π선형대수학
안내

읽기 경로

Lang의 장별 진행 순서와 관통하는 주제

읽음 0/0 갱신 2026-06-26

How 원전 소개 — Lang organizes linear algebra. Its method is abstract and proof-first: everything is developed over an arbitrary field $K$, linear maps come before matrices, and the later chapters bring in the polynomial ring $K[t]$ to reach the deepest structure theorems.

The arc

Foundations (Ch I)

  1. 체와 벡터공간 — fields, the VS axioms, subspaces, linear combinations.
  2. 기저와 차원 — independence, bases, the exchange theorem, invariance of dimension.
  3. 합과 직합 — sums, direct sums, complements, direct products.

Matrices and maps (Ch II–IV)

  1. 행렬과 선형방정식 — matrix algebra; systems read as column-dependence relations.
  2. 선형사상 — the primary objects: kernel, image, rank–nullity.
  3. 선형사상과 행렬 — the map↔matrix isomorphism, change of basis, similarity, the diagonalization problem.

Geometry: products and duality (Ch V)

  1. 스칼라곱과 직교성 — scalar products, Gram–Schmidt, orthogonal complement, rank.
  2. 쌍대공간과 쌍선형형식 — dual space, $V\cong V^*$, bilinear/quadratic forms, Sylvester's law of inertia.

Determinants (Ch VI)

  1. 행렬식 — the unique multilinear alternating normalized form; Leibniz formula, $\det(AB)=\det A\det B$, Cramer.

Spectral theory (Ch VII–VIII)

  1. 대칭·에르미트·유니터리 연산자 — the adjoint and the three operator classes.
  2. 고윳값과 특성다항식 — eigenvectors, characteristic polynomial, similarity invariance.
  3. 스펙트럼 정리 — orthonormal diagonalization of symmetric/Hermitian/unitary operators.

Canonical forms via $K[t]$ (Ch IX–XI)

  1. 연산자의 다항식 — $f\mapsto f(A)$, minimal polynomial, Cayley–Hamilton.
  2. 삼각화와 일차분해 — fans/triangulation, primary decomposition, Schur's lemma.
  3. 조르당 표준형 — the finest canonical form for a complex operator.

Bridge outward (Ch XII)

  1. 볼록집합과 크레인–밀만 정리 — convex sets, separating/supporting hyperplanes, Krein–Milman.

The throughline

Lang repeatedly poses a representation problem and then finds the best basis:

Stage Object Best basis / form
Ch I–IV linear map matrix in chosen bases; similarity classes
Ch V scalar product orthonormal basis (dot product)
Ch VI top alternating form the determinant
Ch VIII symmetric/Hermitian operator orthonormal eigenbasis (spectral theorem)
Ch X complex operator triangular form (fans)
Ch XI complex operator Jordan normal form

The spine bends from "what is a vector space, abstractly?" to "what is the simplest matrix similar to a given one?" — answered completely by Jordan form over $\mathbb{C}$. Two ideas recur as load-bearing tools: direct-sum decomposition into invariant subspaces, and the non-degenerate scalar product that identifies $V$ with $V^*$ and builds adjoints.

Why Lang (and what it is not)

Lang is the theoretical treatment — proofs over a general field, maps before matrices, $K[t]$-module structure behind Jordan form. It deliberately offloads computation (row reduction, explicit elimination) to his more elementary Introduction to Linear Algebra. A reader wanting drills and numerical methods should pair Lang with a computational text (Anton, Strang); a reader wanting the structure and the proofs is in the right place.

See also