chexpert_resnet18.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import os\n",
    "from importlib import reload\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib\n",
    "import seaborn as sns\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "import skimage\n",
    "from skimage import io\n",
    "from skimage.transform import resize\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "from torch.optim import lr_scheduler\n",
    "import numpy as np\n",
    "import torchvision\n",
    "from torchvision import datasets, models, transforms\n",
    "import time\n",
    "import os\n",
    "import copy\n",
    "\n",
    "# local imports\n",
    "import model as _model\n",
    "import utils as _tools"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sampling 50000 records\n",
      "train.shape: (37500, 19)\n",
      "dev.shape: (12500, 19)\n",
      "valid.shape: (234, 19)\n",
      "saved: ./train.csv\n",
      "saved: ./dev.csv\n",
      "saved: ./valid.csv\n",
      "\n",
      "Trainning label distribution\n",
      " NaN    0.792187\n",
      " 1.0    0.121787\n",
      " 0.0    0.049280\n",
      "-1.0    0.036747\n",
      "Name: Cardiomegaly, dtype: float64\n",
      "\n",
      "Development label distribution\n",
      " NaN    0.79312\n",
      " 1.0    0.12280\n",
      " 0.0    0.04952\n",
      "-1.0    0.03456\n",
      "Name: Cardiomegaly, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "control the generation of trainning and dev datasets\n",
    "\"\"\"\n",
    "\n",
    "train, dev, _ = _tools.make_trainning_data(sample=50000, \n",
    "                                           return_frames=True, \n",
    "                                           state=1729)\n",
    "\n",
    "print()\n",
    "print(\"Trainning label distribution\")\n",
    "print(train['Cardiomegaly'].value_counts(normalize=True, dropna=False))\n",
    "\n",
    "print()\n",
    "print(\"Development label distribution\")\n",
    "print(dev['Cardiomegaly'].value_counts(normalize=True, dropna=False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 1/20\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/sw/arcts/centos7/python3.8-anaconda/2020.07/lib/python3.8/site-packages/sklearn/metrics/_classification.py:1221: UndefinedMetricWarning: Recall is ill-defined and being set to 0.0 due to no true samples. Use `zero_division` parameter to control this behavior.\n",
      "  _warn_prf(average, modifier, msg_start, len(result))\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Trainning loss: 0.6191 accuracy: 82.78 %\n",
      "Validation loss: 0.5925 accuracy: 84.79 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 2/20\n",
      "Trainning loss: 0.5727 accuracy: 83.39 %\n",
      "Validation loss: 0.5686 accuracy: 85.43 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 3/20\n",
      "Trainning loss: 0.5421 accuracy: 83.24 %\n",
      "Validation loss: 0.5456 accuracy: 81.68 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 4/20\n",
      "Trainning loss: 0.5223 accuracy: 83.50 %\n",
      "Validation loss: 0.5323 accuracy: 82.06 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 5/20\n",
      "Trainning loss: 0.5019 accuracy: 83.61 %\n",
      "Validation loss: 0.5260 accuracy: 83.24 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 6/20\n",
      "Trainning loss: 0.4867 accuracy: 83.92 %\n",
      "Validation loss: 0.5187 accuracy: 82.23 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 7/20\n",
      "Trainning loss: 0.4731 accuracy: 84.61 %\n",
      "Validation loss: 0.5158 accuracy: 82.43 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 8/20\n",
      "Trainning loss: 0.4589 accuracy: 84.91 %\n",
      "Validation loss: 0.5140 accuracy: 83.67 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 9/20\n",
      "Trainning loss: 0.4429 accuracy: 85.58 %\n",
      "Validation loss: 0.5136 accuracy: 82.93 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 10/20\n",
      "Trainning loss: 0.4297 accuracy: 85.90 %\n",
      "Validation loss: 0.5178 accuracy: 83.74 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 11/20\n",
      "Trainning loss: 0.4141 accuracy: 86.46 %\n",
      "Validation loss: 0.5183 accuracy: 83.86 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 12/20\n",
      "Trainning loss: 0.3991 accuracy: 86.94 %\n",
      "Validation loss: 0.5203 accuracy: 83.54 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 13/20\n",
      "Trainning loss: 0.3811 accuracy: 87.56 %\n",
      "Validation loss: 0.5327 accuracy: 84.43 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 14/20\n",
      "Trainning loss: 0.3647 accuracy: 88.19 %\n",
      "Validation loss: 0.5281 accuracy: 82.74 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 15/20\n",
      "Trainning loss: 0.3473 accuracy: 88.62 %\n",
      "Validation loss: 0.5372 accuracy: 83.42 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 16/20\n",
      "Trainning loss: 0.3291 accuracy: 89.34 %\n",
      "Validation loss: 0.5580 accuracy: 84.58 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 17/20\n",
