def load_dataset(self, task=None): if task is None: self.dataset = load_dataset( self.folder_with_preprocessed_data[self.tasks[0]]) else: self.dataset = load_dataset( self.folder_with_preprocessed_data[task])
def load_dataset(self): self.dataset = load_dataset(self.folder_with_preprocessed_data)
max(params.get('num_threads') // 2, 1), params.get("num_cached_per_thread"), seeds=seeds_val, pin_memory=pin_memory) return batchgenerator_train, batchgenerator_val if __name__ == "__main__": from nnunet.training.dataloading.dataset_loading import DataLoader3D, load_dataset from nnunet.paths import preprocessing_output_dir import os import pickle t = "Task002_Heart" p = os.path.join(preprocessing_output_dir, t) dataset = load_dataset(p, 0) with open(os.path.join(p, "plans.pkl"), 'rb') as f: plans = pickle.load(f) basic_patch_size = get_patch_size( np.array(plans['stage_properties'][0].patch_size), default_3D_augmentation_params['rotation_x'], default_3D_augmentation_params['rotation_y'], default_3D_augmentation_params['rotation_z'], default_3D_augmentation_params['scale_range']) dl = DataLoader3D( dataset, basic_patch_size, np.array(plans['stage_properties'][0].patch_size).astype(int), 1) tr, val = get_default_augmentation( dl, dl,