I don't understand the BA part. Is that one number split between two variables? And I don't know what floating point is exactly.
Your number A is 16 bits (2 bytes). Let's represent a bit by X. So A looks like this
XXXXXXXXXXXXXXXX.
Same for B, B looks like this
XXXXXXXXXXXXXXXX.
Now, what would be BA ? It is B followed by A so BA looks like this
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXAnd you just created a 32 bits number
Floating point is basically numbers with a decimal where the period is not at the same place for each number. So you could work with 53.21 and 589.4 as opposed to
fixed point where the period is always at the same place for everyone