CNNLSTM-model-SpCas9.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from scipy.stats import pearsonr, spearmanr\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "#%matplotlib notebook\n",
    "%matplotlib inline\n",
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "%config InlineBackend.figure_format ='retina'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.keras.layers import *\n",
    "from tensorflow.keras import Input, Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from data_import_preprocessing import import_data_preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "gpus = tf.config.experimental.list_physical_devices('GPU')\n",
    "if gpus:\n",
    "    gpu_num = 1\n",
    "    try:\n",
    "        tf.config.experimental.set_visible_devices(gpus[gpu_num], 'GPU')\n",
    "        tf.config.experimental.set_memory_growth(gpus[gpu_num], True)\n",
    "    except RuntimeError as e:\n",
    "        print(e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU'),\n",
       " PhysicalDevice(name='/physical_device:GPU:1', device_type='GPU'),\n",
       " PhysicalDevice(name='/physical_device:GPU:2', device_type='GPU')]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gpus"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train_data :  ['Barcode', 'Target context sequence', 'Total read count (Day 0)', 'Indel read count\\n(Day 0)', 'Background\\nindel frequencies (%)', 'Total read count\\n(Day 2.9)', 'Indel read count\\n(Day 2.9)', 'Indel frequencies (%)', 'Background subtracted indel']\n",
      "train_data :  ['Target context sequence', 'Number of used (filtered ) barcodes', 'Total read count (average, Day 0)', 'Total read count (average, Day 2.9)', 'Background subtracted indel']\n"
     ]
    }
   ],
   "source": [
    "train_data_name = 'data/HT_Cas9_train.csv'\n",
    "test_data_name = 'data/HT_Cas9_test.csv'\n",
    "preprocessing = import_data_preprocessing(train_data_file_name = train_data_name,\n",
    "                                          test_data_file_name= test_data_name,\n",
    "                                          \n",
    "                                         )\n",
    "print('train_data : ', preprocessing.train_data_file_sample_column)\n",
    "print('train_data : ', preprocessing.test_data_file_sample_column)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = preprocessing(sgRNA_column='Target context sequence',\n",
    "                    indel_column='Background subtracted indel',\n",
    "                    split_data=0.1\n",
    "                    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dict_keys(['train', 'val', 'total', 'test'])\n",
      "dict_keys(['seq', 'indel_rate', 'indel_class', 'read_cnt', 'info'])\n"
     ]
    }
   ],
   "source": [
    "print(data.keys())\n",
    "print(data['train'].keys())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = data['train']['seq']\n",
    "class_train = data['train']['indel_class']\n",
    "rate_train = data['train']['indel_rate']\n",
    "\n",
    "X_val = data['val']['seq']\n",
    "class_val = data['val']['indel_class']\n",
    "rate_val = data['val']['indel_rate']\n",
    "\n",
    "X_test = data['test']['seq']\n",
    "class_test = data['test']['indel_class']\n",
    "rate_test = data['test']['indel_rate']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(542, 30, 4)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 128\n",
    "\n",
    "en_lr = 0.001\n",
    "en_epochs = 500\n",
    "\n",
    "dnn_lr = 0.001\n",
    "dnn_epochs = 400\n",
    "\n",
    "\n",
    "dropout_RNN = 0.5\n",
    "dropout_RNN_out = 0.5\n",
    "dropout_DNN = 0.5\n",
    "dropout2_DNN = 0.5\n",
    "\n",
    "\n",
    "DNN_1 = 100\n",
    "DNN_2 = 100\n",
    "DNN_3 = 100\n",
    "\n",
    "DNN_num_mis_1 = 100\n",
    "DNN_num_mis_2 = 100\n",
    "DNN_num_mis_3 = 5\n",
    "\n",
    "DNN_class_1 = 100\n",
    "DNN_class_2 = 100\n",
    "DNN_class_3 = 11\n",
    "\n",
    "DNN_rate_1 = 100\n",
    "DNN_rate_2 = 100\n",
    "DNN_rate_3 = 1\n",
    "\n",
    "verbose = 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "CNNLSTM_input = Input(shape=(30, 4))\n",
    "#CNNLSTM_input_reshape = Reshape((23,8))(CNNLSTM_input)\n",
    "\n",
    "kernal0 = 16\n",
    "dropout0 = 0.5\n",
    "CNN0_1 = Conv1D(kernal0, [1], activation='relu', padding='same', kernel_initializer='he_normal')(CNNLSTM_input)\n",
    "CNN0_1 = Dropout(0.3)(CNN0_1)\n",
    "\n",
    "CNN0_2 = Conv1D(kernal0, [2], activation='relu', padding='same', kernel_initializer='he_normal')(CNNLSTM_input)\n",
    "CNN0_2 = Dropout(0.3)(CNN0_2)\n",
    "\n",
    "CNN0_3 = Conv1D(kernal0, [3], activation='relu', padding='same', kernel_initializer='he_normal')(CNNLSTM_input)\n",
    "CNN0_3 = Dropout(0.3)(CNN0_3)\n",
    "\n",
    "CNN0_4 = Conv1D(kernal0, [4], activation='relu', padding='same', kernel_initializer='he_normal')(CNNLSTM_input)\n",
    "CNN0_4 = Dropout(0.3)(CNN0_4)\n",
    "\n",
    "CNN0_5 = Conv1D(kernal0, [5], activation='relu', padding='same', kernel_initializer='he_normal')(CNNLSTM_input)\n",
    "CNN0_5 = Dropout(0.3)(CNN0_5)\n",
    "\n",
    "concatenate0 = concatenate([CNN0_1, CNN0_2, CNN0_3, CNN0_4, CNN0_5])\n",
    "\n",
    "#squeezed = Lambda(lambda x: keras.backend.squeeze(x,axis=2))(concatenate0)\n",
    "\n",
    "LSTM0 = LSTM(64, return_sequences=True)(concatenate0)\n",
    "LSTM0 = Dropout(dropout0)(LSTM0)\n",
    "\n",
    "concatenate1 = concatenate([LSTM0, CNNLSTM_input])\n",
    "\n",
    "\n",
    "kernal1 = 32\n",
    "dropout1 = 0.5\n",
    "CNN1_1 = Conv1D(kernal1, [1], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate1)\n",
    "CNN1_1 = Dropout(dropout1)(CNN1_1)\n",
    "\n",
    "CNN1_2 = Conv1D(kernal1, [2], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate1)\n",
    "CNN1_2 = Dropout(dropout1)(CNN1_2)\n",
    "\n",
    "CNN1_3 = Conv1D(kernal1, [3], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate1)\n",
    "CNN1_3 = Dropout(dropout1)(CNN1_3)\n",
    "\n",
    "CNN1_4 = Conv1D(kernal1, [4], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate1)\n",
    "CNN1_4 = Dropout(dropout1)(CNN1_4)\n",
    "\n",
    "CNN1_5 = Conv1D(kernal1, [5], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate1)\n",
    "CNN1_5 = Dropout(dropout1)(CNN1_5)\n",
    "\n",
    "CNN1_concatenate = concatenate([CNN1_1, CNN1_2, CNN1_3, CNN1_4, CNN1_5])\n",
    "\n",
    "LSTM1 = LSTM(128, return_sequences=True)(CNN1_concatenate)\n",
    "LSTM1 = Dropout(dropout1)(LSTM1)\n",
    "\n",
    "\n",
    "#concatenate2 = concatenate([CNNLSTM_input_reshape,LSTM1])\n",
    "concatenate2 = concatenate([LSTM1, LSTM0])\n",
    "kernal2 = 64\n",
    "dropout2 = 0.5\n",
    "CNN2_1 = Conv1D(kernal2, [1], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate2)\n",
    "#CNN2_1 = MaxPool1D()(CNN2_1)\n",
    "CNN2_1 = Dropout(dropout2)(CNN2_1)\n",
