Questions
Practice and exam questions for Numerical Analysis with MATLAB.
1. Newton's Forward Difference Interpolation
Question 1.1: Concept of Interpolation and Extrapolation
- Define the mathematical concept of interpolation and contrast it with extrapolation.
- Given a discrete set of observations , state the condition under which an estimation of at is termed interpolation versus extrapolation.
- Define the forward difference operator and write down the expressions for the first, second, and -th forward differences of a sequence .
Question 1.2: Difference Operator Identities and Difference Tables
-
Prove algebraically that the second forward difference can be expressed directly in terms of functional ordinates as:
-
Construct the symbolic forward difference table for a function given at six equidistant nodes .
-
Given the following data table:
Construct the full forward difference table up to third-order forward differences .
Question 1.3: Newton's Forward Interpolation Numerical Evaluation
-
State Newton's Forward Interpolation Formula for a set of equispaced points with step size , defining the normalized variable .
-
Using the data from Question 1.2:
Calculate the value of at . Explain why Newton's forward formula is the appropriate choice for this specific query point.
Question 1.4: Polynomial Reconstruction via Forward Differences
A second-degree polynomial passes through the points , , , and .
- Construct the forward difference table for the given coordinate points.
- Determine the step size , base value , and normalized parameter .
- Use Newton's forward interpolation formula to find the explicit analytic polynomial .
Question 1.5: Higher-Order Polynomial Reconstruction & Point Estimation
Given the discrete dataset:
- Construct the forward difference table up to third differences.
- What does the constancy of the third forward difference indicate regarding the degree of the underlying polynomial?
- Find the explicit closed-form polynomial .
- Estimate the value of where the point lies on the reconstructed polynomial curve.
Question 1.6: Village Census Population Estimation
The census records of a rural village over six consecutive decades are recorded as follows:
| Year () | ||||||
|---|---|---|---|---|---|---|
| Population () |
- Construct the complete forward difference table.
- Formulate the normalized step parameter for the target year .
- Estimate the village population in the year using Newton's forward difference interpolation formula up to fifth differences.
Question 1.7: Trigonometric Table Interpolation
The values of the trigonometric sine function are tabulated below at intervals:
- Construct the forward difference table up to .
- Calculate the value of using Newton's forward difference interpolation formula.
2. Newton's Backward Difference Interpolation
Question 2.1: Backward Difference Operator and Table Construction
- Define the backward difference operator . Show the relationship between the backward difference and the forward difference .
- Define higher-order backward differences .
- Construct the symbolic backward difference table for a table with nodes .
Question 2.2: Newton's Backward Interpolation Formula
- State Newton's Backward Interpolation Formula for evaluating a function near the end of a tabular dataset.
- Define the normalized backward parameter in terms of the query point , terminal node , and uniform spacing .
- When is Newton's backward interpolation formula preferred over Newton's forward formula?
Question 2.3: Numerical Evaluation using Backward Differences
A physical experiment yields the following measurements:
- Construct the complete backward difference table up to fifth differences .
- Identify the terminal values , , and step size .
- Calculate the normalized parameter for evaluating .
- Estimate by substituting all difference terms into Newton's backward interpolation formula.
Question 2.4: Advantages, Disadvantages & Computer Science Applications
- Enumerate the key advantages and disadvantages of Newton's backward interpolation method in practical computation.
- Explain how interpolation principles are applied in:
- Digital Image Processing: Bilinear and bicubic image resizing / super-resolution.
- Game Development & Computer Graphics: Linear interpolation (
lerp) and trajectory interpolation for physics animations.
3. Lagrange Interpolation for Unequal Intervals
Question 3.1: Derivation of the Lagrange Interpolation Formula
- State the limitation of Newton's forward and backward interpolation formulas regarding node spacing.
- Given distinct points where is not necessarily uniform:
- Assume an approximating polynomial .
- Prove algebraically that the coefficients are given by:
- Formulate the classical Lagrange Interpolation Polynomial:
- State the property satisfied by the Lagrange cardinal basis functions at the nodes.
Question 3.2: Lagrange Evaluation with Unequally Spaced Nodes
Given the following table of unequally spaced observations:
- Verify that the nodes are non-uniformly spaced.
- Calculate the individual Lagrange basis coefficients for evaluating the function at .
- Compute the weighted contributions for each node.
- Determine the interpolated value .
- Formulate the Lagrange basis values to evaluate at .
4. Central Difference Interpolation Methods
Question 4.1: Gauss's Forward Central Difference Formula
-
Describe the motivation for using central difference interpolation formulas when estimating values near the center of a tabular dataset.
-
Express the displacement operator in terms of the forward difference operator .
-
Using the operator identity and expanding in terms of central ordinates, derive Gauss's Forward Interpolation Formula:
-
Given the tabular data:
- Select the central origin and compute for evaluating .
