Equivalence of Definitions of Norm of Linear Functional
![]() | 2022: It has been suggested that this page or section be merged into Equivalence of Definitions of Norm of Linear Transformation. In particular: This result is a special case of that. To discuss this page in more detail, feel free to use the talk page. When this work has been completed, you may remove this instance of {{Mergeto}} from the code. |
![]() | This page has been identified as a candidate for refactoring of basic complexity. In particular: Use the TFAE template as is standard for equivalence proofs such as this. Until this has been finished, please leave {{Refactor}} in the code.
New contributors: Refactoring is a task which is expected to be undertaken by experienced editors only. Because of the underlying complexity of the work needed, it is recommended that you do not embark on a refactoring task until you have become familiar with the structural nature of pages of $\mathsf{Pr} \infty \mathsf{fWiki}$.To discuss this page in more detail, feel free to use the talk page. When this work has been completed, you may remove this instance of {{Refactor}} from the code. |
Theorem
Let $V$ be a normed vector space, and let $L$ be a bounded linear functional on $V$.
Define the following norms of $L$:
- $(1): \quad \norm L_1 = \sup \set {\size {L h}: \norm h \le 1}$
- $(2): \quad \norm L_2 = \sup \set {\size {L h}: \norm h = 1}$
- $(3): \quad \norm L_3 = \sup \set {\dfrac {\size {L h} } {\norm h}: h \in H \setminus \set {\mathbf 0} }$
- $(4): \quad \norm L_4 = \inf \set {c > 0: \forall h \in H: \size {L h} \le c \norm h}$
Then:
- $\norm L_1 = \norm L_2 = \norm L_3 = \norm L_4$
Corollary
For all $v \in V$, the following inequality holds:
- $\size {L v} \le \norm L \norm v$
![]() | This article, or a section of it, needs explaining. In particular: Define $\norm L$ and $\norm v$ You can help $\mathsf{Pr} \infty \mathsf{fWiki}$ by explaining it. To discuss this page in more detail, feel free to use the talk page. When this work has been completed, you may remove this instance of {{Explain}} from the code. |
Proof
We have:
- $\set {v \in V : \norm v = 1} \subseteq \set {v \in V : \norm v \le 1} \subseteq V$
So it follows from the definition of the supremum that
- $\norm L_2 \le \norm L_1 \le \norm L_3$
Next we show that $\norm L_2 = \norm L_3$:
\(\ds \norm L_2\) | \(=\) | \(\ds \sup \set {\norm {L v}: \norm h = 1}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds \sup \set {\norm {L v'}: v \in V \setminus \set {\mathbf 0}, \ v' = v / \norm v}\) | as $\norm {v / \norm v} = \norm v / \norm v = 1$ for all $v \in V \setminus \set {\mathbf 0}$ | |||||||||||
\(\ds \) | \(=\) | \(\ds \sup \set {\norm {\frac 1 {\norm v} L v}: v \in V \setminus \set {\mathbf 0} }\) | as $L$ is linear | |||||||||||
\(\ds \) | \(=\) | \(\ds \sup \set {\frac {\norm {L v} } {\norm v}: v \in V \setminus \set {\mathbf 0} }\) | Norm Axiom $\text N 2$: Positive Homogeneity | |||||||||||
\(\ds \) | \(=\) | \(\ds \norm L_3\) |
Therefore
- $\norm L_1 = \norm L_2 = \norm L_3$
Moreover, if $\size {L v} \le c \norm v$ for all $v \in V \setminus \set {\mathbf 0}$, then we have:
- $\norm L_3 = \sup \set {\dfrac {\size {L v} } {\norm v}: v \in V \setminus \set {\mathbf 0} } \le c$
Taking the infimum over all such $c$ this reads:
- $\norm L_3 \le \norm L_4$
Aiming for a contradiction, suppose $c_0 := \norm L_4 > \norm L_3$.
Then by the definitions of these two norms, this means that there exists $\epsilon > 0$ such that for every $v \in V \setminus \set {\mathbf 0}$:
- $\dfrac {\size {L v} } {\norm v} + \epsilon \le c_0$
But this in turn implies that for every $v \in V \setminus \set {\mathbf 0}$:
- $\size {L v} \le c_0 \norm v - \epsilon \norm v = \paren {c_0 - \epsilon} \norm v$
This contradicts the fact that $c_0$ is the least such number satisfying this inequality.
Therefore:
- $\norm L_3 = \norm L_4$
and the proof is complete.
$\blacksquare$
Sources
- 1990: John B. Conway: A Course in Functional Analysis (2nd ed.) ... (previous) ... (next): $\text{I}$ Hilbert Spaces: $\S 3.$ The Riesz Representation Theorem: Proposition $3.3$