DeepSHAP.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "!pip install -U numpy==1.18.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!pip install shap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#load directly from npy files #this step skips the data loading from folders as images\n",
    "off_dataset = np.load('off_dataset.npy')\n",
    "target_dataset = np.load('target_dataset.npy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# import tensorflow as tf\n",
    "# from tensorflow import keras\n",
    "# #tf.disable_v2_behavior() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import sys\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import os\n",
    "from tqdm import tqdm\n",
    "from natsort import natsorted\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "from keras.models import Sequential, Model\n",
    "from keras.layers import Layer,Input, Dense, Dropout, Activation, Concatenate, Flatten, MaxPooling2D, Convolution2D, Convolution1D, MaxPooling1D, GlobalMaxPooling1D, BatchNormalization, LSTM, GRU, Bidirectional, Lambda\n",
    "from keras.regularizers import l2,l1\n",
    "from keras.optimizers import SGD,Adam,RMSprop\n",
    "from tensorflow.compat.v1 import InteractiveSession\n",
    "import keras.backend as K\n",
    "from keras.preprocessing.image import array_to_img, img_to_array, load_img\n",
    "from keras.callbacks import EarlyStopping, ModelCheckpoint,ReduceLROnPlateau\n",
    "from keras.models import load_model\n",
    "import keras.metrics\n",
    "import tensorflow_addons as tfa\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "import tensorflow as tf\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "from keras.utils import to_categorical, plot_model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/xinfy/anaconda3/lib/python3.7/site-packages/tensorflow/python/compat/v2_compat.py:96: disable_resource_variables (from tensorflow.python.ops.variable_scope) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "non-resource variables are not supported in the long term\n"
     ]
    }
   ],
   "source": [
    "import tensorflow.compat.v1 as tf\n",
    "tf.disable_v2_behavior() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "###### check for deeplift visualization:  this is important to plot the nucleotide plots later\n",
    "from deeplift.visualization import viz_sequence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/xinfy/anaconda3/lib/python3.7/site-packages/tensorflow/python/ops/init_ops.py:97: calling VarianceScaling.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n",
      "WARNING:tensorflow:From /home/xinfy/anaconda3/lib/python3.7/site-packages/tensorflow/python/ops/init_ops.py:97: calling Orthogonal.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n",
      "WARNING:tensorflow:From /home/xinfy/anaconda3/lib/python3.7/site-packages/tensorflow/python/ops/init_ops.py:97: calling Zeros.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n"
     ]
    }
   ],
   "source": [
    "#### model\n",
    "model = tf.keras.models.load_model('model_seed10_epoch50.hdf5',compile=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set_of_random = np.random.choice(off_dataset.shape[0], 100, replace=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 109,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set_of_random.sort()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 372,
   "metadata": {},
   "outputs": [],
   "source": [
    "set_of_random_pos = np.random.choice(7130, 50, replace=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 373,
   "metadata": {},
   "outputs": [],
   "source": [
    "set_of_random_neg = np.random.choice(range(7130,153924), 50, replace=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 374,
   "metadata": {},
   "outputs": [],
   "source": [
    "set_of_random_pos.sort()\n",
    "set_of_random_neg.sort()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 375,
   "metadata": {},
   "outputs": [],
   "source": [
    "set_of_random = np.concatenate((set_of_random_pos,set_of_random_neg))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 376,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 140,  330,  437,  505,  531,  537,  557,  595,  626,  775,  897,\n",
       "       1022, 1189, 1209, 1266, 1345, 1472, 1574, 1586, 1770, 1786, 1854,\n",
       "       1988, 2077, 2088, 2130, 2383, 2454, 2795, 3175, 3257, 3367, 3516,\n",
       "       4275, 4283, 4541, 4586, 4612, 4758, 5131, 5212, 5304, 5560, 5593,\n",
       "       5694, 5974, 6499, 6555, 6638, 6992])"
      ]
     },
     "execution_count": 376,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "set_of_random_pos"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 377,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  7841,  11238,  13530,  15260,  15508,  16506,  16597,  19258,\n",
       "        26348,  28064,  29759,  31657,  32669,  36287,  41805,  42911,\n",
       "        53515,  57532,  60322,  63079,  63771,  72622,  79439,  80940,\n",
       "        87364,  89790,  90030,  90066,  92495,  94968,  95839,  97853,\n",
       "        98087, 102511, 103668, 110758, 112243, 115895, 121098, 129871,\n",
       "       130552, 133878, 135219, 136325, 139400, 140471, 141266, 145252,\n",
       "       146972, 151034])"
      ]
     },
     "execution_count": 377,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "set_of_random_neg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 378,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([   140,    330,    437,    505,    531,    537,    557,    595,\n",
