rezapci/Ionosphere_Nearest_Neighbour.ipynb

## Ionosphere_Nearest_Neighbour.ipynb
{
  "cells": [
    {
      "metadata": {},
      "cell_type": "markdown",
      "source": "| Name | Description | Date \n| :- |-------------: | :-:\n|Reza Hashemi| Data Mining: Lonosphere Neares Neighbour | September 23rd 2019 |"
    },
    {
      "metadata": {},
      "cell_type": "markdown",
      "source": "<img src=\"https://live.staticflickr.com/7511/16149273389_32d9fc3081_k.jpg\" width=\"400\" align=\"center\"></a>\n\n## The dataset Ionosphere, which high-frequency antennas. The aim of the antennas is to determine whether there is a structure in the ionosphere and a region in the upper atmosphere. We consider readings with a structure to be good, while those that do not have structure are deemed bad. The aim of this application is to build a data mining classifier that can determine whether an image is good or bad."
    },
    {
      "metadata": {},
      "cell_type": "markdown",
      "source": "#### For each row in the dataset, there are 35 values. The first 34 are measurements taken from the 17 antennas (two values for each antenna). The last is either 'g' or 'b'; that stands for good and bad, respectively."
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:45:13.456569Z",
          "start_time": "2019-09-23T20:45:13.452665Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "# check scikit-learn version\nimport sklearn\nprint('sklearn: %s' % sklearn.__version__)",
      "execution_count": 45,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "sklearn: 0.21.2\n"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:33.164389Z",
          "start_time": "2019-09-23T20:46:33.160485Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "import warnings\nwarnings.filterwarnings(\"ignore\", category=FutureWarning)",
      "execution_count": 46,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:34.028173Z",
          "start_time": "2019-09-23T20:46:34.024243Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "%matplotlib inline",
      "execution_count": 47,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:34.611794Z",
          "start_time": "2019-09-23T20:46:34.607890Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "import os\nhome_folder = os.path.expanduser(\"~\")\nprint(home_folder)",
      "execution_count": 48,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "C:\\Users\\Reza Hashemi\n"
        }
      ]
    },
    {
      "metadata": {},
      "cell_type": "markdown",
      "source": "#### then i create the X and y NumPy arrays to store the dataset in. The sizes of these arrays are known from the dataset. Since i  don't know the size of future datasets - i will use other methods to load the dataset in future chapters and i don't need to know this size beforehand:"
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:35.267665Z",
          "start_time": "2019-09-23T20:46:35.264736Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "# Change this to the location of your dataset\ndata_filename = \"data/ionosphere.data\"\nprint(data_filename)",
      "execution_count": 49,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "data/ionosphere.data\n"
        }
      ]
    },
    {
      "metadata": {},
      "cell_type": "markdown",
      "source": "#### The dataset is in a Comma-Separated Values (CSV) format, which is a commonly used format for datasets. I'm going to use the csv module to load this file. Import it and set up a csv reader object, then loop through the file, setting the appropriate row in X and class value in y for every line in our dataset:"
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:36.004544Z",
          "start_time": "2019-09-23T20:46:35.993807Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "import csv\nimport numpy as np\n\n# Size taken from the dataset and is known already\nX = np.zeros((351, 34), dtype='float')\ny = np.zeros((351,), dtype='bool')\n\nwith open(data_filename, 'r') as input_file:\n    reader = csv.reader(input_file)\n    for i, row in enumerate(reader):\n        # Get the data, converting each item to a float\n        data = [float(datum) for datum in row[:-1]]\n        # Set the appropriate row in our dataset\n        X[i] = data\n        # 1 if the class is 'g', 0 otherwise\n        y[i] = row[-1] == 'g'",
      "execution_count": 50,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:36.613567Z",
          "start_time": "2019-09-23T20:46:36.607710Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "from sklearn.model_selection import train_test_split\nX_train, X_test, y_train, y_test = train_test_split(X, y, random_state=14)\nprint(\"There are {} samples in the training dataset\".format(X_train.shape[0]))\nprint(\"There are {} samples in the testing dataset\".format(X_test.shape[0]))\nprint(\"Each sample has {} features\".format(X_train.shape[1]))",
      "execution_count": 51,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "There are 263 samples in the training dataset\nThere are 88 samples in the testing dataset\nEach sample has 34 features\n"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:37.210878Z",
          "start_time": "2019-09-23T20:46:37.206973Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "from sklearn.neighbors import KNeighborsClassifier\n\nestimator = KNeighborsClassifier()",
      "execution_count": 52,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:37.532957Z",
