University College London · FACULTY OF MATHEMATICS

MATH0005 Chap.6 Eigenvalues, Diagonalisation and Determinants

- one subject, every graph, every model, every mark
7 Chapters5-page Bible
Our own words - no uploaded lecturer files
Updated for this semester
Chapter 6 of 8 · MATH0005

Eigenvalues, Diagonalisation and Determinants

Eigenvalues, Diagonalisation and Determinants identifies directions on which a linear map acts by simple scaling. A nonzero vector v is an eigenvector when Av=lambda v. Rearranging produces a homogeneous system, so lambda is an eigenvalue exactly when lambda I-A is not invertible. Its determinant gives the characteristic polynomial, whose roots are the candidate eigenvalues.

Each eigenspace is a null space and needs its own basis calculation. Diagonalisation requires a full basis of eigenvectors, not merely a list of roots. Distinct eigenvalues produce independent eigenvectors, while repeated eigenvalues require a comparison between algebraic multiplicity and eigenspace dimension. The eigenvector columns of P must stay paired with the matching diagonal entries of D.

Determinants also record invertibility and signed volume scaling, and their row-operation rules provide an efficient calculation route. Trace and determinant offer checks on eigenvalue sums and products. The field matters because a real matrix can have nonreal roots and be diagonalizable only over the complex numbers. The chapter keeps candidate roots, eigenspace evidence and final diagonal form as separate stages.

A row-operation ledger controls determinant signs and scales, while direct eigenpair substitution verifies each preserved direction before the final matrices are assembled.

In this chapter

What this chapter covers

  • 01

    Determinant rules

  • 02

    Characteristic polynomial

  • 03

    Eigenvalues

  • 04

    Eigenspaces

  • 05

    Algebraic multiplicity

  • 06

    Geometric multiplicity

  • 07

    Diagonalisation

Worked example · free

Diagonalise an upper triangular matrix

Q [12 marks]. Let A have rows (2,1) and (0,5). Find its eigenvalues, an eigenbasis and matrices P and D with P inverse A P=D. The mark allocation shown for this eigensystem is not an official university scheme.
Both arrows lie on the same eigenline: the orange image is a scalar multiple of the purple input, so the transformation changes magnitude without rotating away Both arrows lie on the same eigenline: the orange image is a scalar multiple of the purple input, so the transformation changes magnitude without rotating away from that direction.vAv=lambda vthe direction is preserved; only scale or sign changes
Both arrows lie on the same eigenline: the orange image is a scalar multiple of the purple input, so the transformation changes magnitude without rotating away from that direction.
  • 2Use triangular form to write the characteristic polynomial (t-2)(t-5).
  • 4For lambda=2 choose eigenvector (1,0); for lambda=5 choose eigenvector (1,3).
  • 3Put the eigenvectors as columns of P and the matching values 2 and 5 on the diagonal of D.
  • 3Verify AP=PD, which proves the pairing and diagonalisation without first expanding P inverse.
P has columns (1,0) and (1,3), and D=diag(2,5). The eigenvalues are distinct, so the two eigenvectors form a basis. The displayed points are original practice weighting.
Sia tip — Label each eigenpair before assembling P and D; changing column order requires the same change on the diagonal.
Glossary

Key terms

Characteristic polynomial
The determinant of tI minus A, whose roots are the eigenvalues of A.
Diagonalizable matrix
A square matrix that has a basis consisting entirely of eigenvectors.
Determinant
A scalar measuring signed volume scaling and detecting whether a square matrix is invertible.
FAQ

Eigenvalues, Diagonalisation and Determinants FAQ

Do repeated eigenvalues prevent diagonalisation?

Not automatically. A repeated eigenvalue is compatible with diagonalisation when its eigenspace has enough independent vectors. Compare geometric multiplicity with algebraic multiplicity and check whether the total number of independent eigenvectors reaches the matrix dimension.

What is the fastest way to verify a diagonalisation?

Check that P is invertible and compute AP and PD. Their equality verifies every eigenvector column and its paired diagonal value. Trace and determinant provide additional checks because they match the sum and product of eigenvalues with multiplicity.

Why does the scalar field matter?

A real matrix may have a characteristic polynomial with no real roots, so it cannot have a real eigenbasis. Over the complex numbers those roots may appear and diagonalisation may become possible. State explicitly whether the work is over R or C.

Study strategy

Exam move

Use the same staged routine for every spectral problem. First compute the characteristic polynomial with a clearly stated sign convention and check its degree. Factor it and list roots with algebraic multiplicity. For each root, solve the null space of lambda I-A and give an eigenspace basis. Count the independent eigenvectors before declaring diagonalizability.

When enough vectors exist, place them as columns of P and place the matching eigenvalues in D in the same order. Verify AP=PD, then compare trace with the eigenvalue sum and determinant with the eigenvalue product. Practise triangular matrices for speed and repeated-root matrices for judgment.

Track determinant changes under row swaps, row scaling and row replacement, and distinguish a determinant test for invertibility from a proof of diagonalizability. Include the field in every conclusion, especially when complex roots arise.

Keep error categories for the zero vector called an eigenvector, characteristic roots used without eigenspaces, columns of P written as rows and repeated eigenvalues assumed to give repeated directions. In applications to powers, explain that diagonal form separates the action into scalar powers. If diagonalisation fails, name the exact shortage of independent eigenvectors rather than forcing a nonexistent basis.

Maintain a determinant ledger during elimination: record each row interchange, every row scaling and the unchanged status of row replacement. Recompute one small determinant by cofactor expansion to validate the ledger. For repeated characteristic roots, write algebraic and geometric multiplicity side by side and compare them before assembling P.

Include pairs of matrices with the same characteristic polynomial but different eigenspace dimensions so the limitation of eigenvalues alone becomes visible. For powers, verify the case k=1 from the proposed factorisation before using the shortcut at larger exponents.

Working through Eigenvalues, Diagonalisation and Determinants in MATH0005? Sia is AskSia’s AI Mathematics tutor — ask any MATH0005 Eigenvalues, Diagonalisation and Determinants question and get a clear, step-by-step explanation grounded in how MATH0005 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.

A+Everything unlocked
Unlocks this Bible + all 32 of your subjects at University College London, plus 1,000+ more Bibles across the AskSia library.
Sia - your MATH0005 tutor, unlimited, worked the way the exam marks it
The full 5-page Bible + practice bank with worked solutions
Chrome extension - sync your LMS so Sia knows your deadlines
Bilingual EN / Chinese on every Bible and every Sia answer
$0.99 Trial
30-day money-back · cancel in one tap · how it works