Sampling and Model Interpretability Project.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "mlOEzxLwuORm"
   },
   "source": [
    "# Assignment 4: Sampling and Model Interpretability\n",
    "-----------------------------------\n",
    "\n",
    "**Dr. Dave Wanik - OPIM 5512: Data Science Using Python - University of Connecticut**\n",
    "\n",
    "-----------------------------------\n",
    "\n",
    "\n",
    "**Your Name Here:** Adam Syed\n",
    "\n",
    "Put your sampling and interpretability skills to the test!\n",
    "\n",
    "You will build a classification model using a sampling technique of your choice (oversampling, undersampling or SMOTE - I'm not testing how good your model is for this assignment, just that you know how to interpret the model!). We will use CA housing since you have some experience with this from tree-based models. You can use any classification model we have covered in class (DTC, RFC, GBC or logistic regression.)\n",
    "\n",
    "# Rubric\n",
    "* (10 points) I've already made your data 'imbalanced' with the 75th percentile. Split the raw data into train and test partition with an 80/20 split. Then use a sampling technique (oversampling, undersampling, SMOTE) on the train data. Important: do not change the class imbalance in the test partition! Count values of each class before and after your sampling to confirm your understanding of how the sampling worked. Your new train data should be more or less balanced, while the test data is still imbalanced. This allows for fair evaluation of the model.\n",
    "\n",
    "Then, using your fitted model, you will:\n",
    "* (40 points) Print the top 3 features (using permutation importance with 15 repeats) as a boxplot (like we do in class.)\n",
    "* (40 points) Create the partial dependence plots for the top 3 features (you may opt to use the `num_grid_points=10` argument if your code is taking a long time to run). Customize the plots so that the Y axis is consistent on each of the three plots and remember that the Y axis is showing the change in predicted probability.\n",
    "* (10 points) Write five meaningful bullets about what you see in the plots and tables. Does anything surprise you?\n",
    "  * **Optional:** Potential with issues concerning correlated predictors (read here for a cool way on how to address this: https://scikit-learn.org/stable/auto_examples/inspection/plot_permutation_importance_multicollinear.html)?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 206
    },
    "id": "qjiMm6XivpGh",
    "outputId": "35ff66fa-15c3-4c06-bf7b-088d29c8f817"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>housing_median_age</th>\n",
       "      <th>total_rooms</th>\n",
       "      <th>total_bedrooms</th>\n",
       "      <th>population</th>\n",
       "      <th>households</th>\n",
       "      <th>median_income</th>\n",
       "      <th>median_house_value</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-114.31</td>\n",
       "      <td>34.19</td>\n",
       "      <td>15.0</td>\n",
       "      <td>5612.0</td>\n",
       "      <td>1283.0</td>\n",
       "      <td>1015.0</td>\n",
       "      <td>472.0</td>\n",
       "      <td>1.4936</td>\n",
       "      <td>66900.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-114.47</td>\n",
       "      <td>34.40</td>\n",
       "      <td>19.0</td>\n",
       "      <td>7650.0</td>\n",
       "      <td>1901.0</td>\n",
       "      <td>1129.0</td>\n",
       "      <td>463.0</td>\n",
       "      <td>1.8200</td>\n",
       "      <td>80100.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-114.56</td>\n",
       "      <td>33.69</td>\n",
       "      <td>17.0</td>\n",
       "      <td>720.0</td>\n",
       "      <td>174.0</td>\n",
       "      <td>333.0</td>\n",
       "      <td>117.0</td>\n",
       "      <td>1.6509</td>\n",
       "      <td>85700.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-114.57</td>\n",
       "      <td>33.64</td>\n",
       "      <td>14.0</td>\n",
       "      <td>1501.0</td>\n",
       "      <td>337.0</td>\n",
       "      <td>515.0</td>\n",
       "      <td>226.0</td>\n",
       "      <td>3.1917</td>\n",
       "      <td>73400.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-114.57</td>\n",
       "      <td>33.57</td>\n",
       "      <td>20.0</td>\n",
       "      <td>1454.0</td>\n",
       "      <td>326.0</td>\n",
       "      <td>624.0</td>\n",
       "      <td>262.0</td>\n",
       "      <td>1.9250</td>\n",
       "      <td>65500.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   longitude  latitude  housing_median_age  total_rooms  total_bedrooms  \\\n",
       "0    -114.31     34.19                15.0       5612.0          1283.0   \n",
       "1    -114.47     34.40                19.0       7650.0          1901.0   \n",
       "2    -114.56     33.69                17.0        720.0           174.0   \n",
       "3    -114.57     33.64                14.0       1501.0           337.0   \n",
       "4    -114.57     33.57                20.0       1454.0           326.0   \n",
       "\n",
       "   population  households  median_income  median_house_value  \n",
       "0      1015.0       472.0         1.4936             66900.0  \n",
       "1      1129.0       463.0         1.8200             80100.0  \n",
       "2       333.0       117.0         1.6509             85700.0  \n",
       "3       515.0       226.0         3.1917             73400.0  \n",
       "4       624.0       262.0         1.9250             65500.0  "
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "df = pd.read_csv('/Users/adamyeeter/OPIM 5512 Python/Data/california_housing_train.csv')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "MSklVwv794tE"
   },
