:39:23.148949Z",
     "iopub.status.idle": "2022-09-08T14:39:23.757306Z",
     "shell.execute_reply": "2022-09-08T14:39:23.756896Z"
    },
    "papermill": {
     "duration": 0.61306,
     "end_time": "2022-09-08T14:39:23.758439",
     "exception": false,
     "start_time": "2022-09-08T14:39:23.145379",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import datasets\n",
    "from sklearn.metrics import accuracy_score\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.model_selection import ParameterGrid\n",
    "import numpy as np\n",
    "from cleanlab.classification import CleanLearning\n",
    "from cleanlab.benchmarking.noise_generation import generate_noisy_labels\n",
    "from cleanlab.internal.util import value_counts\n",
    "from cleanlab.internal.latent_algebra import compute_inv_noise_matrix"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e78e14e3",
   "metadata": {
    "papermill": {
     "duration": 0.001202,
     "end_time": "2022-09-08T14:39:23.761049",
     "exception": false,
     "start_time": "2022-09-08T14:39:23.759847",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Visualizing the Iris dataset distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "46977708",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-08T14:39:23.763928Z",
     "iopub.status.busy": "2022-09-08T14:39:23.763695Z",
     "iopub.status.idle": "2022-09-08T14:39:24.139783Z",
     "shell.execute_reply": "2022-09-08T14:39:24.139492Z"
    },
    "papermill": {
     "duration": 0.378673,
     "end_time": "2022-09-08T14:39:24.140825",
     "exception": false,
     "start_time": "2022-09-08T14:39:23.762152",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Seed for reproducibility\n",
    "seed = 2\n",
    "clf = LogisticRegression(solver=\"lbfgs\", multi_class=\"auto\", max_iter=1000)\n",
    "cl = CleanLearning(clf=clf, seed=seed, verbose=False)\n",
    "np.random.seed(seed=seed)\n",
    "\n",
    "# Get iris dataset\n",
    "iris = datasets.load_iris()\n",
    "X = iris.data  # we only take the first two features.\n",
    "y = iris.target\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)\n",
    "\n",
    "try:\n",
    "    get_ipython().run_line_magic(\"matplotlib\", \"inline\")\n",
    "    from matplotlib import pyplot as plt\n",
    "\n",
    "    _ = plt.figure(figsize=(12, 8))\n",
    "    color_list = plt.cm.tab10(np.linspace(0, 1, 6))\n",
    "    _ = plt.scatter(\n",
    "        X_train[:, 1], X_train[:, 3], color=[color_list[z] for z in y_train], s=50\n",
    "    )\n",
    "    ax = plt.gca()\n",
    "    #     plt.axis('off')\n",
    "    ax.spines[\"top\"].set_visible(False)\n",
    "    ax.spines[\"right\"].set_visible(False)\n",
    "    _ = ax.get_xaxis().set_ticks([])\n",
    "    _ = ax.get_yaxis().set_ticks([])\n",
    "    _ = plt.title(\"Iris dataset (feature 3 vs feature 1)\", fontsize=30)\n",
    "except Exception as e:\n",
    "    print(e)\n",
    "    print(\"Plotting is only supported in an iPython interface.\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "692ad886",
   "metadata": {
    "papermill": {
     "duration": 0.00145,
     "end_time": "2022-09-08T14:39:24.144665",
     "exception": false,
     "start_time": "2022-09-08T14:39:24.143215",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Generating and adding noise to the Iris dataset then visualizing the distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "1905bf23",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-08T14:39:24.147997Z",
     "iopub.status.busy": "2022-09-08T14:39:24.147841Z",
     "iopub.status.idle": "2022-09-08T14:39:24.307012Z",
     "shell.execute_reply": "2022-09-08T14:39:24.306709Z"
    },
    "papermill": {
     "duration": 0.161979,
     "end_time": "2022-09-08T14:39:24.308005",
     "exception": false,
     "start_time": "2022-09-08T14:39:24.146026",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Generate lots of noise.\n",
