if not os.path.exists(data_folder): os.makedirs(data_folder) if next(os.walk(data_folder))[1]: retrain = False if retrain: input, target, classes = data.sample(data_folder) model = classifier.build(input.shape, target.shape) classifier.train(model, input, target) classifier.save(models_folder, model, classes) else: model, classes = classifier.load(models_folder, sorted(os.listdir(models_folder))[-1]) for root, dirs, files in os.walk(data_folder): for file in files: if not file.startswith('.'): with open(root+'/'+file) as f: input = data.str2mat(f.read()) output = classifier.run(model, input) <<<<<<< HEAD data.backtest(save_loc, classes, input, output) ======= data.backtest(save_loc+'/'+file, classes, input, output) >>>>>>> refs/remotes/nathankjer/master else: print("""\nNo data found.\nPut subfolders of files by class, within the 'data' folder.""") if __name__ == "__main__": main()