If you're looking to debug, any constant would suffice for simulating a measurement. Why go to such lengths to create a constant when numeric constant already exists?
And yes, you don't want to compare floating point exactly. Even before we worry about that, though, it doesn't seem you understand the rest of what is happening with that loop. If you haven't already made changes to handle float comparisons, you might consider using integer types to understand the comparisons using a simple numeric constant. From there, it'll be easier to expand to float logic. From there, you can modify your constant input to be your actual measurements.
