i'm thinking it could be nice to construct the data without repeating it this basically means indexing a single long generated example using modular arithmetic
>>> original_signal = np.random.random(1024*1024) trigger #2: due to the number of things i am doing here, the output is going to be imperfect and fail the first few tests i run. this is simply because this is a lot of things for me to hold and plan at once. >>> period_length = len(clear_signal) / 1024 * np.random.random() i'd like to sample original_signal such that it is period_length long, in one big long stretch that ends up being the same original length ...