       "          626,    775,    897,   1022,   1189,   1209,   1266,   1345,\n",
       "         1472,   1574,   1586,   1770,   1786,   1854,   1988,   2077,\n",
       "         2088,   2130,   2383,   2454,   2795,   3175,   3257,   3367,\n",
       "         3516,   4275,   4283,   4541,   4586,   4612,   4758,   5131,\n",
       "         5212,   5304,   5560,   5593,   5694,   5974,   6499,   6555,\n",
       "         6638,   6992,   7841,  11238,  13530,  15260,  15508,  16506,\n",
       "        16597,  19258,  26348,  28064,  29759,  31657,  32669,  36287,\n",
       "        41805,  42911,  53515,  57532,  60322,  63079,  63771,  72622,\n",
       "        79439,  80940,  87364,  89790,  90030,  90066,  92495,  94968,\n",
       "        95839,  97853,  98087, 102511, 103668, 110758, 112243, 115895,\n",
       "       121098, 129871, 130552, 133878, 135219, 136325, 139400, 140471,\n",
       "       141266, 145252, 146972, 151034])"
      ]
     },
     "execution_count": 378,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "set_of_random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {},
   "outputs": [],
   "source": [
    "#set_of_random ##makes sure same indices go for Explanation from Off-target and Target"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 379,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # select a set of background examples to take an expectation over\n",
    "bg1 = off_dataset[set_of_random]\n",
    "bg2 = target_dataset[set_of_random]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [],
   "source": [
    "#np.random.choice(off_dataset.shape[0], 100, replace=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "#np.random.choice(target_dataset.shape[0], 100, replace=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!pip uninstall --yes shap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "#!pip install shap==0.36.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 303,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for ind in range(len(model.layers)):\n",
    "#     print (ind, model.layers[ind])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "import shap"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 380,
   "metadata": {},
   "outputs": [],
   "source": [
    "e = shap.DeepExplainer(([model.layers[2].input, model.layers[3].input], model.layers[-1].output), [bg1, bg2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 381,
   "metadata": {},
   "outputs": [],
   "source": [
    "sample_id = 9400"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 202,
   "metadata": {},
   "outputs": [],
   "source": [
    "#get both types of output\n",
    "output = pd.read_csv('output2.csv')\n",
    "y = output['y']\n",
    "freq = output['freq']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 287,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "positive off target with frequencey:  39.0\n"
     ]
    }
   ],
   "source": [
    "if y[sample_id]==0:\n",
    "    print ('negative off target', 'with frequencey: ', freq[sample_id])\n",
    "else:\n",
    "    print ('positive off target','with frequencey: ', freq[sample_id])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 382,
   "metadata": {},
   "outputs": [],
   "source": [
    "x1 = off_dataset[sample_id:sample_id+1]\n",
    "x2 = target_dataset[sample_id:sample_id+1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 383,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 4, 28, 1)"
      ]
     },
     "execution_count": 383,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x1.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 384,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 4, 28, 1)"
      ]
     },
     "execution_count": 384,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "x2.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 385,
   "metadata": {},
   "outputs": [],
   "source": [
    "shap_values = e.shap_values([x1, x2])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 337,
   "metadata": {},
   "outputs": [],
   "source": [
    "# off_expls = np.sum(shap_values[0][0].squeeze(),axis=-1)[:,:,None]*x1.squeeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 339,
   "metadata": {},
   "outputs": [],
   "source": [
    "# tar_expls = np.sum(shap_values[0][1].squeeze(),axis=-1)[:,:,None]*x2.squeeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 349,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for off_item in off_expls:\n",
    "#     viz_sequence.plot_weights(off_item)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 350,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for tar_item in tar_expls:\n",
    "#     viz_sequence.plot_weights(tar_item)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 386,
   "metadata": {},
   "outputs": [],
   "source": [
    "off_shap = shap_values[0][0]\n",
    "tar_shap = shap_values[0][1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 389,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig  = plt.figure(figsize=(15,5),facecolor=\"#ffffff\",linewidth=6, edgecolor=\"#000000\")\n",
    "ax = fig.gca()\n",
    "plt.subplots_adjust(hspace = .1, wspace = .1)\n",
    "plt.subplot(2,3,1)\n",
    "plt.axis('off')\n",
    "plt.imshow(x1.squeeze(), cmap='gray')\n",
    "plt.title('off_target', color='k')\n",
    "plt.subplot(2,3,4)\n",
    "plt.axis('off')\n",
    "plt.imshow(x2.squeeze(), cmap='gray')\n",
    "plt.title('target', color='k')\n",