    "noise_matrix = np.array(\n",
    "    [\n",
    "        [0.5, 0.0, 0.0],\n",
    "        [0.5, 1.0, 0.5],\n",
    "        [0.0, 0.0, 0.5],\n",
    "    ]\n",
    ")\n",
    "\n",
    "py = value_counts(y_train)\n",
    "# Create noisy labels\n",
    "s = generate_noisy_labels(y_train, noise_matrix)\n",
    "\n",
    "try:\n",
    "    get_ipython().run_line_magic(\"matplotlib\", \"inline\")\n",
    "    from matplotlib import pyplot as plt\n",
    "\n",
    "    _ = plt.figure(figsize=(15, 8))\n",
    "    color_list = plt.cm.tab10(np.linspace(0, 1, 6))\n",
    "    for k in range(len(np.unique(y_train))):\n",
    "        X_k = X_train[y_train == k]  # data for class k\n",
    "        _ = plt.scatter(\n",
    "            X_k[:, 1],\n",
    "            X_k[:, 3],\n",
    "            color=[color_list[noisy_label] for noisy_label in s[y_train == k]],\n",
    "            s=200,\n",
    "            marker=r\"${a}$\".format(a=str(k)),\n",
    "            linewidth=1,\n",
    "        )\n",
    "    _ = plt.scatter(\n",
    "        x=X_train[s != y_train][:, 1],\n",
    "        y=X_train[s != y_train][:, 3],\n",
    "        color=[color_list[z] for z in s],\n",
    "        s=400,\n",
    "        facecolors=\"none\",\n",
    "        edgecolors=\"black\",\n",
    "        linewidth=2,\n",
    "        alpha=0.5,\n",
    "    )\n",
    "    ax = plt.gca()\n",
    "    ax.spines[\"top\"].set_visible(False)\n",
    "    ax.spines[\"right\"].set_visible(False)\n",
    "    _ = ax.get_xaxis().set_ticks([])\n",
    "    _ = ax.get_yaxis().set_ticks([])\n",
    "    _ = plt.title(\"Iris dataset (features 3 and 1). Label errors circled.\", fontsize=30)\n",
    "except Exception as e:\n",
    "    print(e)\n",
    "    print(\"Plotting is only supported in an iPython interface.\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "973b2492",
   "metadata": {
    "papermill": {
     "duration": 0.001733,
     "end_time": "2022-09-08T14:39:24.311777",
     "exception": false,
     "start_time": "2022-09-08T14:39:24.310044",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Comparing the performance of the classifier with and without `cleanlab`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "7fd3f126",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-08T14:39:24.315650Z",
     "iopub.status.busy": "2022-09-08T14:39:24.315541Z",
     "iopub.status.idle": "2022-09-08T14:39:25.991552Z",
     "shell.execute_reply": "2022-09-08T14:39:25.991178Z"
    },
    "papermill": {
     "duration": 1.679135,
     "end_time": "2022-09-08T14:39:25.992591",
     "exception": false,
     "start_time": "2022-09-08T14:39:24.313456",
     "status": "completed"
    },
    "pycharm": {
     "name": "#%%\n"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WITHOUT confident learning, Iris dataset test accuracy: 0.6\n",
      "\n",
      "Now we show improvement using cleanlab to characterize the noise\n",
      "and learn on the data that is (with high confidence) labeled correctly.\n",
      "\n",
      "WITH confident learning (noise matrix given), Iris dataset test accuracy: 0.83\n",
      "\n",
      "WITH confident learning (noise / inverse noise matrix given), Iris dataset test accuracy: 0.77\n",
      "\n",
      "WITH confident learning noise not given, Iris dataset test accuracy: 0.77\n"
     ]
    }
   ],
   "source": [
    "print(\"WITHOUT confident learning,\", end=\" \")\n",
    "clf = LogisticRegression(solver=\"lbfgs\", multi_class=\"auto\", max_iter=1000)\n",
    "_ = clf.fit(X_train, s)\n",
    "pred = clf.predict(X_test)\n",
    "print(\"Iris dataset test accuracy:\", round(accuracy_score(pred, y_test), 2))\n",
    "\n",
    "print(\"\\nNow we show improvement using cleanlab to characterize the noise\")\n",
    "print(\"and learn on the data that is (with high confidence) labeled correctly.\")\n",
    "print()\n",
    "\n",
    "print(\"WITH confident learning (noise matrix given),\", end=\" \")\n",
