You are on page 1of 2

NM grp assig Q3

One way to solve linear equation systems is to apply Gaussian elimination with partial
pivoting. This is because it prevents round-off errors that may occur when dividing a row by
an entry that is quite small in relation to the remaining row entries, this strategy is particularly
helpful to solve the variables. There are few steps that need to follow for solving systems of
linear equations when using Gaussian elimination with partial pivoting.

Consider this example,


[ 1−1 3 4−2 1−3−1 4 ] [ x1 x 2 x 3 ]=[ 13 15 8 ]

The matrix must first be written in the augmented matrix form [A|B], where A is the
coefficient matrix, and B is the column vector on the right-hand side of the equations.
[ 1−1 3 4−2 1−3−1 4∨1315 8 ]

In the first column of this matrix, select the pivot element with the largest absolute value. If
required, switch rows so that the pivot is in the first row.
For this example, the pivot element is 4, so have to swap between row 1 and row 2. Thus, we
will get:
[ 4−2 11−13−3−1 4∨1513 8 ]

Next, use the pivot row to eliminate the coefficients below the pivot in first column. Perform
row operations to make the elements below the pivot become zero.
−1 3
For this example, we need to use the equation R1 + R2 → R2 and R1 + R3 → R3 to make
4 4
the elements below the pivot become zero. We will get:

[ 4−2 1 0−
1 11
2 4
0−
5 19
2 4
∨15
37 77
4 4 ]
Repeat the steps for the following column, until the matrix is in upper triangular form.

For this example, the upper triangular form of this matrix is:

[ 4−2 1 0−
5 19
2 4
9
0 0 ∨15
5
77 27
4 5 ]
By using back-substitution, resolve the variables at the end. Solve for the variable that
corresponds to the last column starting with the last equation. Then, put in this value into the
preceding equations and keep going until all the variables are found.
For this example, we will obtain the result: x 1=2 , x 2=−2∧x 3=3

By selecting the largest elements in each column as the pivots, the algorithm lowers the
possibility of division by small numbers and increases numerical stability. In conclusion, this
is very important when solving equation systems whose coefficients have different
magnitudes.

Reference

● 1.5 gaussian elimination with partial pivoting. - University of Michigan. (n.d.). http://www-
personal.umd.umich.edu/~fmassey/math473/Notes/c1/1.5.1%20LU%20decompositions
%20with%20partial%20pivoting.pdf

● HELM. (2008). Introduction to Numerical Methods. Helping Engineers Learn Mathematics.


https://www.mub.eps.manchester.ac.uk/helm/

● Johnson, T. (n.d.). Gaussian Elimination with Partial Pivoting. Gauss elimination with partial
pivoting. https://web.mit.edu/10.001/Web/Course_Notes/GaussElimPivoting.html

● Stephen Andrilli, David Hecker. (2023). Partial pivoting. Partial Pivoting - an overview |
ScienceDirect Topics. https://www.sciencedirect.com/topics/mathematics/partial-
pivoting#:~:text=The%20most%20popular%20strategy%20is,elimination%20to%20solve
%20the%20system

● YouTube. (2009, August 6). Gaussian elimination with partial pivoting. YouTube.
https://www.youtube.com/watch?v=euIXYdyjlqo&t=204s

You might also like