When playing around with floating point numbers, I find a function like the 
following helps a lot to see what they look like underneath:

decode(x::Float32) = (b=bits(x); (b[1], b[2:9], b[10:32]))
decode(x::Float64) = (b=bits(x); (b[1], b[2:12], b[13:64]))

or, very soon, you will be able to use @printf "%a" x to get a hexadecimal 
representation.

s

Reply via email to