- Construct the central difference table.
- Calculate using Gauss's forward formula.
Question 4.2: Gauss's Backward Central Difference Formula
-
Derive Gauss's Backward Interpolation Formula:
-
The population of a town (in thousands) over six decades is given below:
Year Population - Designate the base year and compute for .
- Construct the central difference table up to fifth differences.
- Estimate the population in using Gauss's backward formula.
Question 4.3: Stirling's Central Difference Formula
- Explain how Stirling's Formula is obtained from the arithmetic mean of Gauss's forward and backward interpolation formulas.
- State the explicit algebraic formula for Stirling's interpolation up to fourth differences.
- Given the data points:
- Set the origin at with , and compute for .
- Construct the difference table.
- Evaluate using Stirling's formula.
Question 4.4: Bessel's Central Difference Formula
-
Formulate Bessel's Central Difference Formula, indicating which terms represent mean differences between adjacent rows.
-
State the optimal range of the normalized parameter for which Bessel's formula provides maximum accuracy.
-
The temperature (in ) of a chemical reaction is measured at regular hourly intervals:
Time (hr) Temp - Choose origin and determine for .
- Construct the central difference table.
- Evaluate the temperature using Bessel's formula.
5. Numerical Quadrature: Trapezoidal & Simpson's Rules
Question 5.1: Composite Trapezoidal Rule
- State the mathematical formula for the composite Trapezoidal Rule approximating over equal subintervals.
- Derive the step size and grid points for the limits .
- Apply the Trapezoidal Rule with to evaluate:
- List the eleven grid coordinates and their corresponding functional values rounded to four decimal places.
- Group the boundary ordinates and intermediate ordinates.
- Compute the numerical approximation .
- Compare with the exact analytical value and determine the absolute error.
Question 5.2: Simpson's 1/3 Rule
- State Simpson's 1/3 Rule for approximating .
- State the essential restriction on the number of subintervals for Simpson's 1/3 rule and explain its geometric origin (quadratic polynomial interpolation through node triplets).
- Evaluate the integral:
- Compute the step size and coordinate nodes .
- Evaluate to four decimal places.
- Group terms into boundary points, odd-indexed points (weight 4), and even-indexed points (weight 2).
- Compute the final numerical approximation.
Question 5.3: Simpson's 3/8 Rule
- State Simpson's 3/8 Rule for numerical quadrature.
- State the structural constraint on for Simpson's 3/8 rule (cubic polynomial interpolation through node quadruplets).
- Using the exact same function and grid nodes from Question 5.2 (, , , ):
- Group the ordinates according to weights: boundary ordinates, ordinates with index not divisible by 3 (weight 3), and ordinates with index divisible by 3 (weight 2).
- Substitute into Simpson's 3/8 formula and compute the numerical approximation.
- Compare the results obtained from Simpson's 1/3 rule and Simpson's 3/8 rule.
Question 5.4: Comparison of Newton-Cotes Quadrature Methods
Construct a comprehensive comparative summary table comparing:
- Trapezoidal Rule
- Simpson's 1/3 Rule
- Simpson's 3/8 Rule
Address:
- Degree of approximating interpolating polynomial .
- Interval partition constraint on .
- Local and global truncation error order in terms of step size .
- Degree of exactness (highest polynomial degree integrated without error).
6. Systems of Linear Equations: Iterative Methods
Question 6.1: Strictly Diagonally Dominant Systems
- Define what it means for a square matrix to be strictly diagonally dominant (SDD).
- State the mathematical condition for row diagonal dominance:
- Why is diagonal dominance a sufficient condition to guarantee convergence for both Gauss-Jacobi and Gauss-Seidel iterations?
- Given the system: Rearrange the equations so that the coefficient matrix satisfies strict diagonal dominance.
Question 6.2: Gauss-Jacobi Method Iterative Solution
For the rearranged diagonally dominant system from Question 6.1:
- Derive the iterative formulas for .
- Starting from the initial guess vector :
- Compute the first three iterations step-by-step showing full arithmetic.
- Present the summary table of iteration vectors up to iteration .
- Verify the stopping criterion .
Question 6.3: Gauss-Seidel Method Iterative Solution
Using the same diagonally dominant system:
- Write down the Gauss-Seidel iterative updating equations, explicitly showing how newly calculated values are immediately reused within the current iteration.
- Beginning with :
- Calculate iterations and step-by-step.
- Present the complete iteration summary table from to .
- Verify convergence under the stopping condition .
Question 6.4: Comparison: Gauss-Jacobi vs. Gauss-Seidel
- Compare the memory and computational characteristics of the Gauss-Jacobi method versus the Gauss-Seidel method.
- Explain why Gauss-Seidel typically converges roughly twice as fast as Gauss-Jacobi.
- Discuss the suitability of each method for parallel and distributed computing hardware architectures (e.g., GPU computing).