Chapter 7 — Further Reading

Annotated pointers for going deeper on matrices as linear transformations. Each entry says what to read and why it complements this chapter. Start with the 3Blue1Brown video — it animates the exact picture this chapter is built on.

Watch first (the geometry, animated)

  • 3Blue1Brown, Essence of Linear Algebra, Chapter 3: "Linear transformations and matrices" (free, YouTube). This is the single best companion to Chapter 7. Grant Sanderson animates the unit grid deforming under a matrix and shows, frame by frame, that the columns are where $\hat\imath$ and $\hat\jmath$ (our $\mathbf{e}_1, \mathbf{e}_2$) land — the same "columns are images of the basis vectors" idea, in motion. Watch Chapter 4 ("Matrix multiplication as composition") next; it previews our Chapter 8.

Core textbooks (the standard references for this book)

  • Gilbert Strang, Introduction to Linear Algebra (5th/6th ed.), §§2.1–2.3 and §1.1–1.3. Strang's "column picture" of $A\mathbf{x}$ as a combination of the columns is exactly the weighted-sum-of-columns view we derived; he hammers it harder than any other author, and his framing shaped this chapter. His matching free video lectures (MIT 18.06) Lecture 1 ("The geometry of linear equations") and the early transformation lectures are excellent. Best for: the column picture, applications, and an intuition-first tone that matches ours.

  • Sheldon Axler, Linear Algebra Done Right (4th ed.), Chapter 3 ("Linear Maps"). The rigorous, abstract counterpart to our §7.7. Axler develops linear maps before matrices and proves the "freely and uniquely determined by a basis" theorem (our Math-Major Sidebar) in full generality, with the matrix appearing as a consequence. Best for: math majors who want the proof-first, basis-centric treatment; pairs with our A-track sidebars. Note Axler famously delays determinants, so look elsewhere (Strang) for the $\det$-as-area story.

  • Stephen Boyd & Lieven Vandenberghe, Introduction to Applied Linear Algebra (VMLS) (free PDF), Chapters 6 ("Matrices") and 8 ("Linear equations"). The applied/data-science angle: matrices as the objects that act on vectors, with examples from networks, images, and least squares. Best for: CS/data-science readers who want to see matrices doing real work immediately; complements our applications and the transpose/data-layout discussion.

On specific topics in this chapter

  • Rotation matrices and orthogonal transformations. Strang §7.2 and Boyd–Vandenberghe §5 expand the rotation/reflection story; we generalize it ourselves in Chapter 21. For the angle-addition fact ($R(\alpha)R(\beta) = R(\alpha+\beta)$) seen as composition, see 3Blue1Brown Chapter 4.
  • The transpose. Its geometric meaning (the adjoint, via the dot product) is deferred in this chapter and developed in Part IV; Axler Chapter 7 and Strang §4.1 are the references when you get there. For now, any of the above texts' "definition of transpose" sections suffice.
  • Why translation is not linear / homogeneous coordinates. Picked up in Chapter 12 of this book. For a graphics-flavored treatment, see any computer-graphics text's chapter on transformations (e.g. Fundamentals of Computer Graphics, Marschner & Shirley), which builds the $3\times 3$/$4\times 4$ homogeneous matrices on exactly the $2\times 2$ foundation laid here.

Free, interactive, and visual

  • MIT 18.06 (Strang), OpenCourseWare — full free lecture videos, problem sets, and exams; the canonical free linear-algebra course, geometry-forward like this book.
  • Immersive Math, Immersive Linear Algebra (free, interactive) — Chapter on linear transformations has manipulable figures that echo our visualizer.
  • The recurring toolkit/visualizer.py in this book's repository — re-run it on your own matrices. The fastest way to build intuition for "what does this matrix do?" is to type a matrix and look.

For the applications in this chapter

  • Graphics (Case Study 1): Marschner & Shirley, Fundamentals of Computer Graphics, transformation chapters; or the transformations in video game design material for a game-development framing of the same matrices.
  • Robotics (Case Study 2): Craig, Introduction to Robotics: Mechanics and Control, Chapter 2 ("Spatial descriptions and transformations") — forward kinematics as products of rotation/transformation matrices, generalizing our 2R arm to 3D and many joints.
  • Neural networks (the transpose application): the linear-layer view in how neural networks work — every layer is a matrix transforming an activation vector, the chapter's thesis at industrial scale.

Where to go next in this book

Chapter 8 (matrix multiplication as composition, where the row-times-column rule finally earns its place and we meet non-commutativity), then Chapter 9 (the inverse — undoing a transformation) and Chapter 11 (the determinant — the area-scaling number the visualizer has been whispering all along).