because the input is cast to double and the closest double is 4.0250000000000003552713678800500929355621337890625, which rounds to 4.03 even though the fixed ...