machine_learning_04.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyNLh29SrmQ0MaWDgSGaKAjU",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/matsunagalab/lecture_ML/blob/main/machine_learning_04.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# 第4回 回帰その2：過学習と正則化\n"
      ],
      "metadata": {
        "id": "Y3pXof4n2o2j"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 多項式の回帰"
      ],
      "metadata": {
        "id": "u3R-uajwoPFP"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "授業中に出てきた多項式の回帰を行います"
      ],
      "metadata": {
        "id": "KWX3MGT2o-lu"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.linear_model import LinearRegression\n",
        "np.random.seed(seed=321)\n",
        "\n",
        "# テストデータの作成\n",
        "N = 50\n",
        "x = 2.0 * (np.random.rand(N) - 0.5)\n",
        "x = np.sort(x)\n",
        "X = np.column_stack([x**i for i in range(1, 14)])\n",
        "w_true = np.array([0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0])\n",
        "y = X @ w_true + 0.1 * np.random.randn(N)\n",
        "\n",
        "# データの標準化\n",
        "scaler_y = StandardScaler()\n",
        "y_scaled = scaler_y.fit_transform(y.reshape(-1, 1)).ravel()\n",
        "scaler_X = StandardScaler()\n",
        "X_scaled = scaler_X.fit_transform(X)\n",
        "\n",
        "# 擬似逆行列を使った重回帰の計算\n",
        "reg = LinearRegression(fit_intercept=False)\n",
        "reg.fit(X_scaled, y_scaled)\n",
        "w = reg.coef_\n",
        "print(w)\n",
        "\n",
        "# グラフの描画\n",
        "plt.scatter(X_scaled[:, 0], y_scaled, label=\"data\", marker='o', s=64)\n",
        "plt.plot(X_scaled[:, 0], (X[:, 2] - scaler_X.mean_[2]) / scaler_X.scale_[2], lw=2, label=\"ground-truth\")\n",
        "plt.plot(X_scaled[:, 0], X_scaled @ w, lw=2, label=\"regression\")\n",
        "plt.xlabel(\"x_1\")\n",
        "plt.ylabel(\"y\")\n",
        "plt.legend(loc=\"upper left\")\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "4-wjGsDPNF9I",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 520
        },
        "outputId": "7d17aa4e-95cb-4e30-fc44-6ec6f8401774"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[ 3.03594565e-01 -1.60656949e+00 -1.19809525e+01  1.69064628e+01\n",
            "  8.20583403e+01 -6.64150752e+01 -2.25940054e+02  1.21630042e+02\n",
            "  3.22326313e+02 -1.04797603e+02 -2.34438671e+02  3.42085681e+01\n",
            "  6.86640950e+01]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 交差検証"
      ],
      "metadata": {
        "id": "YBAU2YkWoTPG"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "交差検証を行います。K=10として様々な次数のモデルの交差検証エラーを求めます"
      ],
      "metadata": {
        "id": "Rapa2xiIpDLN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.linear_model import LinearRegression\n",
        "from sklearn.model_selection import KFold, cross_val_score\n",
        "\n",
        "def compute_cross_validation_error(k, X, y, solver=LinearRegression(fit_intercept=False)):\n",
        "    kf = KFold(n_splits=k, shuffle=True, random_state=42)\n",
        "    reg = solver\n",
        "    cv_error = -cross_val_score(reg, X, y, cv=kf, scoring='neg_mean_squared_error')\n",
        "    return np.mean(cv_error)\n",
        "\n",
        "# 交差検証エラーを求める\n",
        "k = 10\n",
        "print(compute_cross_validation_error(k, X, y))\n",
        "\n",
        "# 様々な次数のモデルの交差検証エラーを求める\n",
        "K = 10\n",
        "nth_order_max = 10\n",
        "cv_error = np.zeros(nth_order_max)\n",
        "for n in range(nth_order_max):\n",
        "    cv_error[n] = compute_cross_validation_error(K, X_scaled[:, :(n+1)], y_scaled)\n",
        "\n",
        "plt.plot(range(1, nth_order_max+1), cv_error, lw=2)\n",
        "plt.yscale(\"log\")\n",
        "plt.xlabel(\"polynomial order\")\n",
        "plt.ylabel(\"cross validation error\")\n",
        "plt.show()\n"
      ],
      "metadata": {
        "id": "hQyfJPEio7so",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 467
        },
        "outputId": "9cf7e865-6cba-4ba3-815f-f99eaf8c6dad"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.170889173396334\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Ridge回帰"
      ],
      "metadata": {
        "id": "oiwxE7fkoVsY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.linear_model import Ridge, RidgeCV\n",
        "from sklearn.model_selection import cross_val_score\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "# RidgeCVを用いて交差検証を行い最適なlambdaを求める\n",
        "alphas = np.logspace(-5, 3, 30)\n",
        "ridge_cv = RidgeCV(alphas=alphas, fit_intercept=False, cv=10)\n",
        "ridge_cv.fit(X_scaled, y_scaled)\n",
        "\n",
        "# 最適なlambdaのRidge回帰を求める\n",
        "best_alpha = ridge_cv.alpha_\n",
        "ridge_reg = Ridge(alpha=best_alpha, fit_intercept=False)\n",
        "ridge_reg.fit(X_scaled, y_scaled)\n",
        "print(ridge_reg.coef_)\n",
        "\n",
        "# 交差検証エラーを求める\n",
        "cv_errors = []\n",
        "for alpha in alphas:\n",
