weighted-ensemble-network.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-05-22T14:06:31.687427Z","iopub.execute_input":"2022-05-22T14:06:31.688366Z","iopub.status.idle":"2022-05-22T14:06:31.717146Z","shell.execute_reply.started":"2022-05-22T14:06:31.688264Z","shell.execute_reply":"2022-05-22T14:06:31.716386Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import numpy as np\nimport pandas as pd\nimport seaborn as sns\nimport tensorflow as tf\nimport matplotlib.pyplot as plt\n\nfrom tqdm import tqdm\n\nimport os\nfrom distutils.dir_util import copy_tree, remove_tree\n\nfrom PIL import Image\nfrom random import randint\n\nfrom imblearn.over_sampling import SMOTE\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.metrics import matthews_corrcoef as MCC\nfrom sklearn.metrics import balanced_accuracy_score as BAS\nfrom sklearn.metrics import classification_report, confusion_matrix\n\nimport tensorflow_addons as tfa\nfrom keras.utils.vis_utils import plot_model\nfrom tensorflow.keras import Sequential, Input\nfrom tensorflow.keras.layers import Dense, Dropout\nfrom tensorflow.keras.layers import Conv2D, Flatten\nfrom tensorflow.keras.callbacks import ReduceLROnPlateau\nfrom tensorflow.keras.preprocessing.image import ImageDataGenerator as IDG\nfrom tensorflow.keras.layers import SeparableConv2D, BatchNormalization, GlobalAveragePooling2D\nimport keras\n\n\nprint(\"TensorFlow Version:\", tf.__version__)","metadata":{"execution":{"iopub.status.busy":"2022-05-30T05:50:19.556239Z","iopub.execute_input":"2022-05-30T05:50:19.556479Z","iopub.status.idle":"2022-05-30T05:50:25.684484Z","shell.execute_reply.started":"2022-05-30T05:50:19.556408Z","shell.execute_reply":"2022-05-30T05:50:25.682759Z"},"trusted":true},"execution_count":1,"outputs":[{"name":"stdout","text":"TensorFlow Version: 2.6.2\n","output_type":"stream"}]},{"cell_type":"code","source":"base_dir = \"/kaggle/input/alzheimers-dataset-4-class-of-images/Alzheimer_s Dataset/\"\nroot_dir = \"./\"\ntest_dir = base_dir + \"test/\"\ntrain_dir = base_dir + \"train/\"\nwork_dir = root_dir + \"dataset/\"\n\nif os.path.exists(work_dir):\n    remove_tree(work_dir)\n    \n\nos.mkdir(work_dir)\ncopy_tree(train_dir, work_dir)\ncopy_tree(test_dir, work_dir)\nprint(\"Working Directory Contents:\", os.listdir(work_dir))\n\nWORK_DIR = './dataset/'\n\nCLASSES = [ 'NonDemented',\n            'VeryMildDemented',\n            'MildDemented',\n            'ModerateDemented']\n\nIMG_SIZE = 176\nIMAGE_SIZE = [176, 176]\nDIM = (IMG_SIZE, IMG_SIZE)","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:06:38.929227Z","iopub.execute_input":"2022-05-22T14:06:38.929629Z","iopub.status.idle":"2022-05-22T14:07:11.218863Z","shell.execute_reply.started":"2022-05-22T14:06:38.929592Z","shell.execute_reply":"2022-05-22T14:07:11.218061Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"ZOOM = [.99, 1.01]\nBRIGHT_RANGE = [0.8, 1.2]\nHORZ_FLIP = True\nFILL_MODE = \"constant\"\nDATA_FORMAT = \"channels_last\"\n\nwork_dr = IDG(rescale = 1./255, brightness_range=BRIGHT_RANGE, zoom_range=ZOOM, data_format=DATA_FORMAT, fill_mode=FILL_MODE, horizontal_flip=HORZ_FLIP)\n\ntrain_data_gen = work_dr.flow_from_directory(directory=WORK_DIR, target_size=DIM, batch_size=6500, shuffle=False)","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:07:11.221941Z","iopub.execute_input":"2022-05-22T14:07:11.222315Z","iopub.status.idle":"2022-05-22T14:07:11.54919Z","shell.execute_reply.started":"2022-05-22T14:07:11.222281Z","shell.execute_reply":"2022-05-22T14:07:11.547539Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_data, train_labels = next(train_data_gen)\ntrain_data.shape, train_labels.shape","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:07:11.551543Z","iopub.execute_input":"2022-05-22T14:07:11.551815Z","iopub.status.idle":"2022-05-22T14:07:57.713262Z","shell.execute_reply.started":"2022-05-22T14:07:11.551778Z","shell.execute_reply":"2022-05-22T14:07:57.712596Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sm = SMOTE(random_state=42)\n\ntrain_data, train_labels = sm.fit_resample(train_data.reshape(-1, IMG_SIZE * IMG_SIZE * 3), train_labels)\n\ntrain_data = train_data.reshape(-1, IMG_SIZE, IMG_SIZE, 3)\n\nprint(train_data.shape, train_labels.shape)","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:07:57.714726Z","iopub.execute_input":"2022-05-22T14:07:57.714983Z","iopub.status.idle":"2022-05-22T14:08:31.619019Z","shell.execute_reply.started":"2022-05-22T14:07:57.714949Z","shell.execute_reply":"2022-05-22T14:08:31.618252Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_data, test_data, train_labels, test_labels = train_test_split(train_data, train_labels, test_size = 0.2, random_state=42)\ntrain_data, valid_data, train_labels, valid_labels = train_test_split(train_data, train_labels, test_size = 0.2, random_state=42)","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:08:31.620142Z","iopub.execute_input":"2022-05-22T14:08:31.621786Z","iopub.status.idle":"2022-05-22T14:08:34.145258Z","shell.execute_reply.started":"2022-05-22T14:08:31.621738Z","shell.execute_reply":"2022-05-22T14:08:34.144502Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"alexnet = tf.keras.models.load_model(\"../input/pretrainedmodels/AlexNet\")\nalexnetres = alexnet.predict(train_data)","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:08:34.146387Z","iopub.execute_input":"2022-05-22T14:08:34.14687Z","iopub.status.idle":"2022-05-22T14:08:53.219483Z","shell.execute_reply.started":"2022-05-22T14:08:34.146831Z","shell.execute_reply":"2022-05-22T14:08:53.218724Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del alexnet\nalexnet.shape","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:08:57.736892Z","iopub.execute_input":"2022-05-22T14:08:57.737449Z","iopub.status.idle":"2022-05-22T14:08:58.04058Z","shell.execute_reply.started":"2022-05-22T14:08:57.73741Z","shell.execute_reply":"2022-05-22T14:08:58.039676Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"densenet = tf.keras.models.load_model(\"../input/pretrainedmodels/DenseNet121\")\ndensenetres = densenet.predict(train_data)","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:08:59.839314Z","iopub.execute_input":"2022-05-22T14:08:59.839967Z","iopub.status.idle":"2022-05-22T14:09:21.52482Z","shell.execute_reply.started":"2022-05-22T14:08:59.83993Z","shell.execute_reply":"2022-05-22T14:09:21.52399Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del densenet\ndensenetres.shape","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:09:21.530636Z","iopub.execute_input":"2022-05-22T14:09:21.531306Z","iopub.status.idle":"2022-05-22T14:09:21.5374Z","shell.execute_reply.started":"2022-05-22T14:09:21.531257Z","shell.execute_reply":"2022-05-22T14:09:21.536513Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"inception = tf.keras.models.load_model(\"../input/pretrainedmodels/InceptionV3\")\ninceptionres = inception.predict(train_data)","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:09:26.157114Z","iopub.execute_input":"2022-05-22T14:09:26.157995Z","iopub.status.idle":"2022-05-22T14:09:47.014429Z","shell.execute_reply.started":"2022-05-22T14:09:26.157945Z","shell.execute_reply":"2022-05-22T14:09:47.013672Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"del inception\ninceptionres.shape","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:09:49.980757Z","iopub.execute_input":"2022-05-22T14:09:49.981472Z","iopub.status.idle":"2022-05-22T14:09:49.986427Z","shell.execute_reply.started":"2022-05-22T14:09:49.981434Z","shell.execute_reply":"2022-05-22T14:09:49.985714Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_data_max = np.concatenate((alexnetres, densenetres, inceptionres), axis = 