    "\n",
    "CNN2_2 = Conv1D(kernal2, [2], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate2)\n",
    "#CNN2_2 = MaxPool1D()(CNN2_2)\n",
    "CNN2_2 = Dropout(dropout2)(CNN2_2)\n",
    "\n",
    "CNN2_3 = Conv1D(kernal2, [3], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate2)\n",
    "#CNN2_3 = MaxPool1D()(CNN2_3)\n",
    "CNN2_3 = Dropout(dropout2)(CNN2_3)\n",
    "\n",
    "CNN2_4 = Conv1D(kernal2, [4], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate2)\n",
    "#CNN2_4 = MaxPool1D()(CNN2_4)\n",
    "CNN2_4 = Dropout(dropout2)(CNN2_4)\n",
    "\n",
    "CNN2_5 = Conv1D(kernal2, [5], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate2)\n",
    "#CNN2_5 = MaxPool1D()(CNN2_5)\n",
    "CNN2_5 = Dropout(dropout2)(CNN2_5)\n",
    "\n",
    "CNN2_concatenate = concatenate([CNN2_1, CNN2_2, CNN2_3, CNN2_4, CNN2_5])\n",
    "\n",
    "LSTM2 = LSTM(256, return_sequences=True)(CNN2_concatenate)\n",
    "LSTM2 = Dropout(dropout2)(LSTM2)\n",
    "\n",
    "\n",
    "#concatenate3 = concatenate([CNNLSTM_input_reshape,LSTM2])\n",
    "concatenate3 = concatenate([LSTM2,LSTM1])\n",
    "kernal3 = 128\n",
    "dropout3 = 0.5\n",
    "CNN3_1 = Conv1D(kernal3, [1], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate3)\n",
    "#CNN3_1 = MaxPool1D()(CNN3_1)\n",
    "CNN3_1 = Dropout(dropout3)(CNN3_1)\n",
    "\n",
    "CNN3_2 = Conv1D(kernal3, [2], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate3)\n",
    "#CNN3_2 = MaxPool1D()(CNN3_2)\n",
    "CNN3_2 = Dropout(dropout3)(CNN3_2)\n",
    "\n",
    "CNN3_3 = Conv1D(kernal3, [3], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate3)\n",
    "#CNN3_3 = MaxPool1D()(CNN3_3)\n",
    "CNN3_3 = Dropout(dropout3)(CNN3_3)\n",
    "\n",
    "CNN3_4 = Conv1D(kernal3, [4], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate3)\n",
    "#CNN3_4 = MaxPool1D()(CNN3_4)\n",
    "CNN3_4 = Dropout(dropout3)(CNN3_4)\n",
    "\n",
    "CNN3_5 = Conv1D(kernal3, [5], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate3)\n",
    "#CNN3_5 = MaxPool1D()(CNN3_5)\n",
    "CNN3_5 = Dropout(dropout3)(CNN3_5)\n",
    "\n",
    "CNN3_concatenate = concatenate([CNN3_1, CNN3_2, CNN3_3, CNN3_4, CNN3_5])\n",
    "\n",
    "LSTM3 = LSTM(256, return_sequences=True)(CNN3_concatenate)\n",
    "LSTM3 = Dropout(dropout3)(LSTM3)\n",
    "\n",
    "\n",
    "\n",
    "#concatenate4 = LSTM3#concatenate([CNNLSTM_input_reshape,LSTM2])\n",
    "concatenate4 =concatenate([LSTM3, LSTM2])\n",
    "\n",
    "kernal4 = 128\n",
    "dropout3 = 0.5\n",
    "CNN4_1 = Conv1D(kernal4, [1], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate4)\n",
    "#CNN4_1 = MaxPool1D()(CNN4_1)\n",
    "CNN4_1 = Dropout(dropout3)(CNN4_1)\n",
    "\n",
    "CNN4_2 = Conv1D(kernal4, [2], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate4)\n",
    "#CNN4_2 = MaxPool1D()(CNN4_2)\n",
    "CNN4_2 = Dropout(dropout3)(CNN4_2)\n",
    "\n",
    "CNN4_3 = Conv1D(kernal4, [3], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate4)\n",
    "#CNN4_3 = MaxPool1D()(CNN4_3)\n",
    "CNN4_3 = Dropout(dropout3)(CNN4_3)\n",
    "\n",
    "CNN4_4 = Conv1D(kernal4, [4], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate4)\n",
    "#CNN4_4 = MaxPool1D()(CNN4_4)\n",
    "CNN4_4 = Dropout(dropout3)(CNN4_4)\n",
    "\n",
    "CNN4_5 = Conv1D(kernal4, [5], activation='relu', padding='same', kernel_initializer='he_normal')(concatenate4)\n",
    "#CNN4_5 = MaxPool1D()(CNN4_5)\n",
    "CNN4_5 = Dropout(dropout3)(CNN4_5)\n",
    "\n",
    "CNN4_concatenate = concatenate([CNN4_1, CNN4_2, CNN4_3, CNN4_4, CNN4_5])\n",
    "\n",
    "LSTM4 = LSTM(256, return_sequences=True)(CNN4_concatenate)\n",
    "LSTM4 = Dropout(dropout3)(LSTM4)\n",
    "\n",
    "CNN_model = Model(CNNLSTM_input, LSTM4)\n",
    "#CNN_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Failed to import pydot. You must install pydot and graphviz for `pydotprint` to work.\n"
     ]
    }
   ],
   "source": [
    "graph = tf.keras.utils.plot_model(CNN_model, show_shapes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "LSTM_middel = Lambda(lambda x: x[:,-1,:])(LSTM2) #Flatten()(CNN2_concatenate)\n",
    "LSTM_middel_2 = Lambda(lambda x: x[:,-1,:])(LSTM3) #Flatten()(CNN2_concatenate)\n",
    "LSTM_result = Lambda(lambda x: x[:,-1,:])(LSTM4) #Flatten()(CNN3_concatenate)\n",
    "\"\"\"\n",
    "#DNN model - number of mismatch prediction\n",
    "DNN_nummis_layer_input = LSTM_middel_2\n",
    "DNN_nummis_layer = Dense(DNN_num_mis_1, activation='relu')(DNN_nummis_layer_input)\n",
    "DNN_nummis_layer = Dropout(dropout2_DNN)(DNN_nummis_layer)\n",
    "DNN_nummis_layer = Dense(DNN_num_mis_3, activation='softmax', name='num_mis')(DNN_nummis_layer)\n",
    "\n",
    "\"\"\"\n",
    "#DNN model - class prediction\n",
    "DNN_class_layer_input = LSTM_middel\n",
    "DNN_class_layer_mid = Dense(DNN_class_1, activation='relu')(DNN_class_layer_input)\n",
    "DNN_class_layer = Dropout(dropout2_DNN)(DNN_class_layer_mid)\n",
    "DNN_class_layer = Dense(DNN_class_2, activation='relu')(DNN_class_layer)\n",
    "DNN_class_layer = Dropout(dropout2_DNN)(DNN_class_layer)\n",
    "DNN_class_layer = Dense(DNN_class_3, activation='softmax', name='class_1')(DNN_class_layer)\n",
    "\n",
    "DNN_class_layer_input_2 = LSTM_middel_2\n",
    "DNN_class_layer_mid_2 = Dense(DNN_class_1, activation='relu')(DNN_class_layer_input_2)\n",
    "DNN_class_layer_2 = Dropout(dropout2_DNN)(DNN_class_layer_mid_2)\n",
    "DNN_class_layer_2 = Dense(DNN_class_2, activation='relu')(DNN_class_layer_2)\n",
    "DNN_class_layer_2 = Dropout(dropout2_DNN)(DNN_class_layer_2)\n",
    "DNN_class_layer_2 = Dense(DNN_class_3, activation='softmax', name='class_2')(DNN_class_layer_2)\n",
    "\n",
    "\n",
    "#DNN model - class prediction 2\n",
    "DNN_class_layer_input_result = LSTM_result\n",
    "DNN_class_layer_result = Dense(DNN_class_1, activation='relu')(DNN_class_layer_input_result)\n",
    "DNN_class_layer_result = Dropout(dropout2_DNN)(DNN_class_layer_result)\n",
    "DNN_class_layer_mid_result = Dense(DNN_class_2, activation='relu')(DNN_class_layer_result)\n",
    "DNN_class_layer_result = Dropout(dropout2_DNN)(DNN_class_layer_mid_result)\n",
    "DNN_class_layer_result = Dense(DNN_class_3, activation='softmax', name='class_result')(DNN_class_layer_result)\n",
    "\n",
    "\n",
    "#DNN model - real value prediction\n",
    "#DNN_rate_layer_input = LSTM_result\n",
    "#DNN_rate_layer = Dense(DNN_rate_1, activation='relu', kernel_initializer='he_normal')(DNN_rate_layer_input)\n",
    "#DNN_rate_layer = Dropout(dropout2_DNN)(DNN_rate_layer)\n",
    "DNN_rate_layer = Dense(DNN_rate_2, activation='relu', kernel_initializer='he_normal')(DNN_class_layer_mid_result)\n",
    "DNN_rate_layer = Dropout(dropout2_DNN)(DNN_rate_layer)\n",
    "DNN_rate_layer = Dense(DNN_rate_3, activation='linear', kernel_initializer='he_normal', name='rate')(DNN_rate_layer)\n",