        "    ridge = Ridge(alpha=alpha, fit_intercept=False)\n",
        "    cv_error = -cross_val_score(ridge, X_scaled, y_scaled, cv=10, scoring='neg_mean_squared_error').mean()\n",
        "    cv_errors.append(cv_error)\n",
        "\n",
        "# Plot 1: 交差検証エラーのプロット\n",
        "plt.plot(alphas, cv_errors, lw=2)\n",
        "plt.xscale(\"log\")\n",
        "plt.xlabel(\"lambda\")\n",
        "plt.ylabel(\"cross validation error\")\n",
        "plt.show()\n",
        "\n",
        "# Plot 2: データの散布図と、モデルの予測をラインで重ねてプロット\n",
        "plt.scatter(X_scaled[:, 0], y_scaled, label=\"data\", marker='o', s=64)\n",
        "plt.plot(X_scaled[:, 0], (X[:, 2] - scaler_X.mean_[2]) / scaler_X.scale_[2], lw=2, label=\"ground-truth\")\n",
        "plt.plot(X_scaled[:, 0], X_scaled @ ridge_reg.coef_, lw=2, label=\"ridge\")\n",
        "plt.xlabel(\"x\")\n",
        "plt.ylabel(\"y\")\n",
        "plt.legend(loc=\"upper left\")\n",
        "plt.show()\n"
      ],
      "metadata": {
        "id": "3DjmzW4EtwNc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 940
        },
        "outputId": "0ca30fa8-90de-462e-d9e6-9106b6c58d8d"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[-0.40546466  0.02557338  1.6087629  -0.26564249  0.85268646  0.3678646\n",
            " -0.79944817  0.28183708 -1.08560435 -0.21491592 -0.29034448 -0.27902174\n",
            "  1.15406269]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Lasso回帰"
      ],
      "metadata": {
        "id": "QTiNE4fGDKzk"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.linear_model import Lasso, LassoCV\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "# LassoCVを用いて交差検証を行い最適なlambdaを求める\n",
        "alphas = np.logspace(-5, 0, 30)\n",
        "lasso_cv = LassoCV(alphas=alphas, fit_intercept=False, max_iter=10000, tol=1e-5, cv=10, random_state=42)\n",
        "lasso_cv.fit(X_scaled, y_scaled)\n",
        "\n",
        "# 最適なlambdaのLasso回帰を求める\n",
        "best_alpha = lasso_cv.alpha_\n",
        "lasso_reg = Lasso(alpha=best_alpha, fit_intercept=False, max_iter=10000, tol=1e-5)\n",
        "lasso_reg.fit(X_scaled, y_scaled)\n",
        "print(best_alpha)\n",
        "print(lasso_reg.coef_)\n",
        "\n",
        "# Plot 1: 交差検証エラーのプロット\n",
        "plt.plot(lasso_cv.alphas_, lasso_cv.mse_path_.mean(axis=-1), lw=2)\n",
        "plt.xscale(\"log\")\n",
        "plt.xlabel(\"lambda\")\n",
        "plt.ylabel(\"cross validation error\")\n",
        "plt.show()\n",
        "\n",
        "# Plot 2: 様々なalphaでのいくつかの偏回帰係数のプロット\n",
        "alphas_path, coef_path, _ = lasso_cv.path(X_scaled, y_scaled)\n",
        "coef_path = coef_path.T\n",
        "for i in range(10):\n",
        "    plt.plot(alphas_path, coef_path[:, i], label=f\"{i+1}st order\")\n",
        "plt.xscale(\"log\")\n",
        "plt.xlabel(\"lambda\")\n",
        "plt.ylabel(\"coefficients\")\n",
        "plt.legend(loc=\"upper right\", bbox_to_anchor=(1.30, 1))\n",
        "plt.show()\n",
        "\n",
        "# Plot 3: データの散布図と、モデルの予測をラインで重ねてプロット\n",
        "plt.scatter(X_scaled[:, 0], y_scaled, label=\"data\", marker='o', s=64)\n",
        "plt.plot(X_scaled[:, 0], (X[:, 2] - scaler_X.mean_[2]) / scaler_X.scale_[2], lw=2, label=\"ground-truth\")\n",
        "plt.plot(X_scaled[:, 0], X_scaled @ lasso_reg.coef_, lw=2, label=\"lasso\")\n",
        "plt.xlabel(\"x\")\n",
        "plt.ylabel(\"y\")\n",
        "plt.legend(loc=\"upper left\")\n",
        "plt.show()\n"
      ],
      "metadata": {
        "id": "jQHL-YJUjMg3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "81f0fde1-78a2-4087-cbdd-f5a387d5620e"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0016619826411972127, tolerance: 0.0003593416355522188\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.002534561717457695, tolerance: 0.0003593416355522188\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.001339180555446795, tolerance: 0.0003593416355522188\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.00994147894332853, tolerance: 0.0003593416355522188\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.09350769509759616, tolerance: 0.0003593416355522188\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.150669771776212, tolerance: 0.0003593416355522188\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.19128135417048275, tolerance: 0.0003593416355522188\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.22268134068928735, tolerance: 0.0003593416355522188\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.243876752115727, tolerance: 0.0003593416355522188\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0008104770820471785, tolerance: 0.000422562000465476\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.003309116450993521, tolerance: 0.000422562000465476\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.009644203531826712, tolerance: 0.000422562000465476\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.006871864524804927, tolerance: 