Eigenvalues and matrix properties
Find eigenvalues by solving det(A − λI) = 0. For an n×n matrix, eigenvalue sum equals the trace and eigenvalue product equals the determinant. A real symmetric matrix has all real eigenvalues. Diagonalisation requires n linearly independent eigenvectors.
Check yourself: Does eigenvalue sum match the trace and eigenvalue product match the determinant?
Multivariable calculus and vector calculus
Gradient gives the direction of steepest ascent. Divergence of a vector field measures net outward flux per unit volume; Stokes' theorem relates a surface integral of curl to a line integral around its boundary. Confirm orientation and surface normals before applying integral theorems.
Check yourself: Are the surface normal and boundary traversal consistently oriented?
Ordinary differential equations
For the linear first-order ODE dy/dx + P(x)y = Q(x), the integrating factor is e^(∫P dx). Second-order constant-coefficient ODEs reduce to a characteristic equation; the form of the particular solution depends on whether the forcing term duplicates a homogeneous solution.
Check yourself: Does the candidate solution satisfy both the equation and the initial conditions?
Complex analysis
A function is analytic where the Cauchy–Riemann equations hold and partial derivatives are continuous. Residues at isolated singularities determine contour integrals via the residue theorem: the integral equals 2πj times the sum of enclosed residues. Identify singularity type before computing residues.
Check yourself: Is the singularity a simple pole, a higher-order pole, or an essential singularity?
Probability and random variables
For equally likely outcomes, probability is favourable over total. Use conditional probability P(A|B) = P(A ∩ B)/P(B). Expected value is linear: E(X + Y) = E(X) + E(Y) regardless of dependence. Variance of independent variables adds: Var(X + Y) = Var(X) + Var(Y).
Check yourself: Are the events independent, or do you need a joint distribution?