Sum and intersection of anihalator spaces

  • Thread starter Thread starter talolard
  • Start date Start date
  • Tags Tags
    Intersection Sum
Join the discussion
Registration is free. Ask a follow-up in this thread, or start your own.
3 replies · 2K views
talolard
Messages
119
Reaction score
0

Homework Statement




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


First prove That
(U[tex]\bigcap[/tex]W)[tex]^{\circ}[/tex][tex]\supseteq[/tex]W[tex]^{\circ}[/tex]+U[tex]^{\circ}[/tex]

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

(U[tex]\bigcap[/tex]W)[tex]^{\circ}[/tex][tex]\supseteq[/tex]W[tex]^{\circ}[/tex]+U[tex]^{\circ}[/tex]

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

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

because

[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?

Thanks
Tal
 
Physics news on Phys.org
Need some more information here. What are [tex]U[/tex] and [tex]W[/tex]?
 
Sorry Finite subspaces of the finite space V
 
talolard said:
Sorry Finite subspaces of the finite space V

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].