JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

From R^n to infinite dimensions: norms, completeness, and the leap

Vol I lived in R^n, where every basis is finite and every Cauchy sequence converges. This guide asks what happens when vectors become functions or infinite sequences — and shows that completeness, not dimension, is the property worth keeping.

Where Volume I's comfort ran out

All of Volume I took place in a finite-dimensional vector space: pick a basis, and every vector is a finite list of coordinates. Length came from a norm, angle from an inner product, and life was easy because the dimension was a single finite number. But the spaces that matter in analysis — spaces of functions, of infinite sequences, of signals — have no finite basis at all.

Consider the set of square-summable sequences: x = (x_1, x_2, x_3, ...) with sum |x_n|^2 finite. This is a genuine vector space — you can add sequences and scale them — but it is infinite-dimensional: the standard unit vectors e_1, e_2, e_3, ... are linearly independent and there are infinitely many of them. No finite basis could span it. We need new tools that survive this jump.

Completeness: the property that replaces dimension

In R^n there is a fact so basic you never named it: every Cauchy sequence converges to a limit *inside the space*. That is completeness, and in finite dimensions it comes for free. In infinite dimensions it can fail — and when it fails, calculus breaks, because limits of approximations can leak out of the space entirely.

Incompleteness in the wild — continuous functions on [-1, 1]
with the L^2 "length"  ||f|| = sqrt( integral |f|^2 ).

Approximate a step function by continuous ramps f_n:

  f_n(x) = -1            for x <= -1/n
         =  n*x          for -1/n < x < 1/n   (a steep ramp)
         = +1            for x >= 1/n

  Each f_n is CONTINUOUS.
  The sequence (f_n) is Cauchy in the L^2 norm.
  But its limit is the sign function sgn(x), which JUMPS at 0
  -> the limit is NOT continuous.

So (continuous functions, L^2 norm) is INCOMPLETE:
the Cauchy sequence converges to something outside the space.
A Cauchy sequence of continuous functions whose L^2 limit is discontinuous — the space has holes.

Two new homes: Banach and Hilbert spaces

A complete normed vector space is a Banach space: it has lengths (a norm) and no holes. If in addition the norm comes from an inner product — so you also have angles, orthogonality, and the parallelogram law — the complete space is a Hilbert space. Every Hilbert space is a Banach space; the converse fails, and that extra inner-product geometry is exactly what makes Hilbert spaces feel like an infinite-dimensional R^n.

Family tree of the spaces in this volume

  vector space
      |  add a length
  normed space            ||x||
      |  add: no holes (every Cauchy seq converges)
  BANACH space            complete + norm
      |  add: the norm comes from <x, y>
  HILBERT space           complete + inner product

Canonical examples:
  ell^2  = square-summable sequences      -> HILBERT   ||x||^2 = sum |x_n|^2
  L^2    = square-integrable functions    -> HILBERT   ||f||^2 = integral |f|^2
  ell^p, L^p  (p != 2)                    -> BANACH but NOT Hilbert
  C[0,1] with sup norm                    -> BANACH but NOT Hilbert

Test: does the parallelogram law hold?
   ||x+y||^2 + ||x-y||^2 = 2||x||^2 + 2||y||^2
   YES  -> norm comes from an inner product (Hilbert)
   NO   -> Banach only.
Banach = complete + norm; Hilbert = complete + inner product. The parallelogram law tells them apart.

From here on, ell^2 and L^2 are our running examples: both are Hilbert spaces, infinite-dimensional, and complete. The plan for this track is to carry the geometry of Vol I — orthonormal bases, projections, eigenvalues — across the infinite-dimensional divide, keeping what survives and flagging what shatters.