    "\n",
    "\n",
    "\n",
    "MTL_model = Model(CNNLSTM_input, [DNN_class_layer, DNN_class_layer_2,\n",
    "                                  DNN_class_layer_result, DNN_rate_layer])\n",
    "\n",
    "\n",
    "cluster_model = Model(CNNLSTM_input,DNN_class_layer)\n",
    "regression_model = Model(CNNLSTM_input, DNN_rate_layer)\n",
    "cluster_tsne_model = Model(CNNLSTM_input, DNN_class_layer_mid_result)\n",
    "\n",
    "\n",
    "#MTL_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"\\nregression_model.compile(optimizer='adam',\\n                        loss={'rate' : 'mean_squared_error'},\\n                        )\\n\""
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#optimizer = Adam(lr=0.0001)\n",
    "#optimizer_2 = Adam(lr=0.0001)\n",
    "\n",
    "MTL_model.compile(optimizer='adam',\n",
    "              loss={\n",
    "                  #'num_mis': 'categorical_crossentropy',\n",
    "                  'class_1': 'categorical_crossentropy',\n",
    "                  'class_2': 'categorical_crossentropy',\n",
    "                  'class_result': 'categorical_crossentropy',\n",
    "                  'rate': 'mean_squared_error'},\n",
    "              loss_weights={\n",
    "                  #'num_mis': 0.5,\n",
    "                  'class_1': 1,\n",
    "                  'class_2': 1,\n",
    "                  'class_result': 0.5,\n",
    "                  'rate': 1},\n",
    "              metrics={\n",
    "                  #'num_mis': 'accuracy',\n",
    "                  'class_1': \"accuracy\",\n",
    "                  'class_2': \"accuracy\",\n",
    "                  'class_result': \"accuracy\"}\n",
    "             )\n",
    "'''\n",
    "regression_model.compile(optimizer='adam',\n",
    "                        loss={'rate' : 'mean_squared_error'},\n",
    "                        )\n",
    "'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "class_weight\n",
      "{0: 1.171672077922078, 1: 0.9026811537559603, 2: 0.7846174752004348, 3: 0.7098162148872088, 4: 0.6924922043655553, 5: 0.635098718583292, 6: 0.6500422178440755, 7: 0.8690547862733293, 8: 1.841786283891547, 9: 9.631359466221852, 10: 1049.8181818181818}\n",
      "sample_weight\n",
      "{0: 1.0, 1: 1.0, 2: 1.0, 3: 1.0, 4: 1.0, 5: 1.0, 6: 0.8235090959499638, 7: 0.975946828981805, 8: 1.8421154881397088, 9: 3.998060936178457, 10: 4.0}\n"
     ]
    }
   ],
   "source": [
    "from sklearn.utils import class_weight\n",
    "class_train_num = class_train.argmax(axis=-1)\n",
    "class_weights = class_weight.compute_class_weight('balanced',\n",
    "                                                  np.unique(class_train_num),\n",
    "                                                  class_train_num\n",
    "                                                 )\n",
    "class_weights_dict = dict(enumerate(class_weights))#{ i : class_weights[i] for i in range(11)}\n",
    "print(\"class_weight\")\n",
    "print(class_weights_dict)\n",
    "\n",
    "class_wieghts_dict_tuned = class_weights_dict\n",
    "for i in range(0,11):\n",
    "    multiple_constant = 2\n",
    "    cutoff_class = 5\n",
    "    \n",
    "    if i <=cutoff_class:\n",
    "        class_wieghts_dict_tuned[i] = 1.0\n",
    "    else:\n",
    "        class_wieghts_dict_tuned[i] = multiple_constant*np.tanh((class_weights_dict[i]-2)/2) + multiple_constant\n",
    "    \n",
    "print(\"sample_weight\")\n",
    "print(class_wieghts_dict_tuned)\n",
    "\n",
    "sample_weight = np.array([class_wieghts_dict_tuned[i] for i in class_train_num])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 11548 samples, validate on 1284 samples\n",
      "Epoch 1/200\n",
      "11548/11548 [==============================] - 20s 2ms/sample - loss: 6.1912 - class_1_loss: 2.4282 - class_2_loss: 2.4383 - class_result_loss: 2.4479 - rate_loss: 0.0975 - class_1_accuracy: 0.1368 - class_2_accuracy: 0.1321 - class_result_accuracy: 0.1206 - val_loss: 6.0736 - val_class_1_loss: 2.4256 - val_class_2_loss: 2.4272 - val_class_result_loss: 2.4502 - val_rate_loss: 0.0797 - val_class_1_accuracy: 0.1706 - val_class_2_accuracy: 0.1659 - val_class_result_accuracy: 0.1410\n",
      "Epoch 2/200\n",
      "11548/11548 [==============================] - 6s 519us/sample - loss: 5.9154 - class_1_loss: 2.3162 - class_2_loss: 2.3309 - class_result_loss: 2.3848 - rate_loss: 0.0718 - class_1_accuracy: 0.1641 - class_2_accuracy: 0.1638 - class_result_accuracy: 0.1376 - val_loss: 5.7058 - val_class_1_loss: 2.2884 - val_class_2_loss: 2.2939 - val_class_result_loss: 2.3419 - val_rate_loss: 0.0478 - val_class_1_accuracy: 0.1869 - val_class_2_accuracy: 0.1869 - val_class_result_accuracy: 0.1838\n",
      "Epoch 3/200\n",
      "11548/11548 [==============================] - 6s 520us/sample - loss: 5.7517 - class_1_loss: 2.2647 - class_2_loss: 2.2786 - class_result_loss: 2.2975 - rate_loss: 0.0644 - class_1_accuracy: 0.1816 - class_2_accuracy: 0.1781 - class_result_accuracy: 0.1676 - val_loss: 5.6396 - val_class_1_loss: 2.2851 - val_class_2_loss: 2.2757 - val_class_result_loss: 2.2992 - val_rate_loss: 0.0422 - val_class_1_accuracy: 0.2072 - val_class_2_accuracy: 0.1947 - val_class_result_accuracy: 0.1916\n",
      "Epoch 4/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.6392 - class_1_loss: 2.2256 - class_2_loss: 2.2321 - class_result_loss: 2.2460 - rate_loss: 0.0576 - class_1_accuracy: 0.1894 - class_2_accuracy: 0.1886 - class_result_accuracy: 0.1886 - val_loss: 5.4331 - val_class_1_loss: 2.1781 - val_class_2_loss: 2.2037 - val_class_result_loss: 2.1972 - val_rate_loss: 0.0398 - val_class_1_accuracy: 0.2040 - val_class_2_accuracy: 0.2009 - val_class_result_accuracy: 0.1986\n",
      "Epoch 5/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.4663 - class_1_loss: 2.1554 - class_2_loss: 2.1662 - class_result_loss: 2.1867 - rate_loss: 0.0504 - class_1_accuracy: 0.2090 - class_2_accuracy: 0.2007 - class_result_accuracy: 0.2011 - val_loss: 5.2503 - val_class_1_loss: 2.1202 - val_class_2_loss: 2.1485 - val_class_result_loss: 2.1616 - val_rate_loss: 0.0361 - val_class_1_accuracy: 0.2017 - val_class_2_accuracy: 0.2126 - val_class_result_accuracy: 0.2134\n",
      "Epoch 6/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.3655 - class_1_loss: 2.1163 - class_2_loss: 2.1269 - class_result_loss: 2.1424 - rate_loss: 0.0456 - class_1_accuracy: 0.2070 - class_2_accuracy: 0.2116 - class_result_accuracy: 0.2077 - val_loss: 5.1835 - val_class_1_loss: 2.0966 - val_class_2_loss: 2.0884 - val_class_result_loss: 2.1027 - val_rate_loss: 0.0386 - val_class_1_accuracy: 0.2079 - val_class_2_accuracy: 0.2095 - val_class_result_accuracy: 0.2033\n",