0.000422562000465476\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.014846686477113291, tolerance: 0.000422562000465476\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.07551948969341238, tolerance: 0.000422562000465476\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.1549868748517047, tolerance: 0.000422562000465476\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.26210374792959157, tolerance: 0.000422562000465476\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.4612733900877459, tolerance: 0.000422562000465476\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0019930072094451745, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0016364023632315927, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.021390325648149577, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.039411235602738515, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.05158771616694002, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.05523802953394141, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0825682699367154, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.12108374054884541, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.17134209223746333, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.38292904683366413, tolerance: 0.0004779939192530614\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0018070933245795118, tolerance: 0.0004951820055583039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.003979553747754494, tolerance: 0.0004951820055583039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.006593518679560262, tolerance: 0.0004951820055583039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0387878089988174, tolerance: 0.0004951820055583039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.14125544715958838, tolerance: 0.0004951820055583039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.10592495868907248, tolerance: 0.0004951820055583039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.28865159015819786, tolerance: 0.0004951820055583039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.43656892805675773, tolerance: 0.0004951820055583039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0007121990502421127, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0032681723984628874, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.01169049088569274, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.027387678065164067, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.03680682922808787, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.05238980930044512, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.09906867635457695, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.16613539371227337, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.2577910214340189, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.367828424707773, tolerance: 0.0004971927387034728\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0005302961754285462, tolerance: 0.0004933838935571039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0026655655662572997, tolerance: 0.0004933838935571039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.004662785737781228, tolerance: 0.0004933838935571039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0069938793033177404, tolerance: 0.0004933838935571039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.009953520720263231, tolerance: 0.0004933838935571039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.013643145888422303, tolerance: 0.0004933838935571039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.01785752705621424, tolerance: 0.0004933838935571039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.024284961828317186, tolerance: 0.0004933838935571039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.03190982623139149, tolerance: 0.0004933838935571039\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.001459551377255508, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0032532662766273646, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.011581146985811586, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.08430071539766226, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.03857032907998659, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.1978179585378954, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.2376827357107434, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.4742067185352212, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.67402869040128, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.8121756747595803, tolerance: 0.0004986167174770334\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.001966874000109975, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.006334968278492781, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.007239547988145301, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.03298237973981344, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.06070011718738311, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.11444767856846072, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.18213938524969842, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.262672310737722, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.3673015048781616, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.5024381834050422, tolerance: 