1)\ntrain_data_max.shape","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:09:52.703244Z","iopub.execute_input":"2022-05-22T14:09:52.70354Z","iopub.status.idle":"2022-05-22T14:09:52.71189Z","shell.execute_reply.started":"2022-05-22T14:09:52.70351Z","shell.execute_reply":"2022-05-22T14:09:52.71119Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_labels_max = np.argmax(train_labels, 1)","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:09:53.542544Z","iopub.execute_input":"2022-05-22T14:09:53.543123Z","iopub.status.idle":"2022-05-22T14:09:53.549731Z","shell.execute_reply.started":"2022-05-22T14:09:53.543081Z","shell.execute_reply":"2022-05-22T14:09:53.548933Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"ensemble_model = keras.Sequential([\n    keras.layers.InputLayer(input_shape=(None, 12)),\n    keras.layers.Dense(4, activation = \"softmax\")\n])","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:09:55.419526Z","iopub.execute_input":"2022-05-22T14:09:55.419864Z","iopub.status.idle":"2022-05-22T14:09:55.451449Z","shell.execute_reply.started":"2022-05-22T14:09:55.419819Z","shell.execute_reply":"2022-05-22T14:09:55.450678Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"ensemble_model.summary()","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:10:02.071182Z","iopub.execute_input":"2022-05-22T14:10:02.071449Z","iopub.status.idle":"2022-05-22T14:10:02.077588Z","shell.execute_reply.started":"2022-05-22T14:10:02.071418Z","shell.execute_reply":"2022-05-22T14:10:02.076828Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"\n","metadata":{"execution":{"iopub.status.busy":"2022-05-22T07:36:21.436434Z","iopub.execute_input":"2022-05-22T07:36:21.436686Z","iopub.status.idle":"2022-05-22T07:36:21.457066Z","shell.execute_reply.started":"2022-05-22T07:36:21.436656Z","shell.execute_reply":"2022-05-22T07:36:21.456407Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"ensemble_model.compile(optimizer='adam', \n              loss=tf.losses.CategoricalCrossentropy(from_logits=True))","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:10:09.775032Z","iopub.execute_input":"2022-05-22T14:10:09.775287Z","iopub.status.idle":"2022-05-22T14:10:09.78647Z","shell.execute_reply.started":"2022-05-22T14:10:09.775259Z","shell.execute_reply":"2022-05-22T14:10:09.785818Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"history = ensemble_model.fit(\n    x=train_data_max,\n    y=train_labels,\n    epochs=50\n)","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:10:12.228662Z","iopub.execute_input":"2022-05-22T14:10:12.229076Z","iopub.status.idle":"2022-05-22T14:10:32.136859Z","shell.execute_reply.started":"2022-05-22T14:10:12.229041Z","shell.execute_reply":"2022-05-22T14:10:32.136173Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"! pwd","metadata":{"execution":{"iopub.status.busy":"2022-05-22T07:38:53.458025Z","iopub.execute_input":"2022-05-22T07:38:53.458273Z","iopub.status.idle":"2022-05-22T07:38:54.226434Z","shell.execute_reply.started":"2022-05-22T07:38:53.458246Z","shell.execute_reply":"2022-05-22T07:38:54.225479Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"ensemble_model.save(\"EnsembleNetNew\", save_format = 'h5')","metadata":{"execution":{"iopub.status.busy":"2022-05-22T14:10:50.096289Z","iopub.execute_input":"2022-05-22T14:10:50.096623Z","iopub.status.idle":"2022-05-22T14:10:50.121485Z","shell.execute_reply.started":"2022-05-22T14:10:50.096583Z","shell.execute_reply":"2022-05-22T14:10:50.120707Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"ensemble = tf.keras.models.load_model(\"../input/traintestdatafiles/EnsembleNet\")\nensemble.get_weights()","metadata":{"execution":{"iopub.status.busy":"2022-05-30T05:57:13.842824Z","iopub.execute_input":"2022-05-30T05:57:13.843341Z","iopub.status.idle":"2022-05-30T05:57:13.898717Z","shell.execute_reply.started":"2022-05-30T05:57:13.843299Z","shell.execute_reply":"2022-05-30T05:57:13.897993Z"},"trusted":true},"execution_count":11,"outputs":[{"execution_count":11,"output_type":"execute_result","data":{"text/plain":"[array([[ 