      "Epoch 7/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.3237 - class_1_loss: 2.1039 - class_2_loss: 2.1158 - class_result_loss: 2.1253 - rate_loss: 0.0445 - class_1_accuracy: 0.2107 - class_2_accuracy: 0.2117 - class_result_accuracy: 0.2095 - val_loss: 5.0975 - val_class_1_loss: 2.0717 - val_class_2_loss: 2.0821 - val_class_result_loss: 2.0925 - val_rate_loss: 0.0319 - val_class_1_accuracy: 0.2126 - val_class_2_accuracy: 0.2056 - val_class_result_accuracy: 0.2150\n",
      "Epoch 8/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 5.2305 - class_1_loss: 2.0666 - class_2_loss: 2.0799 - class_result_loss: 2.0821 - rate_loss: 0.0403 - class_1_accuracy: 0.2104 - class_2_accuracy: 0.2193 - class_result_accuracy: 0.2154 - val_loss: 5.0695 - val_class_1_loss: 2.0515 - val_class_2_loss: 2.0705 - val_class_result_loss: 2.0720 - val_rate_loss: 0.0299 - val_class_1_accuracy: 0.2134 - val_class_2_accuracy: 0.2173 - val_class_result_accuracy: 0.2259\n",
      "Epoch 9/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.2068 - class_1_loss: 2.0577 - class_2_loss: 2.0689 - class_result_loss: 2.0832 - rate_loss: 0.0404 - class_1_accuracy: 0.2193 - class_2_accuracy: 0.2154 - class_result_accuracy: 0.2136 - val_loss: 5.3089 - val_class_1_loss: 2.1639 - val_class_2_loss: 2.1797 - val_class_result_loss: 2.1682 - val_rate_loss: 0.0346 - val_class_1_accuracy: 0.1822 - val_class_2_accuracy: 0.1706 - val_class_result_accuracy: 0.1690\n",
      "Epoch 10/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.1556 - class_1_loss: 2.0401 - class_2_loss: 2.0471 - class_result_loss: 2.0517 - rate_loss: 0.0387 - class_1_accuracy: 0.2187 - class_2_accuracy: 0.2180 - class_result_accuracy: 0.2147 - val_loss: 4.9813 - val_class_1_loss: 2.0408 - val_class_2_loss: 2.0438 - val_class_result_loss: 2.0478 - val_rate_loss: 0.0302 - val_class_1_accuracy: 0.2298 - val_class_2_accuracy: 0.2290 - val_class_result_accuracy: 0.2391\n",
      "Epoch 11/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.1164 - class_1_loss: 2.0201 - class_2_loss: 2.0392 - class_result_loss: 2.0422 - rate_loss: 0.0383 - class_1_accuracy: 0.2272 - class_2_accuracy: 0.2235 - class_result_accuracy: 0.2232 - val_loss: 4.9639 - val_class_1_loss: 2.0279 - val_class_2_loss: 2.0475 - val_class_result_loss: 2.0425 - val_rate_loss: 0.0289 - val_class_1_accuracy: 0.2321 - val_class_2_accuracy: 0.2251 - val_class_result_accuracy: 0.2298\n",
      "Epoch 12/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.0793 - class_1_loss: 2.0066 - class_2_loss: 2.0190 - class_result_loss: 2.0275 - rate_loss: 0.0365 - class_1_accuracy: 0.2326 - class_2_accuracy: 0.2271 - class_result_accuracy: 0.2287 - val_loss: 4.9555 - val_class_1_loss: 2.0376 - val_class_2_loss: 2.0322 - val_class_result_loss: 2.0422 - val_rate_loss: 0.0278 - val_class_1_accuracy: 0.2336 - val_class_2_accuracy: 0.2422 - val_class_result_accuracy: 0.2399\n",
      "Epoch 13/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.0741 - class_1_loss: 2.0108 - class_2_loss: 2.0162 - class_result_loss: 2.0223 - rate_loss: 0.0361 - class_1_accuracy: 0.2315 - class_2_accuracy: 0.2352 - class_result_accuracy: 0.2268 - val_loss: 5.0773 - val_class_1_loss: 2.0777 - val_class_2_loss: 2.0849 - val_class_result_loss: 2.0941 - val_rate_loss: 0.0316 - val_class_1_accuracy: 0.2227 - val_class_2_accuracy: 0.2095 - val_class_result_accuracy: 0.2126\n",
      "Epoch 14/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 5.0344 - class_1_loss: 1.9914 - class_2_loss: 2.0026 - class_result_loss: 2.0041 - rate_loss: 0.0355 - class_1_accuracy: 0.2354 - class_2_accuracy: 0.2362 - class_result_accuracy: 0.2346 - val_loss: 4.9757 - val_class_1_loss: 2.0329 - val_class_2_loss: 2.0349 - val_class_result_loss: 2.0404 - val_rate_loss: 0.0298 - val_class_1_accuracy: 0.2461 - val_class_2_accuracy: 0.2414 - val_class_result_accuracy: 0.2290\n",
      "Epoch 15/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 5.0213 - class_1_loss: 1.9842 - class_2_loss: 1.9999 - class_result_loss: 2.0037 - rate_loss: 0.0346 - class_1_accuracy: 0.2355 - class_2_accuracy: 0.2365 - class_result_accuracy: 0.2294 - val_loss: 4.8547 - val_class_1_loss: 1.9932 - val_class_2_loss: 2.0049 - val_class_result_loss: 2.0095 - val_rate_loss: 0.0275 - val_class_1_accuracy: 0.2453 - val_class_2_accuracy: 0.2430 - val_class_result_accuracy: 0.2500\n",
      "Epoch 16/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.9907 - class_1_loss: 1.9751 - class_2_loss: 1.9828 - class_result_loss: 1.9888 - rate_loss: 0.0337 - class_1_accuracy: 0.2433 - class_2_accuracy: 0.2450 - class_result_accuracy: 0.2439 - val_loss: 4.8372 - val_class_1_loss: 2.0041 - val_class_2_loss: 2.0102 - val_class_result_loss: 2.0067 - val_rate_loss: 0.0276 - val_class_1_accuracy: 0.2407 - val_class_2_accuracy: 0.2313 - val_class_result_accuracy: 0.2321\n",
      "Epoch 17/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.9770 - class_1_loss: 1.9741 - class_2_loss: 1.9739 - class_result_loss: 1.9792 - rate_loss: 0.0338 - class_1_accuracy: 0.2464 - class_2_accuracy: 0.2526 - class_result_accuracy: 0.2372 - val_loss: 4.8231 - val_class_1_loss: 1.9812 - val_class_2_loss: 1.9714 - val_class_result_loss: 1.9789 - val_rate_loss: 0.0258 - val_class_1_accuracy: 0.2578 - val_class_2_accuracy: 0.2632 - val_class_result_accuracy: 0.2453\n",
      "Epoch 18/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.9516 - class_1_loss: 1.9630 - class_2_loss: 1.9733 - class_result_loss: 1.9780 - rate_loss: 0.0332 - class_1_accuracy: 0.2480 - class_2_accuracy: 0.2422 - class_result_accuracy: 0.2445 - val_loss: 4.8788 - val_class_1_loss: 2.0189 - val_class_2_loss: 2.0224 - val_class_result_loss: 2.0213 - val_rate_loss: 0.0279 - val_class_1_accuracy: 0.2399 - val_class_2_accuracy: 0.2477 - val_class_result_accuracy: 0.2383\n",
      "Epoch 19/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.9700 - class_1_loss: 1.9718 - class_2_loss: 1.9756 - class_result_loss: 1.9892 - rate_loss: 0.0342 - class_1_accuracy: 0.2387 - class_2_accuracy: 0.2437 - class_result_accuracy: 0.2314 - val_loss: 4.9114 - val_class_1_loss: 2.0368 - val_class_2_loss: 2.0312 - val_class_result_loss: 2.0426 - val_rate_loss: 0.0304 - val_class_1_accuracy: 0.2274 - val_class_2_accuracy: 0.2344 - val_class_result_accuracy: 0.2344\n",
      "Epoch 20/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.9090 - class_1_loss: 1.9427 - class_2_loss: 1.9537 - class_result_loss: 1.9602 - rate_loss: 0.0325 - class_1_accuracy: 0.2425 - class_2_accuracy: 0.2445 - class_result_accuracy: 0.2458 - val_loss: 4.8642 - val_class_1_loss: 1.9935 - val_class_2_loss: 1.9986 - val_class_result_loss: 1.9947 - val_rate_loss: 0.0280 - val_class_1_accuracy: 0.2445 - val_class_2_accuracy: 0.2484 - val_class_result_accuracy: 0.2329\n",
