Computer Oriented Numerical Methods with Python (S2320C1) · Continuous Evaluation Test 1 · 31 August 2026 · 30 marks · 10:10 – 11:40 am ITM (SLS) Baroda University · B.Tech all branches, odd semester 2026-27 · Semester 3. Non-programmable calculator allowed. Questions transcribed from the original paper. Solutions are written by CampusLoop and every number was checked by computer; check the method against your class notes.
Q1. Select the correct option [3]
1. The convergence of which method is prone to the starting value? (a) Regula-Falsi · (b) Newton-Raphson · (c) Bisection · (d) All Answer: (b) Newton-Raphson.
2. Find the odd one out: (a) Bisection method · (b) Bolzano's method · (c) Midpoint method · (d) Regula-Falsi method Answer: (d) Regula-Falsi. Bisection, Bolzano's method and the midpoint method are three names for the same interval-halving method; Regula-Falsi uses the chord (false position) instead of the midpoint.
3. In the successive approximation method, for rapid convergence φ′(x) should be: Answer: (b) smaller than 1.
4. Which of the following is also known as the Newton-Raphson method? (a) Chord method · (b) Tangent method · (c) Slope method · (d) Fixed-point method Answer: (b) Tangent method.
5. For a polynomial, Budan's theorem is used to estimate, within an interval, the number of: (a) Complex roots · (b) Real roots · (c) Total roots · (d) Positive and negative roots Answer: (b) Real roots (in a given interval [a, b]; Descartes' rule counts positive and negative roots).
6. The Gauss-Seidel method converges faster than Gauss-Jacobi because: Answer: (c) it uses the latest available values.
Q2. Attempt any three [9]
Q2(a) List the rules for estimating errors.
Absolute, relative and percentage error; rounding to n decimals gives error ≤ ½ × 10⁻ⁿ; n significant digits gives relative error < ½ × 10⁻ⁿ⁺¹; errors add for sums/differences; relative errors add for products/quotients. Full answer: CONMP CET-1 2025, Q2(b).
Q2(b) Find a root of x³ + x = 1 correct to three decimals using bisection.
f(x) = x³ + x − 1; f(0) = −1, f(1) = 1, so the root lies in (0, 1).
| n | a | b | m | f(m) |
|---|---|---|---|---|
| 1 | 0.00000 | 1.00000 | 0.50000 | −0.37500 |
| 2 | 0.50000 | 1.00000 | 0.75000 | +0.17188 |
| 3 | 0.50000 | 0.75000 | 0.62500 | −0.13086 |
| 4 | 0.62500 | 0.75000 | 0.68750 | +0.01245 |
| 5 | 0.62500 | 0.68750 | 0.65625 | −0.06113 |
| 6 | 0.65625 | 0.68750 | 0.67188 | −0.02483 |
| 7 | 0.67188 | 0.68750 | 0.67969 | −0.00631 |
| 8 | 0.67969 | 0.68750 | 0.68359 | +0.00304 |
| 9 | 0.67969 | 0.68359 | 0.68164 | −0.00165 |
| 10 | 0.68164 | 0.68359 | 0.68262 | +0.00069 |
| 11 | 0.68164 | 0.68262 | 0.68213 | −0.00048 |
The root lies in (0.68213, 0.68262): x ≈ 0.682 (exact 0.68233).
Q2(c) Jacobi iteration: 10x + y − z = 11.19, x + 10y + z = 28.08, −x + y + 10z = 35.61.
x = 1.23, y = 2.34, z = 3.45. Full table: CONMP CET-2 2025, Q3(d).
Q2(d) Budan's theorem: real roots of x³ − 3x² − 4x + 13 = 0 in the given intervals.
V(−3) = 3, V(−2) = V(−1) = V(0) = V(1) = V(2) = 2, V(3) = 0 → one root in [−3, −2], none in [−2, −1] to [1, 2], two in [2, 3]. Full table: CONMP CET-2 2025, Q2(e).
