MHB Measurability of Open Sets in [0,1]

  • Thread starter Thread starter Fermat1
  • Start date Start date
  • Tags Tags
    Sets
Fermat1
Messages
180
Reaction score
0
I am trying to show that an open set in [0,1] is measurable, given that [0,x] is measurable set for each x in [0,1]. So I need to show (a,b) is measurable. Using the fact that measurable sets form a sigma algebra, I have managed to show that (a,b] is measurable. So (a,b+t] is measurable for any t>0. letting t ->0, can I then conclude that (a,b) is measurable? It seems a bit easy that I can just relax closed intervals to open intervals in this way.
 
Physics news on Phys.org
You are given, then, that (0, a) and (0, b) are measurable and you can write (a,b)= (0, b)\ (0, a).
 
Fermat said:
I am trying to show that an open set in [0,1] is measurable, given that [0,x] is measurable set for each x in [0,1]. So I need to show (a,b) is measurable. Using the fact that measurable sets form a sigma algebra, I have managed to show that (a,b] is measurable. So (a,b+t] is measurable for any t>0. letting t ->0, can I then conclude that (a,b) is measurable?
Not quite, there are two things wrong here. First, in a sigma-algebra, you must use countable unions and intersections. So instead of $t$ you should have $1/n$. Then you can take the intersection of the sets $\bigl(a,b+\frac1n\bigr]$. That leads us to the second thing that is wrong, namely the fact that $$\bigcap_n \bigl(a,b+\tfrac1n\bigr] = (a,b]$$: you still have an interval that is closed at the upper end!

What you need to do is to take the union of the sets $\bigl(a,b-\frac1n\bigr]$, which is $(a,b)$.
 
Back
Top