Linear algebra: eigenvalues and rank
Eigenvalues satisfy det(A − λI) = 0. The trace equals the eigenvalue sum and the determinant equals the eigenvalue product for square matrices. Rank is the number of linearly independent rows (or columns); a system Ax = b is consistent when rank(A) equals rank([A|b]).
Check yourself: Have you checked whether the determinant is zero before inverting a matrix?
Calculus: multivariable and vector
A partial derivative holds other variables constant. The gradient points in the steepest-ascent direction. Green's theorem connects a line integral around a simple closed curve to a double integral over the enclosed region, assuming continuously differentiable components.
Check yourself: Are the partial derivatives continuous in the region where you apply the theorem?
Ordinary differential equations
A first-order linear ODE dy/dx + P(x)y = Q(x) has integrating factor e^(∫P dx). For second-order constant-coefficient ODEs, the characteristic equation determines the general solution form: distinct real roots, repeated roots, or complex conjugate roots each produce a different basis.
Check yourself: Does your particular solution satisfy the original equation, not just the homogeneous part?
Probability and statistics
For mutually exclusive events, P(A ∪ B) = P(A) + P(B). For independent events, P(A ∩ B) = P(A)P(B). The mean of a continuous distribution is ∫x f(x) dx over the support. Variance is E(X²) − [E(X)]², which is always non-negative.
Check yourself: Are the events mutually exclusive, independent, or neither?
Numerical methods
Newton-Raphson iteration x_{n+1} = x_n − f(x_n)/f'(x_n) requires f'(x_n) ≠ 0 and a starting point sufficiently close to the root. The trapezoidal rule approximates ∫f dx using linear segments; Simpson's 1/3 rule uses quadratic segments and requires an even number of subintervals.
Check yourself: Is the derivative non-zero at the current iterate, and is the interval count compatible with the rule?