      "Epoch 21/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.8981 - class_1_loss: 1.9458 - class_2_loss: 1.9456 - class_result_loss: 1.9495 - rate_loss: 0.0327 - class_1_accuracy: 0.2500 - class_2_accuracy: 0.2538 - class_result_accuracy: 0.2455 - val_loss: 4.8296 - val_class_1_loss: 2.0066 - val_class_2_loss: 2.0004 - val_class_result_loss: 2.0038 - val_rate_loss: 0.0269 - val_class_1_accuracy: 0.2617 - val_class_2_accuracy: 0.2632 - val_class_result_accuracy: 0.2469\n",
      "Epoch 22/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.8881 - class_1_loss: 1.9404 - class_2_loss: 1.9427 - class_result_loss: 1.9504 - rate_loss: 0.0323 - class_1_accuracy: 0.2523 - class_2_accuracy: 0.2523 - class_result_accuracy: 0.2484 - val_loss: 4.8464 - val_class_1_loss: 2.0069 - val_class_2_loss: 2.0179 - val_class_result_loss: 2.0135 - val_rate_loss: 0.0281 - val_class_1_accuracy: 0.2438 - val_class_2_accuracy: 0.2500 - val_class_result_accuracy: 0.2430\n",
      "Epoch 23/200\n",
      "11548/11548 [==============================] - 6s 534us/sample - loss: 4.8891 - class_1_loss: 1.9354 - class_2_loss: 1.9486 - class_result_loss: 1.9537 - rate_loss: 0.0326 - class_1_accuracy: 0.2553 - class_2_accuracy: 0.2485 - class_result_accuracy: 0.2515 - val_loss: 4.8798 - val_class_1_loss: 2.0289 - val_class_2_loss: 2.0213 - val_class_result_loss: 2.0240 - val_rate_loss: 0.0279 - val_class_1_accuracy: 0.2391 - val_class_2_accuracy: 0.2430 - val_class_result_accuracy: 0.2414\n",
      "Epoch 24/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.8635 - class_1_loss: 1.9257 - class_2_loss: 1.9367 - class_result_loss: 1.9422 - rate_loss: 0.0322 - class_1_accuracy: 0.2471 - class_2_accuracy: 0.2553 - class_result_accuracy: 0.2481 - val_loss: 4.8132 - val_class_1_loss: 1.9888 - val_class_2_loss: 1.9945 - val_class_result_loss: 1.9898 - val_rate_loss: 0.0265 - val_class_1_accuracy: 0.2500 - val_class_2_accuracy: 0.2516 - val_class_result_accuracy: 0.2547\n",
      "Epoch 25/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.8608 - class_1_loss: 1.9258 - class_2_loss: 1.9377 - class_result_loss: 1.9327 - rate_loss: 0.0317 - class_1_accuracy: 0.2555 - class_2_accuracy: 0.2565 - class_result_accuracy: 0.2562 - val_loss: 4.8151 - val_class_1_loss: 2.0020 - val_class_2_loss: 2.0151 - val_class_result_loss: 2.0201 - val_rate_loss: 0.0272 - val_class_1_accuracy: 0.2710 - val_class_2_accuracy: 0.2601 - val_class_result_accuracy: 0.2593\n",
      "Epoch 26/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.8401 - class_1_loss: 1.9227 - class_2_loss: 1.9243 - class_result_loss: 1.9297 - rate_loss: 0.0312 - class_1_accuracy: 0.2517 - class_2_accuracy: 0.2584 - class_result_accuracy: 0.2607 - val_loss: 4.8807 - val_class_1_loss: 2.0032 - val_class_2_loss: 2.0097 - val_class_result_loss: 2.0035 - val_rate_loss: 0.0263 - val_class_1_accuracy: 0.2438 - val_class_2_accuracy: 0.2336 - val_class_result_accuracy: 0.2438\n",
      "Epoch 27/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.8435 - class_1_loss: 1.9186 - class_2_loss: 1.9280 - class_result_loss: 1.9255 - rate_loss: 0.0307 - class_1_accuracy: 0.2610 - class_2_accuracy: 0.2549 - class_result_accuracy: 0.2592 - val_loss: 4.8913 - val_class_1_loss: 2.0116 - val_class_2_loss: 2.0211 - val_class_result_loss: 2.0197 - val_rate_loss: 0.0288 - val_class_1_accuracy: 0.2453 - val_class_2_accuracy: 0.2500 - val_class_result_accuracy: 0.2438\n",
      "Epoch 28/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.8194 - class_1_loss: 1.9084 - class_2_loss: 1.9139 - class_result_loss: 1.9171 - rate_loss: 0.0308 - class_1_accuracy: 0.2512 - class_2_accuracy: 0.2574 - class_result_accuracy: 0.2564 - val_loss: 4.7607 - val_class_1_loss: 1.9756 - val_class_2_loss: 1.9747 - val_class_result_loss: 1.9796 - val_rate_loss: 0.0258 - val_class_1_accuracy: 0.2617 - val_class_2_accuracy: 0.2734 - val_class_result_accuracy: 0.2664\n",
      "Epoch 29/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.7987 - class_1_loss: 1.9000 - class_2_loss: 1.9129 - class_result_loss: 1.9121 - rate_loss: 0.0302 - class_1_accuracy: 0.2588 - class_2_accuracy: 0.2648 - class_result_accuracy: 0.2526 - val_loss: 4.8477 - val_class_1_loss: 1.9967 - val_class_2_loss: 1.9989 - val_class_result_loss: 2.0012 - val_rate_loss: 0.0279 - val_class_1_accuracy: 0.2523 - val_class_2_accuracy: 0.2578 - val_class_result_accuracy: 0.2586\n",
      "Epoch 30/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.7909 - class_1_loss: 1.9027 - class_2_loss: 1.9038 - class_result_loss: 1.9087 - rate_loss: 0.0305 - class_1_accuracy: 0.2561 - class_2_accuracy: 0.2641 - class_result_accuracy: 0.2655 - val_loss: 4.9277 - val_class_1_loss: 2.0427 - val_class_2_loss: 2.0432 - val_class_result_loss: 2.0466 - val_rate_loss: 0.0315 - val_class_1_accuracy: 0.2399 - val_class_2_accuracy: 0.2578 - val_class_result_accuracy: 0.2500\n",
      "Epoch 31/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.7806 - class_1_loss: 1.8962 - class_2_loss: 1.8998 - class_result_loss: 1.9054 - rate_loss: 0.0302 - class_1_accuracy: 0.2640 - class_2_accuracy: 0.2587 - class_result_accuracy: 0.2587 - val_loss: 4.8139 - val_class_1_loss: 1.9853 - val_class_2_loss: 1.9915 - val_class_result_loss: 1.9897 - val_rate_loss: 0.0271 - val_class_1_accuracy: 0.2593 - val_class_2_accuracy: 0.2593 - val_class_result_accuracy: 0.2539\n",
      "Epoch 32/200\n",
      "11548/11548 [==============================] - 6s 536us/sample - loss: 4.7567 - class_1_loss: 1.8915 - class_2_loss: 1.8946 - class_result_loss: 1.9014 - rate_loss: 0.0300 - class_1_accuracy: 0.2623 - class_2_accuracy: 0.2607 - class_result_accuracy: 0.2591 - val_loss: 4.8207 - val_class_1_loss: 2.0010 - val_class_2_loss: 1.9925 - val_class_result_loss: 1.9928 - val_rate_loss: 0.0261 - val_class_1_accuracy: 0.2578 - val_class_2_accuracy: 0.2586 - val_class_result_accuracy: 0.2562\n",
      "Epoch 33/200\n",
      "11548/11548 [==============================] - 6s 526us/sample - loss: 4.7563 - class_1_loss: 1.8844 - class_2_loss: 1.8938 - class_result_loss: 1.8931 - rate_loss: 0.0296 - class_1_accuracy: 0.2656 - class_2_accuracy: 0.2619 - class_result_accuracy: 0.2624 - val_loss: 4.7952 - val_class_1_loss: 1.9878 - val_class_2_loss: 1.9893 - val_class_result_loss: 1.9879 - val_rate_loss: 0.0267 - val_class_1_accuracy: 0.2601 - val_class_2_accuracy: 0.2671 - val_class_result_accuracy: 0.2617\n",
      "Epoch 34/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.7570 - class_1_loss: 1.8868 - class_2_loss: 1.8920 - class_result_loss: 1.8934 - rate_loss: 0.0295 - class_1_accuracy: 0.2599 - class_2_accuracy: 0.2642 - class_result_accuracy: 0.2566 - val_loss: 4.8183 - val_class_1_loss: 1.9873 - val_class_2_loss: 1.9879 - val_class_result_loss: 1.9897 - val_rate_loss: 0.0273 - val_class_1_accuracy: 0.2469 - val_class_2_accuracy: 0.2438 - val_class_result_accuracy: 0.2422\n",
