Concrete infinite-dimensional vectors: sequences and functions
Objective: Give a sequence in ℓ² but not ℓ¹.
A vector need not be a short arrow or a finite column. If addition and scalar multiplication are defined component by component or point by point, an entire infinite sequence or function can be one vector; a norm then says what “small” and “close” mean.
First identify the objects: The symbol ℓᵖ (read “little ell p”) denotes sequences x=(x₁,x₂,…) with Σ|x_n|ᵖ<∞ for 1≤p<∞, using ||x||ₚ=(Σ|x_n|ᵖ)^{1/p}; ℓ∞ contains bounded sequences with ||x||∞=sup_n|x_n|. The space C[0,1] contains continuous functions on [0,1], while Lᵖ[0,1] contains measurable functions with finite p-power integral, identified when they differ only on a null set. The result to establish is: For sequences on the positive integers, if 1≤p<q<∞ then ℓᵖ⊆ℓ^q and ||x||_q≤||x||_p. Thus absolute summability implies square summability, but the converse can fail. Read each symbol as a statement about a named space, norm or topology, and convergence mode.
The worked question is: For x_n=2^{-n}, compute its ℓ¹ and ℓ² norms and explain what the two numbers measure. Begin with “The ℓ¹ norm adds absolute coordinates: Σ_{n=1}^∞2^{-n}=1.” and finish with “The first norm measures total absolute coordinate size; the second measures Euclidean energy spread across infinitely many coordinates.” Then compare the result with the nearby failure case before deciding what has actually been proved.
For sequences on the positive integers, if 1≤p<q<∞ then ℓᵖ⊆ℓ^q and ||x||_q≤||x||_p. Thus absolute summability implies square summability, but the converse can fail.
If x∈ℓᵖ, every coordinate satisfies |x_n|≤||x||ₚ; otherwise one coordinate alone would exceed the full p-power sum.
Normalize first to ||x||ₚ=1. Then |x_n|≤1, so |x_n|^q≤|x_n|ᵖ for q>p.
Sum over n to get ||x||_q^q≤1, then restore the original scale and obtain ||x||_q≤||x||ₚ.
For x_n=2^{-n}, compute its ℓ¹ and ℓ² norms and explain what the two numbers measure.
- The ℓ¹ norm adds absolute coordinates: Σ_{n=1}^∞2^{-n}=1.
- The squared ℓ² norm is Σ4^{-n}=(1/4)/(1−1/4)=1/3, so ||x||₂=1/√3.
- The first norm measures total absolute coordinate size; the second measures Euclidean energy spread across infinitely many coordinates.
Result: The same sequence is one vector in both spaces, with ||x||₁=1 and ||x||₂=1/√3. The norm changes the geometry even when the underlying coordinate list is unchanged.