          "start_time": "2019-09-23T20:46:37.527101Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "estimator.fit(X_train, y_train)",
      "execution_count": 53,
      "outputs": [
        {
          "data": {
            "text/plain": "KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n                     metric_params=None, n_jobs=None, n_neighbors=5, p=2,\n                     weights='uniform')"
          },
          "execution_count": 53,
          "metadata": {},
          "output_type": "execute_result"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:38.008295Z",
          "start_time": "2019-09-23T20:46:37.999484Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "y_predicted = estimator.predict(X_test)\naccuracy = np.mean(y_test == y_predicted) * 100\nprint(\"The accuracy is {0:.1f}%\".format(accuracy))",
      "execution_count": 54,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "The accuracy is 86.4%\n"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:38.547025Z",
          "start_time": "2019-09-23T20:46:38.544091Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "from sklearn.model_selection import cross_val_score",
      "execution_count": 55,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:38.948155Z",
          "start_time": "2019-09-23T20:46:38.927687Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "scores = cross_val_score(estimator, X, y, scoring='accuracy')\naverage_accuracy = np.mean(scores) * 100\nprint(\"The average accuracy is {0:.1f}%\".format(average_accuracy))",
      "execution_count": 56,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "The average accuracy is 82.3%\n"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:40.371158Z",
          "start_time": "2019-09-23T20:46:40.054936Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "avg_scores = []\nall_scores = []\nparameter_values = list(range(1, 21))  # Including 20\nfor n_neighbors in parameter_values:\n    estimator = KNeighborsClassifier(n_neighbors=n_neighbors)\n    scores = cross_val_score(estimator, X, y, scoring='accuracy')\n    avg_scores.append(np.mean(scores))\n    all_scores.append(scores)",
      "execution_count": 57,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:41.053382Z",
          "start_time": "2019-09-23T20:46:41.025078Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "plt.plot?",
      "execution_count": 58,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:41.955204Z",
          "start_time": "2019-09-23T20:46:41.599941Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "from matplotlib import pyplot as plt\nplt.figure(figsize=(32,20))\nplt.plot(parameter_values, avg_scores, '-o', linewidth=5, markersize=24)\n#plt.axis([0, max(parameter_values), 0, 1.0])",
      "execution_count": 59,
      "outputs": [
        {
          "data": {
            "text/plain": "[<matplotlib.lines.Line2D at 0x1c0b6ab7320>]"
          },
          "execution_count": 59,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": "<Figure size 2304x1440 with 1 Axes>"
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:44.683119Z",
          "start_time": "2019-09-23T20:46:44.526959Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "for parameter, scores in zip(parameter_values, all_scores):\n    n_scores = len(scores)\n    plt.plot([parameter] * n_scores, scores, '-o')",
      "execution_count": 60,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": "<Figure size 432x288 with 1 Axes>"
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:46:45.514670Z",
          "start_time": "2019-09-23T20:46:45.391694Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "plt.plot(parameter_values, all_scores, 'bx')",
      "execution_count": 61,
      "outputs": [
        {
          "data": {
            "text/plain": "[<matplotlib.lines.Line2D at 0x1c0b76251d0>,\n <matplotlib.lines.Line2D at 0x1c0b6a50160>,\n <matplotlib.lines.Line2D at 0x1c0b76252e8>]"
          },
          "execution_count": 61,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": "<Figure size 432x288 with 1 Axes>"
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:42.929716Z",
          "start_time": "2019-09-23T20:46:46.079773Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "from collections import defaultdict\nall_scores = defaultdict(list)\nparameter_values = list(range(1, 21))  # Including 20\nfor n_neighbors in parameter_values:\n    for i in range(100):\n        estimator = KNeighborsClassifier(n_neighbors=n_neighbors)\n        scores = cross_val_score(estimator, X, y, scoring='accuracy', cv=10)\n        all_scores[n_neighbors].append(scores)\nfor parameter in parameter_values:\n    scores = all_scores[parameter]\n    n_scores = len(scores)\n    plt.plot([parameter] * n_scores, scores, '-o')",
      "execution_count": 62,
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": "<Figure size 432x288 with 1 Axes>"
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.107347Z",
          "start_time": "2019-09-23T20:47:42.931667Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "plt.plot(parameter_values, avg_scores, '-o')",
      "execution_count": 63,
      "outputs": [
        {
          "data": {
            "text/plain": "[<matplotlib.lines.Line2D at 0x1c0b91d2e10>]"
          },
          "execution_count": 63,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": "<Figure size 432x288 with 1 Axes>"
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.112227Z",