      "Epoch 35/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.7452 - class_1_loss: 1.8811 - class_2_loss: 1.8877 - class_result_loss: 1.8932 - rate_loss: 0.0296 - class_1_accuracy: 0.2632 - class_2_accuracy: 0.2654 - class_result_accuracy: 0.2666 - val_loss: 4.8332 - val_class_1_loss: 2.0087 - val_class_2_loss: 1.9939 - val_class_result_loss: 1.9823 - val_rate_loss: 0.0277 - val_class_1_accuracy: 0.2555 - val_class_2_accuracy: 0.2469 - val_class_result_accuracy: 0.2516\n",
      "Epoch 36/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.7233 - class_1_loss: 1.8718 - class_2_loss: 1.8761 - class_result_loss: 1.8904 - rate_loss: 0.0298 - class_1_accuracy: 0.2633 - class_2_accuracy: 0.2632 - class_result_accuracy: 0.2626 - val_loss: 4.8000 - val_class_1_loss: 1.9899 - val_class_2_loss: 1.9936 - val_class_result_loss: 1.9779 - val_rate_loss: 0.0257 - val_class_1_accuracy: 0.2632 - val_class_2_accuracy: 0.2492 - val_class_result_accuracy: 0.2718\n",
      "Epoch 37/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.7224 - class_1_loss: 1.8741 - class_2_loss: 1.8757 - class_result_loss: 1.8830 - rate_loss: 0.0291 - class_1_accuracy: 0.2637 - class_2_accuracy: 0.2635 - class_result_accuracy: 0.2586 - val_loss: 4.7863 - val_class_1_loss: 2.0033 - val_class_2_loss: 1.9939 - val_class_result_loss: 1.9858 - val_rate_loss: 0.0266 - val_class_1_accuracy: 0.2414 - val_class_2_accuracy: 0.2484 - val_class_result_accuracy: 0.2523\n",
      "Epoch 38/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.7124 - class_1_loss: 1.8722 - class_2_loss: 1.8723 - class_result_loss: 1.8854 - rate_loss: 0.0292 - class_1_accuracy: 0.2635 - class_2_accuracy: 0.2664 - class_result_accuracy: 0.2641 - val_loss: 4.8557 - val_class_1_loss: 2.0211 - val_class_2_loss: 2.0165 - val_class_result_loss: 2.0246 - val_rate_loss: 0.0275 - val_class_1_accuracy: 0.2477 - val_class_2_accuracy: 0.2547 - val_class_result_accuracy: 0.2523\n",
      "Epoch 39/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.6997 - class_1_loss: 1.8622 - class_2_loss: 1.8713 - class_result_loss: 1.8760 - rate_loss: 0.0287 - class_1_accuracy: 0.2704 - class_2_accuracy: 0.2680 - class_result_accuracy: 0.2693 - val_loss: 4.7842 - val_class_1_loss: 1.9971 - val_class_2_loss: 1.9940 - val_class_result_loss: 1.9756 - val_rate_loss: 0.0258 - val_class_1_accuracy: 0.2656 - val_class_2_accuracy: 0.2578 - val_class_result_accuracy: 0.2648\n",
      "Epoch 40/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.6864 - class_1_loss: 1.8571 - class_2_loss: 1.8662 - class_result_loss: 1.8652 - rate_loss: 0.0290 - class_1_accuracy: 0.2705 - class_2_accuracy: 0.2687 - class_result_accuracy: 0.2669 - val_loss: 4.7926 - val_class_1_loss: 2.0384 - val_class_2_loss: 2.0015 - val_class_result_loss: 1.9996 - val_rate_loss: 0.0258 - val_class_1_accuracy: 0.2586 - val_class_2_accuracy: 0.2695 - val_class_result_accuracy: 0.2578\n",
      "Epoch 41/200\n",
      "11548/11548 [==============================] - 6s 534us/sample - loss: 4.6913 - class_1_loss: 1.8652 - class_2_loss: 1.8669 - class_result_loss: 1.8712 - rate_loss: 0.0285 - class_1_accuracy: 0.2696 - class_2_accuracy: 0.2662 - class_result_accuracy: 0.2639 - val_loss: 4.8781 - val_class_1_loss: 2.0423 - val_class_2_loss: 2.0366 - val_class_result_loss: 2.0525 - val_rate_loss: 0.0279 - val_class_1_accuracy: 0.2679 - val_class_2_accuracy: 0.2671 - val_class_result_accuracy: 0.2586\n",
      "Epoch 42/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.6933 - class_1_loss: 1.8641 - class_2_loss: 1.8675 - class_result_loss: 1.8694 - rate_loss: 0.0287 - class_1_accuracy: 0.2732 - class_2_accuracy: 0.2653 - class_result_accuracy: 0.2675 - val_loss: 4.8122 - val_class_1_loss: 2.0430 - val_class_2_loss: 2.0258 - val_class_result_loss: 2.0242 - val_rate_loss: 0.0267 - val_class_1_accuracy: 0.2570 - val_class_2_accuracy: 0.2484 - val_class_result_accuracy: 0.2578\n",
      "Epoch 43/200\n",
      "11548/11548 [==============================] - 6s 529us/sample - loss: 4.6752 - class_1_loss: 1.8534 - class_2_loss: 1.8634 - class_result_loss: 1.8629 - rate_loss: 0.0285 - class_1_accuracy: 0.2782 - class_2_accuracy: 0.2740 - class_result_accuracy: 0.2734 - val_loss: 4.7855 - val_class_1_loss: 1.9823 - val_class_2_loss: 1.9730 - val_class_result_loss: 1.9839 - val_rate_loss: 0.0268 - val_class_1_accuracy: 0.2360 - val_class_2_accuracy: 0.2508 - val_class_result_accuracy: 0.2469\n",
      "Epoch 44/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.6550 - class_1_loss: 1.8491 - class_2_loss: 1.8570 - class_result_loss: 1.8549 - rate_loss: 0.0280 - class_1_accuracy: 0.2690 - class_2_accuracy: 0.2752 - class_result_accuracy: 0.2678 - val_loss: 4.8396 - val_class_1_loss: 2.0346 - val_class_2_loss: 2.0135 - val_class_result_loss: 2.0241 - val_rate_loss: 0.0266 - val_class_1_accuracy: 0.2617 - val_class_2_accuracy: 0.2625 - val_class_result_accuracy: 0.2477\n",
      "Epoch 45/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.6363 - class_1_loss: 1.8348 - class_2_loss: 1.8447 - class_result_loss: 1.8498 - rate_loss: 0.0279 - class_1_accuracy: 0.2769 - class_2_accuracy: 0.2778 - class_result_accuracy: 0.2765 - val_loss: 4.8297 - val_class_1_loss: 2.0152 - val_class_2_loss: 2.0108 - val_class_result_loss: 2.0092 - val_rate_loss: 0.0277 - val_class_1_accuracy: 0.2352 - val_class_2_accuracy: 0.2430 - val_class_result_accuracy: 0.2383\n",
      "Epoch 46/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.6372 - class_1_loss: 1.8412 - class_2_loss: 1.8435 - class_result_loss: 1.8502 - rate_loss: 0.0274 - class_1_accuracy: 0.2761 - class_2_accuracy: 0.2713 - class_result_accuracy: 0.2689 - val_loss: 4.8196 - val_class_1_loss: 1.9947 - val_class_2_loss: 1.9937 - val_class_result_loss: 1.9803 - val_rate_loss: 0.0255 - val_class_1_accuracy: 0.2445 - val_class_2_accuracy: 0.2609 - val_class_result_accuracy: 0.2531\n",
      "Epoch 47/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.6225 - class_1_loss: 1.8350 - class_2_loss: 1.8415 - class_result_loss: 1.8421 - rate_loss: 0.0278 - class_1_accuracy: 0.2723 - class_2_accuracy: 0.2735 - class_result_accuracy: 0.2726 - val_loss: 4.8100 - val_class_1_loss: 1.9980 - val_class_2_loss: 1.9957 - val_class_result_loss: 1.9882 - val_rate_loss: 0.0261 - val_class_1_accuracy: 0.2671 - val_class_2_accuracy: 0.2710 - val_class_result_accuracy: 0.2702\n",
      "Epoch 48/200\n",
      "11548/11548 [==============================] - 6s 528us/sample - loss: 4.6501 - class_1_loss: 1.8423 - class_2_loss: 1.8500 - class_result_loss: 1.8544 - rate_loss: 0.0279 - class_1_accuracy: 0.2736 - class_2_accuracy: 0.2697 - class_result_accuracy: 0.2717 - val_loss: 4.8275 - val_class_1_loss: 2.0095 - val_class_2_loss: 2.0162 - val_class_result_loss: 1.9959 - val_rate_loss: 0.0286 - val_class_1_accuracy: 0.2492 - val_class_2_accuracy: 0.2461 - val_class_result_accuracy: 0.2461\n",