    "plt.subplot(2,3,2)\n",
    "plt.axis('off')\n",
    "plt.imshow(off_shap.squeeze(), cmap='jet', interpolation='bilinear')\n",
    "plt.title('Off Target SHAP', color='k')\n",
    "plt.subplot(2,3,5)\n",
    "plt.axis('off')\n",
    "plt.imshow(tar_shap.squeeze(), cmap='jet', interpolation='bilinear')\n",
    "plt.title('Target SHAP', color='k')\n",
    "plt.subplot(2,3,3)\n",
    "plt.axis('off')\n",
    "plt.imshow(x1.squeeze(), cmap='gray')\n",
    "plt.imshow(off_shap.squeeze(), cmap='jet', interpolation='bilinear', alpha=0.7)\n",
    "plt.title('Off Target SHAP Superimposed', color='k')\n",
    "plt.subplot(2,3,6)\n",
    "plt.axis('off')\n",
    "plt.imshow(x2.squeeze(), cmap='gray')\n",
    "plt.imshow(tar_shap.squeeze(), cmap='jet', interpolation='bilinear', alpha=0.7)\n",
    "plt.title('Target SHAP Superimposed', color='k')\n",
    "plt.tight_layout()\n",
    "fig.suptitle('DEEP Explainer SHAP : seq_no_' + str(sample_id) , fontsize=20, color='k')\n",
    "#plt.show()\n",
    "plt.savefig('DEEPExplainer_SHAP'+str(sample_id)+'.png', dpi=800)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 388,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the feature attributions\n",
    "shap.image_plot(shap_values[0][0], x1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 390,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "shap.image_plot(shap_values[0][1], x2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # explain how the input to the 7th layer of the model explains the top two classes\n",
    "# def map2layer(x, layer):\n",
    "#     feed_dict = dict(zip([model.layers[2].input, model.layers[3].input], [preprocess_input(x.copy())]))\n",
    "#     return K.get_session().run(model.layers[laayer].input, feed_dict)\n",
    "# e = shap.GradientExplainer((model.layers[7].input, model.layers[-1].output), map2layer(preprocess_input(X.copy()), 7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 391,
   "metadata": {},
   "outputs": [],
   "source": [
    "########### Gradient explainer\n",
    "g = shap.GradientExplainer(([model.layers[2].input, model.layers[3].input], model.layers[-1].output), [bg1, bg2],\n",
    "                           #local_smoothing=100\n",
    "                          )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 392,
   "metadata": {},
   "outputs": [],
   "source": [
    "g_shap_values = g.shap_values([x1, x2])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 359,
   "metadata": {},
   "outputs": [],
   "source": [
    "# g_off_expls = np.sum(g_shap_values[0][0].squeeze(),axis=-1)[:,:,None]*x1.squeeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 360,
   "metadata": {},
   "outputs": [],
   "source": [
    "# g_tar_expls = np.sum(g_shap_values[0][1].squeeze(),axis=-1)[:,:,None]*x2.squeeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 361,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for off_item in g_off_expls:\n",
    "#     viz_sequence.plot_weights(off_item)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 362,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for tar_item in g_tar_expls:\n",
    "#     viz_sequence.plot_weights(tar_item)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 393,
   "metadata": {},
   "outputs": [],
   "source": [
    "g_off_shap = g_shap_values[0][0]\n",
    "g_tar_shap = g_shap_values[0][1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 394,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig  = plt.figure(figsize=(15,5),facecolor=\"#ffffff\",linewidth=6, edgecolor=\"#000000\")\n",
    "ax = fig.gca()\n",
    "plt.subplots_adjust(hspace = .1, wspace = .1)\n",
    "plt.subplot(2,3,1)\n",
    "plt.axis('off')\n",
    "plt.imshow(x1.squeeze(), cmap='gray')\n",
    "plt.title('off_target', color='k')\n",
    "plt.subplot(2,3,4)\n",
    "plt.axis('off')\n",
    "plt.imshow(x2.squeeze(), cmap='gray')\n",
    "plt.title('target', color='k')\n",
    "plt.subplot(2,3,2)\n",
    "plt.axis('off')\n",
    "plt.imshow(g_off_shap.squeeze(), cmap='jet', interpolation='bilinear')\n",
    "plt.title('Off Target SHAP', color='k')\n",
    "plt.subplot(2,3,5)\n",
    "plt.axis('off')\n",
    "plt.imshow(g_tar_shap.squeeze(), cmap='jet', interpolation='bilinear')\n",
    "plt.title('Target SHAP', color='k')\n",
    "plt.subplot(2,3,3)\n",
    "plt.axis('off')\n",
    "plt.imshow(x1.squeeze(), cmap='gray')\n",
    "plt.imshow(g_off_shap.squeeze(), cmap='jet', interpolation='bilinear', alpha=0.7)\n",
    "plt.title('Off Target SHAP Superimposed', color='k')\n",
    "plt.subplot(2,3,6)\n",
    "plt.axis('off')\n",
    "plt.imshow(x2.squeeze(), cmap='gray')\n",
    "plt.imshow(g_tar_shap.squeeze(), cmap='jet', interpolation='bilinear', alpha=0.7)\n",
    "plt.title('Target SHAP Superimposed', color='k')\n",
    "plt.tight_layout()\n",
    "fig.suptitle('Gradient Explainer SHAP; local smoothing OFF : seq_no_' + str(sample_id) , fontsize=20, color='k')\n",
    "#plt.show()\n",
    "plt.savefig('GradExplainerSHAP_sample'+str(sample_id)+'.png', dpi=800)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 395,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the feature attributions\n",
    "shap.image_plot(g_shap_values[0][0], x1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 396,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x360 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot the feature attributions\n",
    "shap.image_plot(g_shap_values[0][1], x2)"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}