{"cells":[{"metadata":{},"cell_type":"markdown","source":"In this kernel, I will resize all the train images to (512, 512) because I/O with this dataset is not great."},{"metadata":{},"cell_type":"markdown","source":"## Import necessary libraries"},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"import os\nimport gc\nimport cv2\n\nimport numpy as np\nimport pandas as pd\nfrom joblib import delayed, Parallel\nfrom tqdm.notebook import tqdm, trange","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"N = 16\nH = 512\nW = 512\nC = cv2.COLOR_BGR2RGB","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Define paths and load .csv files"},{"metadata":{"trusted":true},"cell_type":"code","source":"#TEST_IMG_PATH = '../input/siim-isic-melanoma-classification/jpeg/test/'\nTRAIN_IMG_PATH = '../input/ocular-disease-recognition-odir5k/ODIR-5K/ODIR-5K/Training Images/'","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#test_df = pd.read_csv('../input/siim-isic-melanoma-classification/test.csv')#\ntrain_df = pd.read_csv('../input/oc-csv-file/ocular.csv')","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Resize images to (512, 512) and save (with multi-threading)"},{"metadata":{"trusted":true},"cell_type":"code","source":"def save_to_zip(out_path):\n\n    if 'train_1' in out_path:\n        !zip -r train_1.zip train_1\n        !rm -rf train_1\n\n    if 'train_2' in out_path:\n        !zip -r train_2.zip train_2\n        !rm -rf train_2","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"def save(ids, in_path, out_path):\n    # Resize images to (512, 512) and save\n\n    ids = tqdm(ids)\n\n    for idx, image_name in enumerate(ids):\n        input_read_path = in_path + image_name\n        image = cv2.imread(input_read_path + '.jpg')\n        image = cv2.resize(cv2.cvtColor(image, C), (H, W))\n        output_write_path = out_path + image_name + '.jpg'\n        cv2.imwrite(output_write_path, image); del image; gc.collect()","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"# Create directories and image ID lists\n\n\n!mkdir train_1\n!mkdir train_2\n\nlength = int(0.5*len(train_df))\ntrain_ids_1 = np.array_split(np.array(train_df.filename[:length]), N)\ntrain_ids_2 = np.array_split(np.array(train_df.filename[length:]), N)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"#\npath = \"train_1/\"\nparallel = Parallel(n_jobs=N, backend=\"threading\")\nparallel(delayed(save)(ids, TRAIN_IMG_PATH, path) for ids in train_ids_1)\n\npath = \"train_2/\"\nparallel = Parallel(n_jobs=N, backend=\"threading\")\nparallel(delayed(save)(ids, TRAIN_IMG_PATH, path) for ids in train_ids_2)","execution_count":null,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"\n# Save train images to ZIP\n\nsave_to_zip(\"train_1\")\nsave_to_zip(\"train_2\")","execution_count":null,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"## Thank you :D"}],"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}