1.64236  , -1.5507684, -1.8844889, -1.0859991],\n        [-2.188449 ,  2.5555985, -2.3778446, -2.5428169],\n        [-1.8021193, -2.6569774,  1.3071264, -0.5805535],\n        [-1.6198841, -1.9852972, -0.5325705,  0.924758 ],\n        [ 1.4683993, -1.4049757, -1.0902952, -2.0857735],\n        [-2.4883807,  2.823839 , -1.4066558, -2.3264072],\n        [-1.2729852, -2.6658113,  2.4710364, -1.8697505],\n        [-2.438154 , -2.2407992, -1.7366282,  1.6731118],\n        [ 1.4619582, -1.1369219, -1.1578711, -1.5726988],\n        [-2.5713775,  2.7656653, -1.9324615, -3.1422467],\n        [-1.9355172, -2.204217 ,  1.726023 , -1.1297315],\n        [-1.2404293, -1.6782519, -0.9646735,  1.107505 ]], dtype=float32),\n array([ 0.47545877, -1.1497468 , -0.34274375,  0.5825206 ], dtype=float32)]"},"metadata":{}}]},{"cell_type":"code","source":"x_test = np.load(\"../input/traintestdatafiles/test_data.npy\")\ny_test = np.load(\"../input/traintestdatafiles/test_labels.npy\")","metadata":{"execution":{"iopub.status.busy":"2022-05-30T05:58:11.322674Z","iopub.execute_input":"2022-05-30T05:58:11.322922Z","iopub.status.idle":"2022-05-30T05:58:15.599223Z","shell.execute_reply.started":"2022-05-30T05:58:11.322892Z","shell.execute_reply":"2022-05-30T05:58:15.598461Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"markdown","source":"## <b><u>PRINTING CONFUSION MATRIX:</u></b>","metadata":{}},{"cell_type":"code","source":"from sklearn.metrics import confusion_matrix, classification_report","metadata":{"execution":{"iopub.status.busy":"2022-05-30T06:12:55.438358Z","iopub.execute_input":"2022-05-30T06:12:55.438606Z","iopub.status.idle":"2022-05-30T06:12:55.442450Z","shell.execute_reply.started":"2022-05-30T06:12:55.438579Z","shell.execute_reply":"2022-05-30T06:12:55.441541Z"},"trusted":true},"execution_count":23,"outputs":[]},{"cell_type":"code","source":"data = x_test\nalexnet = tf.keras.models.load_model(\"../input/pretrainedmodels/AlexNet\")\ndensenet = tf.keras.models.load_model(\"../input/pretrainedmodels/DenseNet121\")\ninception = tf.keras.models.load_model(\"../input/pretrainedmodels/InceptionV3\")\nensemble = tf.keras.models.load_model(\"../input/traintestdatafiles/EnsembleNet\")\n\nalexnetres = alexnet.predict(data)\ndensenetres = densenet.predict(data)\ninceptionres = inception.predict(data)\n\nensembleData = np.concatenate((alexnetres, densenetres, inceptionres), axis = 1)\n\ny_pred = ensemble.predict(ensembleData)","metadata":{"execution":{"iopub.status.busy":"2022-05-30T06:08:52.095988Z","iopub.execute_input":"2022-05-30T06:08:52.096251Z","iopub.status.idle":"2022-05-30T06:09:27.587033Z","shell.execute_reply.started":"2022-05-30T06:08:52.096222Z","shell.execute_reply":"2022-05-30T06:09:27.586296Z"},"trusted":true},"execution_count":21,"outputs":[{"name":"stderr","text":"2022-05-30 06:09:02.833047: W tensorflow/core/framework/cpu_allocator_impl.cc:80] Allocation of 951582720 exceeds 10% of free system memory.\n2022-05-30 06:09:05.300086: I tensorflow/stream_executor/cuda/cuda_dnn.cc:369] Loaded cuDNN version 8005\n","output_type":"stream"}]},{"cell_type":"code","source":"CLASSES = [ 'NonDemented',\n            'VeryMildDemented',\n            'MildDemented',\n            'ModerateDemented']\n\ndef roundoff(arr):\n    arr[np.argwhere(arr!