      "Trainning loss: 0.3094 accuracy: 90.06 %\n",
      "Validation loss: 0.5553 accuracy: 82.10 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 18/20\n",
      "Trainning loss: 0.2882 accuracy: 90.85 %\n",
      "Validation loss: 0.5715 accuracy: 82.75 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 19/20\n",
      "Trainning loss: 0.2665 accuracy: 91.75 %\n",
      "Validation loss: 0.6015 accuracy: 83.85 %\n",
      "\n",
      "-------------------------------\n",
      "Cardiomegaly Model epoch 20/20\n",
      "Trainning loss: 0.2452 accuracy: 92.36 %\n",
      "Validation loss: 0.6307 accuracy: 84.11 %\n",
      "\n",
      "Best dev Acc: 0.854320\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "build and train the model\n",
    "\"\"\"\n",
    "\n",
    "reload(_model)\n",
    "\n",
    "# build the models\n",
    "resnet = _model.TransferModel(use_cpu=False)\n",
    "\n",
    "# train + evaluate the model\n",
    "resnet.train()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(12500, 3)\n",
      "\n",
      "0.0    0.8772\n",
      "1.0    0.1228\n",
      "Name: y_true, dtype: float64\n",
      "\n",
      "0    0.91744\n",
      "1    0.08256\n",
      "Name: y_pred, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# get results on dev set\n",
    "results = resnet.evaluate_model(resnet.best_model, \n",
    "                                resnet.dataloader_dev, \n",
    "                                resnet.dev_map)\n",
    "print(results.shape)\n",
    "\n",
    "# get distributions of true labels\n",
    "print()\n",
    "print(results['y_true'].value_counts(normalize=True))\n",
    "\n",
    "# get distributions of pred labels\n",
    "print()\n",
    "print(results['y_pred'].value_counts(normalize=True))\n",
    "\n",
    "outpath = f\"results/dev_results.csv\"\n",
    "results.to_csv(outpath, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "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>y_prob</th>\n",
       "      <th>y_pred</th>\n",
       "      <th>y_true</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.211662</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.307920</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.349837</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.305924</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.270863</td>\n",
       "      <td>0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     y_prob y_pred y_true\n",
       "0  0.211662      0    0.0\n",
       "1  0.307920      0    0.0\n",
       "2  0.349837      0    0.0\n",
       "3  0.305924      0    0.0\n",
       "4  0.270863      0    0.0"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Distribution of Propensities')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "matplotlib.rcParams['figure.dpi'] = 150\n",
    "results['y_prob'].hist(edgecolor='black', bins=30)\n",
    "plt.title('Distribution of Propensities')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "_time = list(range(len(resnet.train_loss_history)))\n",
    "\n",
    "matplotlib.rcParams['figure.dpi'] = 150\n",
    "plt.plot(_time, resnet.train_loss_history, c='black', label=\"Trainning\")\n",
    "plt.plot(_time, resnet.dev_loss_history, c='black', ls=\":\", label=\"Testing\")\n",
    "plt.title(\"Training Loss\")\n",
    "plt.xlabel(\"Epoch\")\n",
    "plt.ylabel(\"Cross Entropy Loss\")\n",
    "plt.legend()\n",
    "outpath = f\"results/training_loss.png\"\n",
    "plt.savefig(outpath, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 900x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "_time = list(range(len(resnet.train_acc_history)))\n",
    "\n",
    "matplotlib.rcParams['figure.dpi'] = 150\n",
    "plt.plot(_time, resnet.train_acc_history,  c='black', label=\"Trainning\")\n",
    "plt.plot(_time, resnet.dev_acc_history, c='black', ls=\":\", label=\"Testing\")\n",
    "plt.title(\"Trainning Accuracy\")\n",
    "plt.xlabel(\"Epoch\")\n",
    "plt.ylabel(\"Accuracy\")\n",
    "plt.legend()\n",
    "outpath = f\"results/training_accuracy.png\"\n",
    "plt.savefig(outpath, bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>optimal_threshold</th>\n",
       "      <td>0.373523</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>true negatives</th>\n",
       "      <td>7002.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>true positives</th>\n",
       "      <td>985.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>false positives</th>\n",
       "      <td>3963.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>false negatives</th>\n",
       "      <td>550.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensitivity</th>\n",
       "      <td>0.641694</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>specificity</th>\n",
       "      <td>0.638577</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>F1-score</th>\n",
       "      <td>0.303872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.199070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.641694</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AUCROC</th>\n",
