i-zaitsev

def generate_dataset(n_rows, num_count, cat_count, max_nan=0.1, max_cat_size=100):
    """Randomly generate datasets with numerical and categorical features.
    
    The numerical features are taken from the normal distribution X ~ N(0, 1).
    The categorical features are generated as random uuid4 strings with 
    cardinality C where 2 <= C <= max_cat_size.
    
    Also, a max_nan proportion of both numerical and categorical features is replaces
    with NaN values.
    """