Skip to Content
PrepPath logoPrepPath

Quick Revision

High-yield formula sheet, comparison tables, and decision flowcharts for Numerical Analysis.

1. High-Yield Formula Reference

Interpolation Methods

MethodGrid RequirementTarget Point LocationCore Formula / Parameter
Newton's ForwardEquispaced (h=consth = \text{const})Near the start of table (x≈x0x \approx x_0)u=x−x0hu = \frac{x - x_0}{h}
y(x)=y0+u Δy0+u(u−1)2! Δ2y0+⋯y(x) = y_0 + u\,\Delta y_0 + \frac{u(u-1)}{2!}\,\Delta^2 y_0 + \cdots
Newton's BackwardEquispaced (h=consth = \text{const})Near the end of table (x≈xnx \approx x_n)p=x−xnhp = \frac{x - x_n}{h}
y(x)=yn+p ∇yn+p(p+1)2! ∇2yn+⋯y(x) = y_n + p\,\nabla y_n + \frac{p(p+1)}{2!}\,\nabla^2 y_n + \cdots
Lagrange FormulaUnequal or Equal intervalsAnywhere in tabular domainPn(x)=∑i=0nyi∏j≠ix−xjxi−xjP_n(x) = \sum_{i=0}^n y_i \prod_{j \ne i} \frac{x - x_j}{x_i - x_j}
Gauss ForwardEquispaced (h=consth = \text{const})Near center (0<p<0.50 < p < 0.5)p=x−x0hp = \frac{x - x_0}{h}
yp=y0+p Δy0+p(p−1)2! Δ2y−1+⋯y_p = y_0 + p\,\Delta y_0 + \frac{p(p-1)}{2!}\,\Delta^2 y_{-1} + \cdots
Gauss BackwardEquispaced (h=consth = \text{const})Near center (−0.5<p<0-0.5 < p < 0)p=x−x0hp = \frac{x - x_0}{h}
yp=y0+p Δy−1+p(p+1)2! Δ2y−1+⋯y_p = y_0 + p\,\Delta y_{-1} + \frac{p(p+1)}{2!}\,\Delta^2 y_{-1} + \cdots
Stirling FormulaEquispaced (h=consth = \text{const})Symmetric about center (−0.5≤p≤0.5-0.5 \le p \le 0.5)Mean of Gauss Forward and Gauss Backward
Bessel FormulaEquispaced (h=consth = \text{const})Center interval (0.25≤p≤0.750.25 \le p \le 0.75)Mean differences with offset origin x0+12hx_0 + \frac{1}{2}h

Numerical Quadrature (Integration)

MethodStrip ConstraintWeighting StructureLocal Truncation ErrorDegree of Exactness
Trapezoidal RuleAny integer n≥1n \ge 1h2[y0+yn+2∑i=1n−1yi]\frac{h}{2}\left[ y_0 + y_n + 2\sum_{i=1}^{n-1} y_i \right]−(b−a)h212f′′(ξ)=O(h2)-\frac{(b-a)h^2}{12} f''(\xi) = \mathcal{O}(h^2)11
Simpson's 1/3 Rulenn must be even (2∣n2 \mid n)h3[y0+yn+4∑oddyi+2∑evenyi]\frac{h}{3}\left[ y_0 + y_n + 4\sum_{\text{odd}} y_i + 2\sum_{\text{even}} y_i \right]−(b−a)h4180f(4)(ξ)=O(h4)-\frac{(b-a)h^4}{180} f^{(4)}(\xi) = \mathcal{O}(h^4)33
Simpson's 3/8 Rulenn multiple of 3 (3∣n3 \mid n)3h8[y0+yn+3∑i≠3kyi+2∑i=3kyi]\frac{3h}{8}\left[ y_0 + y_n + 3\sum_{i \ne 3k} y_i + 2\sum_{i = 3k} y_i \right]−(b−a)h480f(4)(ξ)=O(h4)-\frac{(b-a)h^4}{80} f^{(4)}(\xi) = \mathcal{O}(h^4)33

Linear Systems: Iterative Methods

Strict Diagonal Dominance (SDD):∣aii∣>∑j≠i∣aij∣∀iGauss-Jacobi Update:xi(k+1)=1aii(bi−∑j≠iaijxj(k))Gauss-Seidel Update:xi(k+1)=1aii(bi−∑j<iaijxj(k+1)−∑j>iaijxj(k))\begin{aligned} \text{Strict Diagonal Dominance (SDD):} & \quad |a_{ii}| > \sum_{j \ne i} |a_{ij}| \quad \forall i \\ \text{Gauss-Jacobi Update:} & \quad x_i^{(k+1)} = \frac{1}{a_{ii}}\left( b_i - \sum_{j \ne i} a_{ij} x_j^{(k)} \right) \\ \text{Gauss-Seidel Update:} & \quad x_i^{(k+1)} = \frac{1}{a_{ii}}\left( b_i - \sum_{j < i} a_{ij} x_j^{(k+1)} - \sum_{j > i} a_{ij} x_j^{(k)} \right) \end{aligned}

2. Decision Flowcharts (Column-Wise)

Interpolation Method Selection

Loading diagram...

Numerical Integration Method Selection

Loading diagram...

Iterative Linear System Solver Flowchart

Loading diagram...

3. Quick Revision Cheat Sheet

Summary of Key Pitfalls to Avoid

Summary of Key Pitfalls to Avoid

  • Simpson's 1/3 Rule with Odd nn: Never apply Simpson's 1/3 rule if nn is odd. It requires an even number of strips (odd number of grid points).
  • Simpson's 3/8 Rule Divisibility: Always check that nn is a multiple of 3 before applying Simpson's 3/8 rule.
  • Sign of pp in Newton's Backward Interpolation: When evaluating f(x)f(x) using Newton's backward formula with x<xnx < x_n, p=x−xnhp = \frac{x - x_n}{h} is negative. Remember that the polynomial signs alternate accordingly.
  • Diagonal Dominance Check: Before running Gauss-Jacobi or Gauss-Seidel iterations, always test row-wise diagonal dominance. If violated, re-order rows first; otherwise, iterations may diverge.

On this page