   "source": [
    "And we will recode the target variable to a classification problem."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "LvT61xo097ST",
    "outputId": "e97e2fbd-f4da-43a0-d9d8-6b485367e434"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    12753\n",
       "1     4247\n",
       "Name: median_house_value, dtype: int64"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['median_house_value'] = np.where(df['median_house_value'] > np.percentile(df['median_house_value'], q=75), 1, 0)\n",
    "df['median_house_value'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "6Zp3Z4IRAbvi"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    10206\n",
       "1     3394\n",
       "Name: median_house_value, dtype: int64"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "y = df['median_house_value']\n",
    "X = df.drop('median_house_value', axis=1)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y,\n",
    "                                                    test_size = 0.2,\n",
    "                                                    shuffle = True,\n",
    "                                                    random_state = 2735566)\n",
    "y_train.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "P847REMT_l24"
   },
   "source": [
    "# Sampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "id": "iACWT_64_oHm"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({0: 10206, 1: 10206})"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from imblearn.over_sampling import RandomOverSampler\n",
    "from collections import Counter\n",
    "\n",
    "oversample = RandomOverSampler(sampling_strategy='minority')\n",
    "\n",
    "X_train_over, y_train_over = oversample.fit_resample(X_train, y_train)\n",
    "\n",
    "Counter(y_train_over)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xgm__s2pAXW5"
   },
   "source": [
    "# Model and Metrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "id": "RMwCQ7W6Aees"
   },
   "outputs": [],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "\n",
    "LR = LogisticRegression(max_iter=1000, random_state=2735566)\n",
    "\n",
    "LR = LR.fit(X_train_over, y_train_over)\n",
    "\n",
    "train_preds_LR = LR.predict(X_train_over) \n",
    "test_preds_LR = LR.predict(X_test) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[8361, 1845],\n",
       "       [1869, 8337]])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix(y_train_over, train_preds_LR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[2070,  477],\n",
       "       [ 161,  692]])"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "confusion_matrix(y_test, test_preds_LR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.82      0.82      0.82     10206\n",
      "           1       0.82      0.82      0.82     10206\n",
      "\n",
      "    accuracy                           0.82     20412\n",
      "   macro avg       0.82      0.82      0.82     20412\n",
      "weighted avg       0.82      0.82      0.82     20412\n",
      "\n"
     ]
    }
   ],
   "source": [
    "trainReport_LR = classification_report(y_train_over, train_preds_LR)\n",
    "print(trainReport_LR)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       0.93      0.81      0.87      2547\n",
      "           1       0.59      0.81      0.68       853\n",
      "\n",
      "    accuracy                           0.81      3400\n",
      "   macro avg       0.76      0.81      0.78      3400\n",
      "weighted avg       0.84      0.81      0.82      3400\n",
      "\n"
     ]
    }
   ],
   "source": [
    "testReport_LR = classification_report(y_test, test_preds_LR)\n",
    "print(testReport_LR)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "j99ZHu-vAqVK"
   },
   "source": [
    "# Interpretability"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "xN63ERDfAszH"
   },
   "source": [
    "## Permutation Importance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "4TKUEyYhAsNz"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.inspection import permutation_importance\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "clf = LR\n",
    "\n",
    "result = permutation_importance(clf, X_test, y_test, n_repeats=15,\n",
    "                                random_state=2735566)\n",
    "\n",
    "perm_sorted_idx = result.importances_mean.argsort()[::-1]\n",
    "\n",
    "top_3_idx = perm_sorted_idx[:3]\n",
    "\n",
    "fig, ax1 = plt.subplots(1, 1, figsize=(6, 4))\n",
    "ax1.boxplot(result.importances[top_3_idx].T, vert=False,\n",
    "            labels=X.columns[top_3_idx])\n",
    "fig.suptitle('LR Feature Importance', y=1.05)\n",
    "fig.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "53mNce4iAwIX"
   },
   "source": [
    "## Partial Dependence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "IsxgzjxnD5ZI"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>longitude</th>\n",
       "      <th>latitude</th>\n",
       "      <th>housing_median_age</th>\n",
       "      <th>total_rooms</th>\n",
       "      <th>total_bedrooms</th>\n",
       "      <th>population</th>\n",
       "      <th>households</th>\n",
       "      <th>median_income</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-120.06</td>\n",
       "      <td>36.72</td>\n",
       "      <td>32.0</td>\n",
       "      <td>981.0</td>\n",
       "      <td>237.0</td>\n",
       "      <td>736.0</td>\n",
       "      <td>249.0</td>\n",
       "      <td>1.8000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-117.95</td>\n",
       "      <td>33.98</td>\n",
       "      <td>15.0</td>\n",
       "      <td>16042.0</td>\n",
       "      <td>2602.0</td>\n",
       "      <td>7732.0</td>\n",
       "      <td>2552.0</td>\n",
       "      <td>5.6716</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-121.94</td>\n",
       "      <td>37.32</td>\n",
       "      <td>46.0</td>\n",
       "      <td>2451.0</td>\n",
       "      <td>472.0</td>\n",
       "      <td>1163.0</td>\n",
       "      <td>448.0</td>\n",
       "      <td>4.8519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-122.68</td>\n",
       "      <td>41.15</td>\n",
       "      <td>32.0</td>\n",
       "      <td>817.0</td>\n",
       "      <td>206.0</td>\n",
       "      <td>224.0</td>\n",
       "      <td>89.0</td>\n",
       "      <td>3.6310</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-122.27</td>\n",
       "      <td>37.84</td>\n",
       "      <td>52.0</td>\n",
       "      <td>2436.0</td>\n",
       "      <td>541.0</td>\n",
       "      <td>1015.0</td>\n",
       "      <td>478.0</td>\n",
       "      <td>1.7250</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   longitude  latitude  housing_median_age  total_rooms  total_bedrooms  \\\n",
       "0    -120.06     36.72                32.0        981.0           237.0   \n",
       "1    -117.95     33.98                15.0      16042.0          2602.0   \n",
       "2    -121.94     37.32                46.0       2451.0           472.0   \n",
       "3    -122.68     41.15                32.0        817.0           206.0   \n",
       "4    -122.27     37.84                52.0       2436.0           541.0   \n",
       "\n",
       "   population  households  median_income  \n",
       "0       736.0       249.0         1.8000  \n",
       "1      7732.0      2552.0         5.6716  \n",
       "2      1163.0       448.0         4.8519  \n",
       "3       224.0        89.0         3.6310  \n",
       "4      1015.0       478.0         1.7250  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_X_df = pd.DataFrame(X_train_over, columns=X.columns)\n",
    "train_X_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/adamyeeter/anaconda3/lib/python3.11/site-packages/sklearn/base.py:464: UserWarning: X does not have valid feature names, but LogisticRegression was fitted with feature names\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(10, 13600)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from pycebox.ice import ice, ice_plot\n",
    "clfdf = ice(data=train_X_df,\n",
    "            column='median_income',\n",
    "                   predict=clf.predict,\n",
    "                    num_grid_points=10)\n",
    "np.shape(clfdf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ice_plot(clfdf, c='dimgray', linewidth=0.3,\n",
    "                  plot_pdp=True, \n",
    "         pdp_kwargs={'linewidth': 5, 'color':'red'})\n",
    "plt.title('PDP: median_income')\n",
    "plt.ylabel('Predicted median_house_value')\n",
    "plt.xlabel('median_income');\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/adamyeeter/anaconda3/lib/python3.11/site-packages/sklearn/base.py:464: UserWarning: X does not have valid feature names, but LogisticRegression was fitted with feature names\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pycebox.ice import ice, ice_plot\n",
    "clfdf = ice(data=train_X_df,\n",
    "            column='households',\n",
    "                   predict=clf.predict,\n",
    "                    num_grid_points=10)\n",
    "\n",
    "ice_plot(clfdf, c='dimgray', linewidth=0.3,\n",
    "                  plot_pdp=True, \n",
    "         pdp_kwargs={'linewidth': 5, 'color':'red'})\n",
    "plt.title('PDP: households')\n",
    "plt.ylabel('Predicted median_house_value')\n",
    "plt.xlabel('households');\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/adamyeeter/anaconda3/lib/python3.11/site-packages/sklearn/base.py:464: UserWarning: X does not have valid feature names, but LogisticRegression was fitted with feature names\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pycebox.ice import ice, ice_plot\n",
    "clfdf = ice(data=train_X_df,\n",
    "            column='population',\n",
    "                   predict=clf.predict,\n",
    "                    num_grid_points=10)\n",
    "\n",
    "ice_plot(clfdf, c='dimgray', linewidth=0.3,\n",
    "                  plot_pdp=True, \n",
    "         pdp_kwargs={'linewidth': 5, 'color':'red'})\n",
    "plt.title('PDP: population')\n",
    "plt.ylabel('Predicted median_house_value')\n",
    "plt.xlabel('population');\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "8l5c2RQvAygR"
   },
   "source": [
    "# Conclusions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jUJQLeFXAhyk"
   },
   "source": [
    "1. In the permutation importance boxplots, I see that the three most important features are median_income, households, and population.I was not surprised to see median_income as the top feature overall, however I am surprised to see population in the top 3. \n",
    "\n",
    "2. As population increases, the predicted median_house_value decreases sharply at first and then decreases gradually. I was able to see this through the partial dependence plot.\n",
    "\n",
    "3. As median_income increases, the predicted median_house_value increases and then starts to plateau at a predicted median_house_value of 1 and a median_income value of 6. This is not a surpise to me as these two variables seem to have an obvious correlation.\n",
    "\n",
    "4. As the households feature increases, the predicted median_house_value increases sharply at first and then increases gradually.\n",
    "\n",
    "5. In the classification report for the test set, the precision was very high (0.93) when predicting the median_house_value below the median. However, the precision was low (0.59) when predicting the median_house_value above the median."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}