Hello! I am dealing with image annotations for machine-learning purposes. My input-file is a .csv-file, which I need to edit. So output should still be .csv, but the little content that is actually required from the original file should be presented in a different structure. That means, that a lot of content from the original file can be deleted. My challange is that the input file is quite large: Right now, I am just working with a snipped, consisting of  a 8 rows (that's 1 image per row