       "      <td>0.682099</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             0\n",
       "optimal_threshold     0.373523\n",
       "true negatives     7002.000000\n",
       "true positives      985.000000\n",
       "false positives    3963.000000\n",
       "false negatives     550.000000\n",
       "sensitivity           0.641694\n",
       "specificity           0.638577\n",
       "F1-score              0.303872\n",
       "precision             0.199070\n",
       "recall                0.641694\n",
       "AUCROC                0.682099"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reload(_tools)\n",
    "\n",
    "res = _tools.get_classification_metrics(results)\n",
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saved: models/Cardiomegaly_resnet18.pth\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "save the model\n",
    "\"\"\"\n",
    "\n",
    "outpath = f\"models/{resnet.condition}_resnet18.pth\"\n",
    "torch.save(resnet.model.state_dict(), outpath)\n",
    "print(f\"saved: {outpath}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(234, 3)\n",
      "\n",
      "0.0    0.709402\n",
      "1.0    0.290598\n",
      "Name: y_true, dtype: float64\n",
      "\n",
      "0    0.948718\n",
      "1    0.051282\n",
      "Name: y_pred, dtype: float64\n",
      "\n"
     ]
    },
    {
     "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>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>optimal_threshold</th>\n",
       "      <td>0.337538</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>true negatives</th>\n",
       "      <td>109.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>true positives</th>\n",
       "      <td>51.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>false positives</th>\n",
       "      <td>57.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>false negatives</th>\n",
       "      <td>17.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sensitivity</th>\n",
       "      <td>0.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>specificity</th>\n",
       "      <td>0.656627</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>F1-score</th>\n",
       "      <td>0.579545</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>precision</th>\n",
       "      <td>0.472222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>recall</th>\n",
       "      <td>0.750000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>AUCROC</th>\n",
       "      <td>0.756821</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            0\n",
       "optimal_threshold    0.337538\n",
       "true negatives     109.000000\n",
       "true positives      51.000000\n",
       "false positives     57.000000\n",
       "false negatives     17.000000\n",
       "sensitivity          0.750000\n",
       "specificity          0.656627\n",
       "F1-score             0.579545\n",
       "precision            0.472222\n",
       "recall               0.750000\n",
       "AUCROC               0.756821"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get results on valid set\n",
    "results = resnet.evaluate_model(resnet.best_model, \n",
    "                                resnet.dataloader_valid, \n",
    "                                resnet.valid_map)\n",
    "print(results.shape)\n",
    "\n",
    "# get distributions of true labels\n",
    "print()\n",
    "print(results['y_true'].value_counts(normalize=True))\n",
    "\n",
    "# get distributions of pred labels\n",
    "print()\n",
    "print(results['y_pred'].value_counts(normalize=True))\n",
    "\n",
    "outpath = f\"results/validation_results.csv\"\n",
    "results.to_csv(outpath, index=False)\n",
    "\n",
    "print()\n",
    "res = _tools.get_classification_metrics(results)\n",
    "res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "!git add ."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[master 27cfa88] model updates\n",
      " Committer: Cooper Stansbury <cstansbu@gl1019.arc-ts.umich.edu>\n",
      "Your name and email address were configured automatically based\n",
      "on your username and hostname. Please check that they are accurate.\n",
      "You can suppress this message by setting them explicitly:\n",
      "\n",
      "    git config --global user.name \"Your Name\"\n",
      "    git config --global user.email you@example.com\n",
      "\n",
      "After doing this, you may fix the identity used for this commit with:\n",
      "\n",
      "    git commit --amend --reset-author\n",
      "\n",
      " 10 files changed, 63620 insertions(+), 1021 deletions(-)\n",
      " rewrite .ipynb_checkpoints/chexpert_resnet18-checkpoint.ipynb (64%)\n",
      " rewrite chexpert_resnet18.ipynb (64%)\n",
      " rewrite results/dev_results.csv (85%)\n",
      " rewrite results/training_accuracy.png (98%)\n",
      " rewrite results/training_loss.png (99%)\n",
      " rewrite results/validation_results.csv (99%)\n"
     ]
    }
   ],
   "source": [
    "!git commit -m \"model updates\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}