      "Epoch 49/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.5765 - class_1_loss: 1.8183 - class_2_loss: 1.8191 - class_result_loss: 1.8218 - rate_loss: 0.0269 - class_1_accuracy: 0.2800 - class_2_accuracy: 0.2821 - class_result_accuracy: 0.2788 - val_loss: 4.9725 - val_class_1_loss: 2.0904 - val_class_2_loss: 2.0618 - val_class_result_loss: 2.0733 - val_rate_loss: 0.0282 - val_class_1_accuracy: 0.2422 - val_class_2_accuracy: 0.2360 - val_class_result_accuracy: 0.2305\n",
      "Epoch 50/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.6055 - class_1_loss: 1.8292 - class_2_loss: 1.8312 - class_result_loss: 1.8374 - rate_loss: 0.0273 - class_1_accuracy: 0.2719 - class_2_accuracy: 0.2772 - class_result_accuracy: 0.2752 - val_loss: 4.8702 - val_class_1_loss: 2.0471 - val_class_2_loss: 2.0389 - val_class_result_loss: 2.0221 - val_rate_loss: 0.0275 - val_class_1_accuracy: 0.2617 - val_class_2_accuracy: 0.2547 - val_class_result_accuracy: 0.2508\n",
      "Epoch 51/200\n",
      "11548/11548 [==============================] - 6s 531us/sample - loss: 4.6020 - class_1_loss: 1.8251 - class_2_loss: 1.8300 - class_result_loss: 1.8319 - rate_loss: 0.0271 - class_1_accuracy: 0.2710 - class_2_accuracy: 0.2730 - class_result_accuracy: 0.2698 - val_loss: 4.8697 - val_class_1_loss: 2.0366 - val_class_2_loss: 2.0315 - val_class_result_loss: 2.0272 - val_rate_loss: 0.0263 - val_class_1_accuracy: 0.2687 - val_class_2_accuracy: 0.2671 - val_class_result_accuracy: 0.2601\n",
      "Epoch 52/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.5917 - class_1_loss: 1.8226 - class_2_loss: 1.8248 - class_result_loss: 1.8324 - rate_loss: 0.0268 - class_1_accuracy: 0.2786 - class_2_accuracy: 0.2792 - class_result_accuracy: 0.2733 - val_loss: 4.8089 - val_class_1_loss: 2.0155 - val_class_2_loss: 2.0042 - val_class_result_loss: 2.0018 - val_rate_loss: 0.0262 - val_class_1_accuracy: 0.2632 - val_class_2_accuracy: 0.2632 - val_class_result_accuracy: 0.2664\n",
      "Epoch 53/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.5832 - class_1_loss: 1.8164 - class_2_loss: 1.8210 - class_result_loss: 1.8272 - rate_loss: 0.0272 - class_1_accuracy: 0.2734 - class_2_accuracy: 0.2762 - class_result_accuracy: 0.2821 - val_loss: 4.9238 - val_class_1_loss: 2.0529 - val_class_2_loss: 2.0520 - val_class_result_loss: 2.0463 - val_rate_loss: 0.0276 - val_class_1_accuracy: 0.2578 - val_class_2_accuracy: 0.2578 - val_class_result_accuracy: 0.2625\n",
      "Epoch 54/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.5580 - class_1_loss: 1.8073 - class_2_loss: 1.8147 - class_result_loss: 1.8198 - rate_loss: 0.0265 - class_1_accuracy: 0.2806 - class_2_accuracy: 0.2778 - class_result_accuracy: 0.2723 - val_loss: 4.8156 - val_class_1_loss: 2.0191 - val_class_2_loss: 1.9977 - val_class_result_loss: 2.0088 - val_rate_loss: 0.0260 - val_class_1_accuracy: 0.2539 - val_class_2_accuracy: 0.2609 - val_class_result_accuracy: 0.2516\n",
      "Epoch 55/200\n",
      "11548/11548 [==============================] - 6s 526us/sample - loss: 4.5630 - class_1_loss: 1.8106 - class_2_loss: 1.8136 - class_result_loss: 1.8201 - rate_loss: 0.0263 - class_1_accuracy: 0.2826 - class_2_accuracy: 0.2800 - class_result_accuracy: 0.2774 - val_loss: 4.8983 - val_class_1_loss: 2.0137 - val_class_2_loss: 2.0007 - val_class_result_loss: 1.9985 - val_rate_loss: 0.0285 - val_class_1_accuracy: 0.2414 - val_class_2_accuracy: 0.2407 - val_class_result_accuracy: 0.2313\n",
      "Epoch 56/200\n",
      "11548/11548 [==============================] - 6s 542us/sample - loss: 4.5477 - class_1_loss: 1.8111 - class_2_loss: 1.8099 - class_result_loss: 1.8131 - rate_loss: 0.0266 - class_1_accuracy: 0.2781 - class_2_accuracy: 0.2815 - class_result_accuracy: 0.2811 - val_loss: 4.8823 - val_class_1_loss: 2.0056 - val_class_2_loss: 1.9934 - val_class_result_loss: 1.9860 - val_rate_loss: 0.0257 - val_class_1_accuracy: 0.2352 - val_class_2_accuracy: 0.2399 - val_class_result_accuracy: 0.2391\n",
      "Epoch 57/200\n",
      "11548/11548 [==============================] - 6s 526us/sample - loss: 4.5314 - class_1_loss: 1.7952 - class_2_loss: 1.8040 - class_result_loss: 1.8069 - rate_loss: 0.0269 - class_1_accuracy: 0.2833 - class_2_accuracy: 0.2835 - class_result_accuracy: 0.2836 - val_loss: 4.9087 - val_class_1_loss: 2.0394 - val_class_2_loss: 2.0631 - val_class_result_loss: 2.0439 - val_rate_loss: 0.0282 - val_class_1_accuracy: 0.2555 - val_class_2_accuracy: 0.2492 - val_class_result_accuracy: 0.2570\n",
      "Epoch 58/200\n",
      "11548/11548 [==============================] - 6s 527us/sample - loss: 4.4948 - class_1_loss: 1.7847 - class_2_loss: 1.7863 - class_result_loss: 1.7970 - rate_loss: 0.0254 - class_1_accuracy: 0.2918 - class_2_accuracy: 0.2810 - class_result_accuracy: 0.2875 - val_loss: 4.8992 - val_class_1_loss: 2.0419 - val_class_2_loss: 2.0334 - val_class_result_loss: 2.0165 - val_rate_loss: 0.0266 - val_class_1_accuracy: 0.2617 - val_class_2_accuracy: 0.2726 - val_class_result_accuracy: 0.2625\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\"\\nregression_model.fit(x=X_train,\\n                     y=rate_train,\\n                     validation_data=(X_val, rate_val),\\n                     sample_weight={'rate' : sample_weight},\\n                     shuffle=True,\\n                     epochs = 300,\\n                     batch_size=batch_size,\\n                     callbacks=[early_stopping]\\n                    \\n                    )\\n\""
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "early_stopping = tf.keras.callbacks.EarlyStopping(monitor='val_loss',\n",
    "                                                  min_delta=0.0001,\n",
    "                                                  patience=30, verbose=0, mode='auto')\n",
    "\n",
    "MTL_model.fit(x=X_train,\n",
    "              y={\n",
    "                  #'num_mis': num_mis_train,\n",
    "                  'class_1': class_train,\n",
    "                  'class_2': class_train,\n",
    "                  'class_result': class_train,\n",
    "                  'rate': rate_train},\n",
    "              validation_data=(X_val, {#'num_mis': num_mis_val,\n",
    "                                       'class_1': class_val,\n",
    "                                       'class_2': class_val,\n",
    "                                       'class_result': class_val,\n",
    "                                       'rate': rate_val}),\n",
    "              class_weight={\n",
    "                  'class_1' : class_weights_dict,\n",
    "                  'class_2' : class_weights_dict,\n",
    "                  'class_result' : class_weights_dict},\n",
    "              sample_weight={'rate' : sample_weight},\n",
    "              shuffle=True,\n",
    "              epochs=200,\n",
    "              batch_size=batch_size,\n",
    "              verbose=verbose,\n",
    "              callbacks=[early_stopping]\n",