0.0004844827115266013\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.00488320852690638, tolerance: 0.0004206110409397965\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.010367266955309873, tolerance: 0.0004206110409397965\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.015517239912091974, tolerance: 0.0004206110409397965\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.02046120586118505, tolerance: 0.0004206110409397965\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.06418380535098711, tolerance: 0.0004206110409397965\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.12541940142588137, tolerance: 0.0004206110409397965\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.20785266319127915, tolerance: 0.0004206110409397965\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.3223292188823792, tolerance: 0.0004206110409397965\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.45982239433141103, tolerance: 0.0004206110409397965\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.0010379126300286146, tolerance: 0.0003506333369669313\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.027265806953739258, tolerance: 0.0003506333369669313\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.09606608546380313, tolerance: 0.0003506333369669313\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.18108719608677548, tolerance: 0.0003506333369669313\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.29489691814183283, tolerance: 0.0003506333369669313\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n",
            "/usr/local/lib/python3.10/dist-packages/sklearn/linear_model/_coordinate_descent.py:617: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Duality gap: 0.4784735190222795, tolerance: 0.0003506333369669313\n",
            "  model = cd_fast.enet_coordinate_descent_gram(\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0.018873918221350976\n",
            "[ 0.         -0.          0.95537182 -0.          0.         -0.\n",
            "  0.         -0.          0.         -0.         -0.         -0.02819152\n",
            " -0.        ]\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## California housingへの応用"
      ],
      "metadata": {
        "id": "y16U0_kFEXgG"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "このデータセットは、米国カリフォルニア州の各ブロック地区における住宅価格データです。データセットには20640の観測値があり、説明変数(特徴量)xが8つと、目的変数yとして住宅価格が含まれています。\n",
        "\n",
        "具体的な特徴量は以下です：\n",
        "* MedInc：ブロックの中央所得\n",
        "* HouseAge：ブロックの家の中央年齢\n",
        "* AveRooms：世帯あたりの平均部屋数\n",
        "* AveBedrms：世帯あたりの平均寝室数\n",
        "* Population：ブロックの人口\n",
        "* AveOccup：世帯あたりの平均家族人数\n",
        "* Latitude：ブロックの緯度\n",
        "* Longitude：ブロックの経度\n",
        "\n",
        "このデータに対しLasso回帰を行って結果をプロットします。\n",
        "\n"
      ],
      "metadata": {
        "id": "9Mnbd3OwEk_w"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn.datasets import fetch_california_housing\n",
        "from sklearn.linear_model import Lasso, LassoCV\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "\n",
        "# California housing dataを読み込む\n",
        "data = fetch_california_housing()\n",
        "X, y = data.data, data.target\n",
        "\n",
        "# データを標準化する\n",
        "scaler_y = StandardScaler()\n",
        "scaler_X = StandardScaler()\n",
        "y_scaled = scaler_y.fit_transform(y.reshape(-1, 1)).ravel()\n",
        "X_scaled = scaler_X.fit_transform(X)\n",
        "\n",
        "# 最適なalpha (lambda)を見つけるための交差検証\n",
        "alphas = np.logspace(-5, 0, 30)\n",
        "lasso_cv = LassoCV(alphas=alphas, cv=10)\n",
        "lasso_cv.fit(X_scaled, y_scaled)\n",
        "\n",
        "# 交差検証エラーのプロット\n",
        "plt.plot(lasso_cv.alphas_, lasso_cv.mse_path_.mean(axis=-1), lw=2)\n",
        "plt.xscale(\"log\")\n",
        "plt.xlabel(\"lambda\")\n",
        "plt.ylabel(\"cross validation error\")\n",
        "plt.show()\n",
        "\n",
        "# 様々なalphaでのいくつかの偏回帰係数のプロット\n",
        "alphas_path, coef_path, _ = Lasso().path(X_scaled, y_scaled, alphas=alphas)\n",
        "coef_path = coef_path.T\n",
        "#plt.figure(figsize=(10, 6))\n",
        "ax = plt.gca()\n",
        "lines = ax.semilogx(alphas_path, coef_path)\n",
        "plt.xlabel(\"lambda\")\n",
        "plt.ylabel(\"coefficients\")\n",
        "ax.legend(lines, data.feature_names, loc='center left', bbox_to_anchor=(1, 0.5))\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "x1-nGET4D73-",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 891
        },
        "outputId": "20b9011f-0ded-4a9d-9ac0-a6be53d5e553"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "yT6wFISluaLa"
      },
      "execution_count": 5,
      "outputs": []
    }
  ]
}