Solution to Question 2
Part (a): LU Decomposition with Partial Pivoting
We are given:
1.0 −2.5 −2.0
A = −4.0 −1.0 −0.1
3.0 2.0 1.0
Step 1: Pivoting
The largest absolute value in the first column is | − 4.0| = 4.0, so we swap Row 1 and Row 2:
0 1 0 −4.0 −1.0 −0.1
P = 1 0 0 , P A = 1.0 −2.5 −2.0
0 0 1 3.0 2.0 1.0
Step 2: Eliminate Below the First Pivot
Compute the multipliers:
1.0 3.0
m21 = = −0.25, m31 = = −0.75
−4.0 −4.0
Apply row operations R2 ← R2 − m21 R1 and R3 ← R3 − m31 R1 :
−4.0 −1.0 −0.1
0 −2.75 −2.025
0 1.25 0.925
Step 3: Eliminate Below the Second Pivot
Compute the multiplier:
1.25
m32 = ≈ −0.454
−2.75
Apply row operation R3 ← R3 − m32 R2 :
−4.0 −1.0 −0.1
U = 0 −2.75 −2.025
0 0 0.0045
Constructing L
The lower triangular factor L is built from the multipliers:
1 0 0
L = −0.25 1 0
−0.75 −0.454 1
1
Final Result: P A = LU
0 1 0 1 0 0 −4.0 −1.0 −0.1
P = 1 0 0 , L = −0.25 1 0 , U = 0 −2.75 −2.025
0 0 1 −0.75 −0.454 1 0 0 0.0045
Part (b): Significant Digits in the Solution
Given:
• The right-hand side b has 5 significant digits.
• The condition number κ(A) = 2103.2.
Method: The number of digits lost due to ill-conditioning is estimated by:
digits lost ≈ log10 κ(A)
Calculation:
log10 (2103.2) ≈ 3.32 ≈ 3 digits lost
Remaining significant digits:
remaining digits = 5 − 3 = 2 significant digits