1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Homework Help: Sum and intersection of anihalator spaces

  1. Jan 25, 2010 #1
    1. The problem statement, all variables and given/known data

    prove that (U[tex]\bigcap[/tex]W)[tex]^{\circ}[/tex]=W[tex]^{\circ}[/tex]+U[tex]^{\circ}[/tex]

    First prove That

    Take any [tex]f\in (U\bigcap W)^{\circ}[/tex]
    Then it is easy to see that for any [tex] f\in (U\bigcap W) f(v) =0 [/tex]

    but since [tex] v\in U and v \in W [/tex] then [tex] f \in U^{\circ} and f \in W^{\circ}[/tex]

    So we have


    Next prove

    [tex] W^{\circ} + U ^{\circ} = span(W)^{\circ} + span(U)^{\circ} [/tex]


    [tex] S^{\circ} =Span(S)^{\circ} [/tex]

    [tex] span(W)^{\circ} + span(U)^{\circ} = span (span(W)^{\circ} \cup span(u)^{\circ} [/tex]

    By definition of addition of subspaces

    [tex] span (span(W)^{\circ} \cup span(u)^{\circ}= span (W \cup U) ^{\circ} [/tex]

    Which I am not sure of

    And after all of that, we know that [tex] span (W \cap U) ^{\circ} \subseteq span (W \cup U) ^{\circ} [/tex]

    Which proves it if I did not make a mistake? Am I correct?

  2. jcsd
  3. Jan 25, 2010 #2
    Need some more information here. What are [tex]U[/tex] and [tex]W[/tex]?
  4. Jan 25, 2010 #3
    Sorry Finite subspaces of the finite space V
  5. Jan 25, 2010 #4
    I'll assume you mean "finite-dimensional".

    In the first half, your argument is both faulty and goes in the wrong direction. You say you intend to prove [tex](U \cap W)^\circ \supset U^\circ + W^\circ[/tex], and then you give an argument that begins with "take [tex]f \in (U \cap W)^\circ[/tex]" and concludes that "[tex]f \in U^\circ + W^\circ[/tex]". This argument, if correct, would prove [tex](U \cap W)^\circ \subset U^\circ + W^\circ[/tex], not [tex]\supset[/tex] : [tex]A \subset B[/tex] means that [tex]\alpha \in A[/tex] implies [tex]\alpha \in B[/tex].

    However, the argument itself is not correct. If [tex]f \in (U \cap W)^\circ[/tex], and [tex]v \in U \cap W[/tex], then you are correct that [tex]f(v) = 0[/tex]. However, you cannot conclude from this that [tex]f \in U^\circ[/tex] on the grounds that [tex]f(v) = 0[/tex] and [tex]v \in U[/tex]. To conclude that [tex]f \in U^\circ[/tex] you would have to prove that [tex]f(v) = 0[/tex] for every [tex]v \in U[/tex], and this need not be true. What you actually want to do is give an equation [tex]f = g + h[/tex] where [tex]g \in U^\circ[/tex] and [tex]h \in W^\circ[/tex]; this proves that [tex]f \in U^\circ + W^\circ[/tex].

    In the second half, I can't understand at all what you've written. This direction, [tex]U^\circ + W^\circ \subset (U \cap W)^\circ[/tex], actually has a simpler, direct argument like the above: take [tex]f \in U^\circ + W^\circ[/tex], and prove that [tex]f \in (U \cap W)^\circ[/tex].
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook