Numerical Analysis
CSIR NET Mathematical Sciences Unit 3 numerical analysis: root finding, linear systems, interpolation, quadrature and Euler and Runge-Kutta methods.
- 5
- Topics
- 106
- Min read
- 01 Iteration, Newton-Raphson and Rate of ConvergenceStart here CSIR NET numerical analysis: bisection, fixed-point iteration, Newton-Raphson, secant and regula falsi, order of convergence and multiple roots.
- 02 Gauss Elimination and Gauss-Seidel CSIR NET numerical analysis: Gauss elimination, pivoting, LU, operation counts, Jacobi and Gauss-Seidel iteration, spectral radius and convergence tests.
- 03 Finite Differences and Interpolation CSIR NET numerical analysis: difference operators, Newton forward and backward, divided differences, Lagrange, Hermite and cubic spline interpolation.
- 04 Numerical Differentiation and Integration CSIR NET numerical analysis: difference formulas for derivatives, Richardson extrapolation, trapezoidal, Simpson, Newton-Cotes, Gauss quadrature and errors.
- 05 Picard, Euler, Modified Euler and Runge-Kutta CSIR NET numerical analysis: Picard iteration, Euler and modified Euler methods, Runge-Kutta of orders 2 and 4, local and global error, stability for IVPs.
- Numerical Analysis Formula Sheet and Short Notes Revise The whole chapter on one page.
Chapter Overview
Numerical analysis sits in Unit 3 of the CSIR NET Mathematical Sciences syllabus, alongside ODEs, PDEs, calculus of variations, integral equations and classical mechanics. It is the most computational chapter of the paper: almost every question ends in a number (an iterate, a quadrature value, a spectral radius) or in a statement about an order of convergence or an error term.
That makes it a scoring chapter. The methods are few, the formulas are fixed, and the Part C statements are decided by four or five ideas used again and again: the derivative of an iteration function at its fixed point, the spectral radius of an iteration matrix, the interpolation error formula, the degree of precision of a rule, and the order of a one-step method.
The notes assume a first course in numerical methods (books such as Atkinson’s An Introduction to Numerical Analysis, Burden and Faires’ Numerical Analysis, or Jain, Iyengar and Jain’s Numerical Methods for Scientific and Engineering Computation). They state each method exactly, derive the error terms that Part C tests, and give every number checked by computation.
What You’ll Learn
By the end of this chapter you will be able to:
- Decide whether a fixed-point iteration converges, and find its order from \( g'(\alpha), g''(\alpha), \dots \)
- Run Newton-Raphson, explain its quadratic convergence and repair it at multiple roots
- Solve linear systems by Gauss elimination with pivoting, and decide Jacobi and Gauss-Seidel convergence from the spectral radius
- Build difference tables, Newton, Lagrange and Hermite interpolants and natural cubic splines, with error bounds
- Apply difference formulas for derivatives and the trapezoidal, Simpson, Newton-Cotes and Gauss rules, with their errors and degrees of precision
- Generate Picard iterates and take Euler, modified Euler and Runge-Kutta steps, and judge their order and stability
Prerequisites
- Real analysis: the intermediate and mean value theorems, Taylor’s theorem with remainder, uniform convergence (see Real Analysis)
- Linear algebra: Gaussian row reduction, determinants, eigenvalues, positive definite matrices
- First-order ODEs: existence of solutions and the standard solvable types
- Arithmetic speed: Part B numerical questions reward clean fraction work, the same skill as Part A Numerical Ability
Learning Path
graph TD
A[1. Iteration, Newton-Raphson and Rate of Convergence] --> B[2. Gauss Elimination and Gauss-Seidel]
A --> C[3. Finite Differences and Interpolation]
B --> C
C --> D[4. Numerical Differentiation and Integration]
D --> E[5. Picard, Euler, Modified Euler and Runge-Kutta]
A --> E
E --> F[Quick Revision]The order is deliberate. Page 1 sets up the language of iteration and order of convergence that pages 2 and 5 reuse. Page 3’s interpolating polynomial is what page 4 differentiates and integrates, and page 4’s quadrature rules are what page 5’s ODE methods turn out to be.
Topics in This Chapter
| Topic | What it covers | Time |
|---|---|---|
| Iteration, Newton-Raphson and Rate of Convergence | Bisection, fixed-point iteration, contraction theorem, order of convergence, Newton-Raphson, multiple roots, secant and regula falsi | 35 min |
| Gauss Elimination and Gauss-Seidel | Gauss elimination, pivoting, LU and Cholesky, operation counts, conditioning, Jacobi and Gauss-Seidel, spectral radius, diagonal dominance | 35 min |
| Finite Differences and Interpolation | Difference operators and tables, Newton forward and backward, divided differences, Lagrange, error formula, Hermite, cubic splines | 40 min |
| Numerical Differentiation and Integration | Difference formulas and their errors, Richardson extrapolation, trapezoidal, Simpson, Newton-Cotes, Romberg, Gauss-Legendre | 35 min |
| Picard, Euler, Modified Euler and Runge-Kutta | Picard iteration, Euler, Heun and midpoint methods, RK2 family, classical RK4, local and global error, stability | 35 min |
| Quick Revision | Every formula, order and error term on one page | 10 min |
Official syllabus lines and where they are covered
| Official topic line (Unit 3) | Covered on |
|---|---|
| Iteration and Newton-Raphson, rate of convergence | Iteration, Newton-Raphson and Rate of Convergence |
| Gauss elimination and Gauss-Seidel | Gauss Elimination and Gauss-Seidel |
| Finite differences; Lagrange, Hermite, spline interpolation | Finite Differences and Interpolation |
| Numerical differentiation and integration | Numerical Differentiation and Integration |
| Picard, Euler, modified Euler, Runge-Kutta | Picard, Euler, Modified Euler and Runge-Kutta |
How CSIR NET Asks It
Per the NTA information bulletin for December 2026, the Mathematical Sciences paper (code 704) is marked as follows:
| Part B | Part C | |
|---|---|---|
| Questions given | 40 | 60 |
| Questions to attempt | 25 | 20 |
| Marks per correct answer | 3 | 4.75 |
| Marks per wrong answer | −0.75 | 0 |
| Maximum | 75 | 95 |
NTA publishes no topic-wise weightage, so we make no claim about how many questions come from this chapter. The formats it suits are clear:
- Part B (single correct) rewards exact, fast computation: the second Newton iterate, a Gauss-Seidel sweep, the third difference of a cubic, a Simpson’s rule value, one RK4 step, a Picard iterate. Wrong options are usually the values from a neighbouring method (Jacobi instead of Gauss-Seidel, Euler instead of Heun) or from a dropped factor (\( h^n \) in \( \Delta^n \), \( \frac{b - a}{2} \) in Gauss quadrature).
- Part C (one or more correct) rewards structural facts: “this iteration converges linearly with constant \( \frac{4}{5} \)”, “Gauss-Seidel converges but Jacobi does not”, “this piecewise cubic is a natural spline”, “this rule has degree of precision 3”, “this two-stage method has order 2”. Every topic page ends with Level 3 practice in exactly this format, with every option justified.
For the marking scheme and the other units, see the Mathematical Sciences overview. Found an error? See About for how to report it.
Start Learning
Begin with the first topic: