Numerical Methods: Formula & Method
Cheat Sheet
📌 Method Selection Based on Data Structure
Situation Best Method
Unequally spaced data Newton’s Divided Difference or Lagrange
Equally spaced, point near start Newton’s Forward Difference
Equally spaced, point near end Newton’s Backward Difference
Equally spaced, point near middle Central Difference (Gauss, Bessel, Stirling)
Need symbolic polynomial Lagrange or Newton (Divided Differences)
📏 Error Orders
- Trapezoidal Rule: Global Error = O(h²)
- Simpson’s 1/3 Rule: Global Error = O(h⁴)
- Simpson’s 3/8 Rule: Global Error = O(h⁴)
- Euler’s Method: Local Error = O(h²), Global = O(h)
- RK4 Method: Local Error = O(h⁵), Global = O(h⁴)
- Taylor Series Method: Depends on derivative used, often O(hⁿ)
📐 Key Formulas
• Newton's Forward Difference Formula:
f(x) ≈ f(x₀) + pΔf(x₀) + p(p−1)/2! Δ²f(x₀) + ...
• Newton's Backward Difference Formula:
f(x) ≈ f(xn) + p∇f(xn) + p(p+1)/2! ∇²f(xn) + ...
• Newton’s Divided Difference:
f[x₀,x₁] = (f(x₁)−f(x₀))/(x₁−x₀)
f[x₀,x₁,x₂] = (f[x₁,x₂] − f[x₀,x₁])/(x₂ − x₀)
• Lagrange Interpolation Formula:
L(x) = Σ f(xᵢ) * Π (x - xⱼ)/(xᵢ - xⱼ) for j ≠ i
• Trapezoidal Rule:
∫ f(x) dx ≈ h/2 [f(x₀) + 2f(x₁) + ... + 2f(xₙ₋₁) + f(xₙ)]
• Simpson’s 1/3 Rule:
∫ f(x) dx ≈ h/3 [f(x₀) + 4f(x₁) + 2f(x₂) + 4f(x₃) + ... + f(xₙ)]
• Runge-Kutta 4th Order (RK4):
yₙ₊₁ = yₙ + (1/6)(k₁ + 2k₂ + 2k₃ + k₄)
where k₁ = hf(xₙ,yₙ), k₂ = hf(xₙ+h/2, yₙ+k₁/2), etc.