          "start_time": "2019-09-23T20:47:43.109299Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "X_broken = np.array(X)",
      "execution_count": 64,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.168835Z",
          "start_time": "2019-09-23T20:47:43.114178Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "X_broken[:,::2] /= 10",
      "execution_count": 65,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.226420Z",
          "start_time": "2019-09-23T20:47:43.170787Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "estimator = KNeighborsClassifier()\noriginal_scores = cross_val_score(estimator, X, y,\n  scoring='accuracy')\nprint(\"The original average accuracy for is {0:.1f}%\".format(np.mean(original_scores) * 100))\nbroken_scores = cross_val_score(estimator, X_broken, y,\n  scoring='accuracy')\nprint(\"The 'broken' average accuracy for is {0:.1f}%\".format(np.mean(broken_scores) * 100))\n",
      "execution_count": 66,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "The original average accuracy for is 82.3%\nThe 'broken' average accuracy for is 71.5%\n"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.231298Z",
          "start_time": "2019-09-23T20:47:43.227394Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "from sklearn.preprocessing import MinMaxScaler",
      "execution_count": 67,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.240082Z",
          "start_time": "2019-09-23T20:47:43.232274Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "X_transformed = MinMaxScaler().fit_transform(X)",
      "execution_count": 68,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.303523Z",
          "start_time": "2019-09-23T20:47:43.242034Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "X_transformed = MinMaxScaler().fit_transform(X_broken)\nestimator = KNeighborsClassifier()\ntransformed_scores = cross_val_score(estimator, X_transformed, y, \n  scoring='accuracy')\nprint(\"The average accuracy for is {0:.1f}%\".format(np.mean(transformed_scores) * 100))\n",
      "execution_count": 69,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "The average accuracy for is 82.3%\n"
        }
      ]
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.308403Z",
          "start_time": "2019-09-23T20:47:43.304499Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "from sklearn.pipeline import Pipeline",
      "execution_count": 70,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.317187Z",
          "start_time": "2019-09-23T20:47:43.309378Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "scaling_pipeline = Pipeline([('scale', MinMaxScaler()),\n                             ('predict', KNeighborsClassifier())])\n",
      "execution_count": 71,
      "outputs": []
    },
    {
      "metadata": {
        "ExecuteTime": {
          "end_time": "2019-09-23T20:47:43.347443Z",
          "start_time": "2019-09-23T20:47:43.318163Z"
        },
        "trusted": true
      },
      "cell_type": "code",
      "source": "scores = cross_val_score(scaling_pipeline, X_broken, y, scoring='accuracy')\nprint(\"The pipeline scored an average accuracy for is {0:.1f}%\".format(np.mean(transformed_scores) * 100))\n",
      "execution_count": 72,
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": "The pipeline scored an average accuracy for is 82.3%\n"
        }
      ]
    },
    {
      "metadata": {
        "trusted": true
      },
      "cell_type": "code",
      "source": "",
      "execution_count": null,
      "outputs": []
    }
  ],
  "metadata": {
    "anaconda-cloud": {},
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3",
      "language": "python"
    },
    "language_info": {
      "name": "python",
      "version": "3.7.3",
      "mimetype": "text/x-python",
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "pygments_lexer": "ipython3",
      "nbconvert_exporter": "python",
      "file_extension": ".py"
    },
    "latex_envs": {
      "eqNumInitial": 1,
      "eqLabelWithNumbers": true,
      "current_citInitial": 1,
      "cite_by": "apalike",
      "bibliofile": "biblio.bib",
      "LaTeX_envs_menu_present": true,
      "labels_anchors": false,
      "latex_user_defs": false,
      "user_envs_cfg": false,
      "report_style_numbering": false,
      "autoclose": false,
      "autocomplete": true,
      "hotkeys": {
        "equation": "Ctrl-E",
        "itemize": "Ctrl-I"
      }
    },
    "nbTranslate": {
      "hotkey": "alt-t",
      "sourceLang": "en",
      "targetLang": "fr",
      "displayLangs": [
        "*"
      ],
      "langInMainMenu": true,
      "useGoogleTranslate": true
    },
    "toc": {
      "nav_menu": {
        "width": "803px",
        "height": "116px"
      },
      "number_sections": true,
      "sideBar": true,
      "skip_h1_title": false,
      "base_numbering": 1,
      "title_cell": "Table of Contents",
      "title_sidebar": "Contents",
      "toc_cell": false,
      "toc_position": {},
      "toc_section_display": true,
      "toc_window_display": false
    },
    "gist": {
      "id": "adfd14ad2a34001e7c006b4d3c29911c",
      "data": {
        "description": "Ionosphere_Nearest_Neighbour.ipynb",
        "public": true
      }
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "version_major": 2,
        "version_minor": 0,
        "state": {}
      }
    },
    "_draft": {
      "nbviewer_url": "https://gist.github.com/adfd14ad2a34001e7c006b4d3c29911c"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 1
}