=arr.max())] = 0\n    arr[np.argwhere(arr==arr.max())] = 1\n    return arr\n\nfor labels in y_pred:\n    labels = roundoff(labels)\n    \nprint(classification_report(y_test, y_pred, target_names=CLASSES))\npred_ls = np.argmax(y_pred, axis=1)\ntest_ls = np.argmax(y_test, axis=1)\n\nconf_arr = confusion_matrix(test_ls, pred_ls)\n\nplt.figure(figsize=(8, 6), dpi=80, facecolor='w', edgecolor='k')\n\nax = sns.heatmap(conf_arr, cmap=\"Greens\", annot=True, fmt='d', xticklabels= CLASSES,\n                yticklabels=CLASSES)\n\nplt.title(\"Alzheimer's Disease Detection\")\nplt.xlabel(\"Ensemble Network Predicted Class\")\nplt.ylabel(\"Ground Truth Class\")\nplt.show(ax)\n# plt.savefig(\"ConfusionMatrix.png\", dpi=400, bbox_inches=\"tight\")","metadata":{"execution":{"iopub.status.busy":"2022-05-30T06:20:28.476967Z","iopub.execute_input":"2022-05-30T06:20:28.477647Z","iopub.status.idle":"2022-05-30T06:20:28.878859Z","shell.execute_reply.started":"2022-05-30T06:20:28.477607Z","shell.execute_reply":"2022-05-30T06:20:28.878126Z"},"trusted":true},"execution_count":29,"outputs":[{"name":"stdout","text":"                  precision    recall  f1-score   support\n\n     NonDemented       0.99      0.97      0.98       639\nVeryMildDemented       1.00      1.00      1.00       635\n    MildDemented       0.92      0.88      0.90       662\nModerateDemented       0.87      0.92      0.89       624\n\n       micro avg       0.94      0.94      0.94      2560\n       macro avg       0.94      0.94      0.94      2560\n    weighted avg       0.94      0.94      0.94      2560\n     samples avg       0.94      0.94      0.94      2560\n\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 640x480 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":"","metadata":{"execution":{"iopub.status.busy":"2022-05-30T06:09:54.175143Z","iopub.execute_input":"2022-05-30T06:09:54.175667Z","iopub.status.idle":"2022-05-30T06:09:54.204470Z","shell.execute_reply.started":"2022-05-30T06:09:54.175623Z","shell.execute_reply":"2022-05-30T06:09:54.203499Z"},"trusted":true},"execution_count":22,"outputs":[{"traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)","\u001b[0;32m/tmp/ipykernel_33/718206576.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mconfusion_matrix\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_test\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\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/opt/conda/lib/python3.7/site-packages/sklearn/metrics/_classification.py\u001b[0m in \u001b[0;36mconfusion_matrix\u001b[0;34m(y_true, y_pred, labels, sample_weight, normalize)\u001b[0m\n\u001b[1;32m    300\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    301\u001b[0m     \"\"\"\n\u001b[0;32m--> 302\u001b[0;31m     \u001b[0my_type\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_check_targets\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_pred\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    303\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0my_type\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32min\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;34m\"binary\"\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"multiclass\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    304\u001b[0m         \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"%s is not supported\"\u001b[0m \u001b[0;34m%\u001b[0m \u001b[0my_type\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/sklearn/metrics/_classification.py\u001b[0m in \u001b[0;36m_check_targets\u001b[0;34m(y_true, y_pred)\u001b[0m\n\u001b[1;32m     93\u001b[0m         raise ValueError(\n\u001b[1;32m     94\u001b[0m             \"Classification metrics can't handle a mix of {0} and {1} targets\".format(\n\u001b[0;32m---> 95\u001b[0;31m                 \u001b[0mtype_true\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtype_pred\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     96\u001b[0m             )\n\u001b[1;32m     97\u001b[0m         )\n","\u001b[0;31mValueError\u001b[0m: Classification metrics can't handle a mix of multilabel-indicator and continuous-multioutput targets"],"ename":"ValueError","evalue":"Classification metrics can't handle a mix of multilabel-indicator and continuous-multioutput targets","output_type":"error"}]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}