Excellent! See how convenient these numbers are? Still, there is one problem with real and double precision numbers that you must take into account: rounding errors. Why do these happen?
Well, double precision and real numbers store information in the binary system. Values such as 0.2 don't have a precise binary representation. Numbers like 1/3 can't be represented with a precise decimal number at all. This is why such numbers need to be rounded. And when you add a rounded number to a rounded number, and the result is also rounded – well, you can guess what happens.
Have you noticed that we have a rather special user in our table? The one with user_id 6? It's Thumbelina. We will use the information she provided to check rounding errors.
Select her height and weight and calculate the third column called expected_zero in the following way: height - weight - height + weight. It should equal zero, right?
Remember about the condition:
WHERE user_id = 6;