    "_ = cl.fit(X_train, s, noise_matrix=noise_matrix)\n",
    "pred = cl.predict(X_test)\n",
    "print(\"Iris dataset test accuracy:\", round(accuracy_score(pred, y_test), 2))\n",
    "print()\n",
    "\n",
    "print(\"WITH confident learning (noise / inverse noise matrix given),\", end=\" \")\n",
    "inv = compute_inv_noise_matrix(py, noise_matrix)\n",
    "_ = cl.fit(X_train, s, noise_matrix=noise_matrix, inverse_noise_matrix=inv)\n",
    "pred = cl.predict(X_test)\n",
    "print(\"Iris dataset test accuracy:\", round(accuracy_score(pred, y_test), 2))\n",
    "print()\n",
    "\n",
    "print(\"WITH confident learning noise not given,\", end=\" \")\n",
    "clf = LogisticRegression(solver=\"lbfgs\", multi_class=\"auto\", max_iter=1000)\n",
    "cl = CleanLearning(clf=clf, seed=seed, verbose=False)\n",
    "_ = cl.fit(X_train, s)\n",
    "pred = cl.predict(X_test)\n",
    "print(\"Iris dataset test accuracy:\", round(accuracy_score(pred, y_test), 2))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e02ad5d4",
   "metadata": {
    "papermill": {
     "duration": 0.001778,
     "end_time": "2022-09-08T14:39:25.996483",
     "exception": false,
     "start_time": "2022-09-08T14:39:25.994705",
     "status": "completed"
    },
    "pycharm": {
     "name": "#%% md\n"
    },
    "tags": []
   },
   "source": [
    "## Performance of confident learning across varying settings.\n",
    "\n",
    "For more details of the parameters, inspect [`cleanlab/filter.py`](https://docs.cleanlab.ai/stable/cleanlab/filter.html) and [`cleanlab/classification.py`](https://docs.cleanlab.ai/stable/cleanlab/classification.html).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5fd91b26",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-08T14:39:26.000510Z",
     "iopub.status.busy": "2022-09-08T14:39:26.000377Z",
     "iopub.status.idle": "2022-09-08T14:39:30.210494Z",
     "shell.execute_reply": "2022-09-08T14:39:30.210116Z"
    },
    "jupyter": {
     "outputs_hidden": false
    },
    "papermill": {
     "duration": 4.213388,
     "end_time": "2022-09-08T14:39:30.211538",
     "exception": false,
     "start_time": "2022-09-08T14:39:25.998150",
     "status": "completed"
    },
    "pycharm": {
     "name": "#%%\n"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'prune_by_class'}, 'converge_latent_estimates': False}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.83 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'prune_by_class'}, 'converge_latent_estimates': True}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.83 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'both'}, 'converge_latent_estimates': False}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.8 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'both'}, 'converge_latent_estimates': True}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.8 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'prune_by_noise_rate'}, 'converge_latent_estimates': False}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.77 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'prune_by_noise_rate'}, 'converge_latent_estimates': True}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.77 \n",
      "\n"
     ]
    }
   ],
   "source": [
    "param_grid = {\n",
    "    \"find_label_issues_kwargs\": [\n",
    "        {\"filter_by\": \"prune_by_noise_rate\"},\n",
    "        {\"filter_by\": \"prune_by_class\"},\n",
    "        {\"filter_by\": \"both\"},\n",
    "    ],\n",
    "    \"converge_latent_estimates\": [True, False],\n",
    "}\n",
    "\n",
    "# Fit CleanLearning across all parameter settings.\n",
    "params = ParameterGrid(param_grid)\n",
    "scores = []\n",
    "for param in params:\n",
    "    clf = LogisticRegression(solver=\"lbfgs\", multi_class=\"auto\", max_iter=1000)\n",
    "    cl = CleanLearning(clf=clf, verbose=False, **param)\n",