    "             )\n",
    "\"\"\"\n",
    "regression_model.fit(x=X_train,\n",
    "                     y=rate_train,\n",
    "                     validation_data=(X_val, rate_val),\n",
    "                     sample_weight={'rate' : sample_weight},\n",
    "                     shuffle=True,\n",
    "                     epochs = 300,\n",
    "                     batch_size=batch_size,\n",
    "                     callbacks=[early_stopping]\n",
    "                    \n",
    "                    )\n",
    "\"\"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Correlation : 0.7350022981765048\n",
      "Pearson Correlation : 0.7350022981765049\n",
      "Spearman Correlation : 0.7350635719385278\n"
     ]
    }
   ],
   "source": [
    "test_prediction = regression_model.predict(X_test)\n",
    "test_prediction = np.array(test_prediction).reshape(-1,)\n",
    "test_true = rate_test.reshape(-1,)\n",
    "\n",
    "test_correlation = np.corrcoef(test_prediction, test_true).flatten()[1]\n",
    "pearson_corr = pearsonr(test_prediction, test_true)\n",
    "spearman_corr = spearmanr(test_prediction, test_true)\n",
    "print('Correlation : {}'.format(test_correlation))\n",
    "print('Pearson Correlation : {}'.format(pearson_corr[0]))\n",
    "print('Spearman Correlation : {}'.format(spearman_corr[0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "loss             : 4.4791\n",
      "class_1_loss     : 1.7827\n",
      "class_2_loss     : 1.7851\n",
      "class_result_loss : 1.7838\n",
      "rate_loss        : 0.0197\n",
      "class_1_accuracy : 0.2638\n",
      "class_2_accuracy : 0.2583\n",
      "class_result_accuracy : 0.2583\n"
     ]
    }
   ],
   "source": [
    "class_result_acc = MTL_model.evaluate(X_test,\n",
    "                                      {\n",
    "                                          #'num_mis': num_mis_test,\n",
    "                                          'class_1': class_test,\n",
    "                                          'class_2': class_test,\n",
    "                                          'class_result': class_test,\n",
    "                                          'rate': rate_test},\n",
    "                                      verbose=0\n",
    "                                     )\n",
    "#print(MTL_model.metrics_names)\n",
    "#print(class_result_acc)\n",
    "result_label = MTL_model.metrics_names\n",
    "for label, result in zip(result_label, class_result_acc):\n",
    "    print('{:16} : {:>6.4f}'.format(label,result))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "indel_predict = test_prediction\n",
    "indel_true = test_true\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(6,6))\n",
    "ax.scatter(x=indel_true, y=indel_predict)\n",
    "ax.set_title('Test data', fontsize=20)\n",
    "ax.set_xlabel('True', fontsize=15)\n",
    "ax.set_ylabel('Prediction', fontsize=15)\n",
    "ax.set_xlim(0,1)\n",
    "ax.set_ylim(0,1)\n",
    "ax.grid(True)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_prediction = regression_model.predict(X_train).reshape(-1,)\n",
    "train_true = rate_train.reshape(-1,)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "indel_predict = train_prediction\n",
    "indel_true = train_true\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(6,6))\n",
    "ax.scatter(x=indel_true, y=indel_predict)\n",
    "ax.set_title('Train data', fontsize=20)\n",
    "ax.set_xlabel('True', fontsize=15)\n",
    "ax.set_ylabel('Prediction', fontsize=15)\n",
    "ax.set_xlim(0,1)\n",
    "ax.set_ylim(0,1)\n",
    "ax.grid(True)\n",
    "fig.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "sns.set(style=\"whitegrid\")\n",
    "total_indel = data['test']['indel_rate']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x864 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 749,
       "width": 414
      }
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#ax = sns.boxplot(x=\"CAS9\", y=\"Indel frequency\", data=total_indel)\n",
    "#ax = sns.swarmplot(x=\"CAS9\", y=\"Indel frequency\", data=total_indel, color=\".25\")\n",
    "plt.figure(figsize=(6, 12))\n",
    "ax = sns.boxplot(data=total_indel,width=0.3)\n",
    "ax = sns.stripplot(data=total_indel, color=\".3\")\n",
    "#ax.set(xlabel='CAS9', ylabel='Indel frequency')\n",
    "ax.axes.set_title(\"Indel frequency\",fontsize=30)\n",
    "ax.set_xlabel(\"CAS9\",fontsize=20)\n",
    "ax.set_ylabel(\"Indel frequency\",fontsize=20)\n",
    "ax.tick_params(labelsize=20)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "'numpy.ndarray' object has no attribute 'values'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-32-0fded85ab7ca>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0msns\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcatplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtotal_indel\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/seaborn/categorical.py\u001b[0m in \u001b[0;36mcatplot\u001b[0;34m(x, y, hue, data, row, col, col_wrap, estimator, ci, n_boot, units, seed, order, hue_order, row_order, col_order, kind, height, aspect, orient, color, palette, legend, legend_out, sharex, sharey, margin_titles, facet_kws, **kwargs)\u001b[0m\n\u001b[1;32m   3763\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3764\u001b[0m     \u001b[0;31m# Draw the plot onto the facets\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3765\u001b[0;31m     \u001b[0mg\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmap_dataframe\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mplot_func\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mplot_kws\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3766\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3767\u001b[0m     \u001b[0;31m# Special case axis labels for a count type plot\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/seaborn/axisgrid.py\u001b[0m in \u001b[0;36mmap_dataframe\u001b[0;34m(self, func, *args, **kwargs)\u001b[0m\n\u001b[1;32m    808\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    809\u001b[0m             \u001b[0;31m# If this subset is null, move on\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 810\u001b[0;31m             \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mdata_ijk\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mvalues\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    811\u001b[0m                 \u001b[0;32mcontinue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    812\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mAttributeError\u001b[0m: 'numpy.ndarray' object has no attribute 'values'"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "image/png": {
       "height": 348,
       "width": 348
      }
     },
     "output_type": "display_data"
    }
   ],
   "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}