Q2(e) Derive the iterative formula for the Newton-Raphson method.
From Taylor's series f(x₀ + h) ≈ f(x₀) + hf′(x₀) = 0 → h = −f(x₀)/f′(x₀) → xₙ₊₁ = xₙ − f(xₙ)/f′(xₙ). Full derivation: CONMP CET-1 2025, Q2(a).
Q3. Attempt any three [15]
Q3(a) Real root of x log₁₀ x = 1.2 to a tolerance of 0.01 % using Newton-Raphson.
x₀ = 2.5 → 2.746505 → 2.740649 → 2.740646: x ≈ 2.7406. Full table: CONMP CET-1 2025, Q3(d).
Q3(b) Fourth root of 32 correct to three decimals by false position.
f(x) = x⁴ − 32; f(2) = −16, f(3) = 49.
| n | a | b | x | f(x) |
|---|---|---|---|---|
| 1 | 2.00000 | 3.00000 | 2.24615 | −6.54589 |
| 2 | 2.24615 | 3.00000 | 2.33499 | −2.27365 |
| 3 | 2.33499 | 3.00000 | 2.36448 | −0.74330 |
| 4 | 2.36448 | 3.00000 | 2.37398 | −0.23813 |
| 5 | 2.37398 | 3.00000 | 2.37700 | −0.07579 |
| 6 | 2.37700 | 3.00000 | 2.37797 | −0.02407 |
| 7 | 2.37797 | 3.00000 | 2.37827 | −0.00764 |
⁴√32 ≈ 2.378 (exact 2.37841; false position approaches slowly from one side because the right end stays fixed).
Q3(c) Root of 2x − log₁₀ x = 7 by the iteration method, three decimals.
f(x) = 2x − log₁₀ x − 7; f(3) = −1.477, f(4) = 0.398, so the root is in (3, 4). Write x = φ(x) = (7 + log₁₀ x)/2; φ′(x) = 0.4343/(2x) ≈ 0.06 < 1 ✓. Take x₀ = 3.5:
| n | xₙ | xₙ₊₁ |
|---|---|---|
| 0 | 3.50000 | 3.77203 |
| 1 | 3.77203 | 3.78829 |
| 2 | 3.78829 | 3.78922 |
| 3 | 3.78922 | 3.78927 |
x ≈ 3.789.
Q3(d) Quadratic factor x² + px + q of x⁴ − 3x³ + 20x² + 44x + 54 = 0, two Lin-Bairstow iterations.
The paper does not state the starting values; we take p₀ = q₀ = 2 (the values used for this polynomial in the class tutorial).
Iteration 1: b = (1, −5, 28, −2, 2), c = (1, −7, 40, −68). Solving 40Δp − 7Δq = −2 and (−68 + 2)Δp + 40Δq = 2 gives Δp = −0.0580, Δq = −0.0457 → p₁ = 1.9420, q₁ = 1.9543.
Iteration 2: b = (1, −4.9420, 27.6431, −0.0248, 0.0251), c = (1, −6.8840, 39.0575, −62.4212) → Δp = −0.0007, Δq = −0.0005 → p₂ = 1.9413, q₂ = 1.9538.
Quadratic factor: x² + 1.9413x + 1.9538 (the corrections are already tiny, so the factor has converged to about three decimals).
Q3(e) Gauss-Seidel: 30x − 2y + 3z = 75, x + 17y − 2z = 48, x + y + 9z = 15 (two decimals).
x = (75 + 2y − 3z)/30, y = (48 − x + 2z)/17, z = (15 − x − y)/9, using each new value immediately.
| k | x | y | z |
|---|---|---|---|
| 0 | 0 | 0 | 0 |
| 1 | 2.5000 | 2.6765 | 1.0915 |
| 2 | 2.5693 | 2.8008 | 1.0700 |
| 3 | 2.5797 | 2.7977 | 1.0692 |
| 4 | 2.5796 | 2.7976 | 1.0692 |
x = 2.58, y = 2.80, z = 1.07.