    "    _ = cl.fit(X_train, s)  # s is the noisy y_train labels\n",
    "    scores.append(accuracy_score(cl.predict(X_test), y_test))\n",
    "\n",
    "# Print results sorted from best to least\n",
    "for i in np.argsort(scores)[::-1]:\n",
    "    print()\n",
    "    print(\"Param settings:\", params[i])\n",
    "    print(\n",
    "        \"Iris dataset test accuracy (using confident learning):\\t\",\n",
    "        round(scores[i], 2),\n",
    "        \"\\n\",\n",
    "    )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d62a786e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-09-08T14:39:30.216275Z",
     "iopub.status.busy": "2022-09-08T14:39:30.216138Z",
     "iopub.status.idle": "2022-09-08T14:39:34.384385Z",
     "shell.execute_reply": "2022-09-08T14:39:34.384000Z"
    },
    "papermill": {
     "duration": 4.171674,
     "end_time": "2022-09-08T14:39:34.385413",
     "exception": false,
     "start_time": "2022-09-08T14:39:30.213739",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'prune_by_noise_rate'}, 'converge_latent_estimates': True}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.87 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'both'}, 'converge_latent_estimates': False}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.83 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'prune_by_class'}, 'converge_latent_estimates': False}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.83 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'prune_by_noise_rate'}, 'converge_latent_estimates': False}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.77 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'both'}, 'converge_latent_estimates': True}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.77 \n",
      "\n",
      "\n",
      "Param settings: {'find_label_issues_kwargs': {'filter_by': 'prune_by_class'}, 'converge_latent_estimates': True}\n",
      "Iris dataset test accuracy (using confident learning):\t 0.77 \n",
      "\n"
     ]
    }
   ],
   "source": [
    "param_grid = {\n",
    "    \"find_label_issues_kwargs\": [\n",
    "        {\"filter_by\": \"prune_by_noise_rate\"},\n",
    "        {\"filter_by\": \"prune_by_class\"},\n",
    "        {\"filter_by\": \"both\"},\n",
    "    ],\n",
    "    \"converge_latent_estimates\": [True, False],\n",
    "}\n",
    "\n",
    "# Fit CleanLearning across all parameter settings.\n",
    "params = ParameterGrid(param_grid)\n",
    "scores = []\n",
    "for param in params:\n",
    "    clf = LogisticRegression(solver=\"lbfgs\", multi_class=\"auto\", max_iter=1000)\n",
    "    cl = CleanLearning(clf=clf, verbose=False, **param)\n",
    "    _ = cl.fit(X_train, s)  # s is the noisy y_train labels\n",
    "    scores.append(accuracy_score(cl.predict(X_test), y_test))\n",
    "\n",
    "# Print results sorted from best to least\n",
    "for i in np.argsort(scores)[::-1]:\n",
    "    print()\n",
    "    print(\"Param settings:\", params[i])\n",
    "    print(\n",
    "        \"Iris dataset test accuracy (using confident learning):\\t\",\n",
    "        round(scores[i], 2),\n",
    "        \"\\n\",\n",
    "    )\n"
   ]
  }
 ],
 "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": "3.9.13"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 14.865186,
   "end_time": "2022-09-08T14:39:37.007566",
   "environment_variables": {},
   "exception": null,
   "input_path": "./1_iris_simple_example/iris_simple_example.ipynb",
   "output_path": "./1_iris_simple_example/iris_simple_example.ipynb",
   "parameters": {},
   "start_time": "2022-09-08T14:39:22.142380",
   "version": "2.3.4"
  },
  "pycharm": {
   "stem_cell": {
    "cell_type": "raw",
    "metadata": {
     "collapsed": false
    },
    "source": []
   }
  },
  "vscode": {
   "interpreter": {
    "hash": "00885e89789f58e60dbba52a405dc834aaf92411914fde0d391f9b48289a0610"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}