{"cells":[{"metadata":{"_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","trusted":true},"cell_type":"code","source":"import pandas as pd\n\nfrom fastai.vision.all import *\n\nimport os\n\npath = Path('../input/cassava-leaf-disease-classification')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"data = pd.read_csv(path/'train.csv')\ndata.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"new_lbl = {0:\"Bacterial Blight\",\n          1:\"Brown Streak Disease\",\n          2:\"Green Mottle\",\n          3:\"Mosaic Disease\",\n          4:\"Healthy\"}\n\ndata['label'].replace(new_lbl, inplace=True)\ndata.head()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#O código abaixo realiza a remoção de imagens duplicadas.\ntrain = data[~data['image_id'].isin(['1562043567.jpg', '3551135685.jpg', '2252529694.jpg'])]","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"path = Path('../input/cassava-leaf-disease-classification')\n\ndef get_x(r):\n    return path/'train_images'/r['image_id']\n\ndef get_y(r):\n    return r['label']","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def get_data(size=256, bs=64, data_df=train):\n    block = DataBlock(blocks=(ImageBlock, CategoryBlock), \n                      splitter=RandomSplitter(valid_pct= 0.2,seed=42), \n                      get_x=get_x,\n                      get_y=get_y, \n                      item_tfms = RandomResizedCrop(512),\n                      batch_tfms = [*aug_transforms(size=size),\n                                    Normalize.from_stats(*imagenet_stats)])\n\n    \n    return block.dataloaders(data_df, bs=bs)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"dataloader = get_data()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"dataloader.show_batch()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"learn = cnn_learner(dataloader, resnet50, metrics=accuracy)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"learn.lr_find()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#Valor de base_lr escolhido entre o lr_min e lr_steep\n#learn.fine_tune(2,base_lr=0.001)\nlearn.fit_one_cycle(2,0.001)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"learn.unfreeze()\n\nlearn.lr_find()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"learn.fit_one_cycle(4, lr_max=1e-4)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"interp = ClassificationInterpretation.from_learner(learn)\ninterp.plot_confusion_matrix()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"learn = cnn_learner(dataloader, resnet50, metrics=accuracy)\nlearn.fit_one_cycle(2,0.001)\nlearn.unfreeze()\nlearn.fit_one_cycle(10,lr_max=slice(1e-5,1e-4))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#learn = cnn_learner(dataloader, resnet50, metrics=accuracy)\ninterp = ClassificationInterpretation.from_learner(learn)\ninterp.plot_confusion_matrix()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#learn.export(\"../output/models/export.pth\")\nsubmission_df = learn.export(Path(\"/kaggle/working/model_cassava.pth\"))","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submission_df = pd.read_csv('../input/cassava-leaf-disease-classification/sample_submission.csv')","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"test_data_path = submission_df['image_id'].apply(lambda x: path/'test_images'/x)\ntst_dl = learn.dls.test_dl(test_data_path)\npredictions = learn.tta(dl = tst_dl, n=10)\n\nsubmission_df['label'] = np.argmax(predictions[0],axis=1)\nsubmission_df","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"submission_df.to_csv('submission.csv',index=False)","execution_count":null,"outputs":[]}],"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat":4,"nbformat_minor":4}