DSACAN_iris.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "DSACAN_iris.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "TPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/imji0319/PDSH/blob/master/DSACAN_iris.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "metadata": {
        "id": "Q5CrQsSZ0GQG",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# DBSCAN iris 데이터 세트에 적용하기"
      ]
    },
    {
      "metadata": {
        "id": "d0_7sh0mzjVE",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from sklearn.datasets import load_iris\n",
        "from sklearn.cluster import KMeans\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "import pandas as pd \n",
        "%matplotlib inline\n",
        "\n",
        "feature_names = ['sepal_length', 'sepal_width' ,'petal_length', 'petal_width']\n",
        "iris = load_iris()\n",
        "\n",
        "irisDF = pd.DataFrame(data = iris.data , columns = feature_names)\n",
        "irisDF['target'] = iris.target"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "28uAHt5v0PeM",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        },
        "outputId": "8d5a4208-be13-4f15-a945-4b4857108ed5"
      },
      "cell_type": "code",
      "source": [
        "from sklearn.cluster import DBSCAN\n",
        "\n",
        "dbscan = DBSCAN(eps = 0.6, min_samples = 8, metric = 'euclidean')\n",
        "dbscan_labels = dbscan.fit_predict(iris.data)\n",
        "\n",
        "irisDF['dbscan_cluster'] = dbscan_labels\n",
        "\n",
        "iris_result = irisDF.groupby(['target'])['dbscan_cluster'].value_counts()\n",
        "print(iris_result)"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "target  dbscan_cluster\n",
            "0        0                49\n",
            "        -1                 1\n",
            "1        1                46\n",
            "        -1                 4\n",
            "2        1                42\n",
            "        -1                 8\n",
            "Name: dbscan_cluster, dtype: int64\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "UHd8o3x27w9Q",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "2b88b537-1a2c-43eb-c420-628105a83abc"
      },
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/gdrive; to attempt to forcibly remount, call drive.mount(\"/content/gdrive\", force_remount=True).\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "FyCBP6UF8tAT",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import os\n",
        "import sys\n",
        "sys.path.insert(0, '/content/gdrive/My Drive/Colab Notebooks/')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "X8MV-vnW05wC",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 269
        },
        "outputId": "7c8767a8-141f-446e-8dab-53c73e35317b"
      },
      "cell_type": "code",
      "source": [
        "from sklearn.decomposition import PCA\n",
        "from importme import vsp\n",
        "\n",
        "pca = PCA(n_components=2, random_state=0)\n",
        "pca_transformed = pca.fit_transform(iris.data)\n",
        "\n",
        "irisDF['ftr1']  = pca_transformed[:, 0]\n",
        "irisDF['ftr2'] = pca_transformed[:, 1 ]\n",
        "\n",
        "vsp.visualize_cluster_plot(dbscan, irisDF, 'dbscan_cluster', iscenter= False)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAD8CAYAAACfF6SlAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}