Lecture Notes: Compiled by Maqsood Ahmad (A.P. Maths.

) for
students of CUI, Lahore. (FA20-BSE-A&B, FA20-BSE-A).
Objective of Lecture week3:-
• Chapter2: Row Echelon Form (REF) OR Gauss-Elimination Method,
Row Reduced Echelon form (RREF) OR Gauss-Jordan Elimination Method
• Row operations (Allowed).
• Optional (Quadratic interpolation, cubic interpolation, Global Positioning system).
After studying this lecture, You are desired to do
Home Work: Do Questions 1-8 of Exercise 2.1, Questions 1-23, and 26, 27, 28
of Exercise 2.2, Questions 1-21 of Exercise 2.3, following link is extremely
helpful in this regard.
https://www.slader.com/textbook/9780132296540-elementary-linear-algebra-with-
applications-9th-edition/196/

Chapter 2: Solving Linear System

Exercise 2.1.

Solution (b):-
𝟏 𝟎 𝟏 𝟏 𝟎 𝟏 𝟏 𝟎 𝟏
−𝟏 𝟐 −𝟐 𝑹𝟐 + 𝑹𝟏 𝟎 𝟐 −𝟏 𝟎 𝟏 𝟎
𝑨=[ ] ~[ ] 𝑹𝟐𝟑 ~ [ ]
𝟎 𝟏 𝟎 𝑹𝟒 + 𝟐𝑹𝟏 𝟎 𝟏 𝟎 𝟎 𝟐 −𝟏
−𝟐 𝟕 −𝟓 𝟎 𝟕 −𝟑 𝟎 𝟕 −𝟑
𝟏 𝟎 𝟏 𝟏 𝟎 𝟎 𝟏 𝟎 𝟏
𝑹𝟑 − 𝟐𝑹𝟐 𝟎 𝟏 𝟎 𝟎 𝟏 𝟎 𝟎 𝟏 𝟎
~[ ] (−𝟏)𝑹𝟑 ~ [ ] 𝑹𝟒 + 𝟑𝑹𝟑 ~ [ ]
𝑹𝟒 − 𝟕𝑹𝟐 𝟎 𝟎 −𝟏 𝟎 𝟎 𝟏 𝟎 𝟎 𝟏
𝟎 𝟎 −𝟑 𝟎 𝟎 −𝟑 𝟎 𝟎 𝟎
𝟏 𝟎 𝟎
𝟎 𝟏 𝟎
𝑹𝟏 − 𝑹𝟑 ~ [ ]
𝟎 𝟎 𝟏
𝟎 𝟎 𝟎
 Exercise 2.2

Solution 6(c): (a) Row Echelon Form (Gauss Elimination method)

Given system can be written in compact form as

𝑨𝑿 = 𝒃
𝟐 𝟏 𝟏 −𝟐
𝒙 𝟏
𝟑 −𝟐 𝟏 −𝟔
𝒚 −𝟐
Where 𝑨 = 𝟏 𝟏 −𝟏 −𝟏 ; 𝑿 = [ ]; 𝒃 = −𝟏
𝒛
𝟔 𝟎 𝟏 −𝟗
𝒘 −𝟐
[𝟓 −𝟏 𝟐 −𝟖 ] [ 𝟑]
𝟐 𝟏 𝟏 −𝟐 | 𝟏 𝟏 𝟏 −𝟏 −𝟏 | −𝟏
𝟑 −𝟐 𝟏 −𝟔 | −𝟐 𝟑 −𝟐 𝟏 −𝟔 | −𝟐
[𝑨|𝒃] = 𝟏 𝟏 −𝟏 −𝟏 | −𝟏 𝑹𝟏𝟑 ~ 𝟐 𝟏 𝟏 −𝟐 | 𝟏
𝟔 𝟎 𝟏 −𝟗 | −𝟐 𝟔 𝟎 𝟏 −𝟗 | −𝟐
[𝟓 −𝟏 𝟐 −𝟖 | 𝟑] [𝟓 −𝟏 𝟐 −𝟖 | 𝟑]
 |
𝑹𝟐 − 𝟑 𝑹𝟏 𝟏 𝟏 −𝟏 −𝟏 −𝟏
𝟎 −𝟓 𝟒 −𝟑 | 𝟏
𝑹𝟑 − 𝟐 𝑹𝟏
~ 𝟎 −𝟏 𝟑 𝟎 | 𝟑
𝑹𝟒 − 𝟔 𝑹𝟏
𝟎 −𝟔 𝟕 −𝟑 | 𝟒
𝑹𝟓 − 𝟓 𝑹𝟏 [
𝟎 −𝟔 𝟕 −𝟑 | 𝟖]
𝟏 𝟏 −𝟏 −𝟏 | −𝟏
𝟎 −𝟓 𝟒 −𝟑 | 𝟏
𝑹𝟓 − 𝑹𝟒 ~ 𝟎 −𝟏 𝟑 𝟎 | 𝟑
𝟎 −𝟔 𝟕 −𝟑 | 𝟒
[𝟎 𝟎 𝟎 𝟎 | 𝟒]
Read Row 4 and write as 𝟎𝒙 + 𝟎𝒚 + 𝟎𝒛 + 𝟎𝒘 = 𝟒 → 𝟎 = 𝟒 (𝑵𝒐 𝑺𝒐𝒍𝒖𝒕𝒊𝒐𝒏).

A variation of Question 6(c):

𝟐𝒙 + 𝒚 + 𝒛 − 𝟐𝒘 = 𝟏
𝟑𝒙 − 𝟐𝒚 + 𝒛 − 𝟔𝒘 = −𝟐
𝒙 + 𝒚 − 𝒛 − 𝒘 = −𝟏
𝟔𝒙 + 𝒛 − 𝟗𝒘 = −𝟐
𝟓𝒙 − 𝒚 + 𝟐𝒛 − 𝟖𝒘 = −𝟏
Given system can be written in compact form as

𝑨𝑿 = 𝒃
𝟐 𝟏 𝟏 −𝟐
𝒙 𝟏
𝟑 −𝟐 𝟏 −𝟔
𝒚 −𝟐
Where 𝑨 = 𝟏 𝟏 −𝟏 −𝟏 ; 𝑿 = [ ]; 𝒃 = −𝟏
𝒛
𝟔 𝟎 𝟏 −𝟗
𝒘 −𝟐
[𝟓 −𝟏 𝟐 −𝟖 ] [−𝟏]
𝟐 𝟏 𝟏 −𝟐 | 𝟏 𝟏 𝟏 −𝟏 −𝟏 | −𝟏
𝟑 −𝟐 𝟏 −𝟔 | −𝟐 𝟑 −𝟐 𝟏 −𝟔 | −𝟐
[𝑨|𝒃] = 𝟏 𝟏 −𝟏 −𝟏 | −𝟏 𝑹𝟏𝟑 ~ 𝟐 𝟏 𝟏 −𝟐 | 𝟏
𝟔 𝟎 𝟏 −𝟗 | −𝟐 𝟔 𝟎 𝟏 −𝟗 | −𝟐
[𝟓 −𝟏 𝟐 −𝟖 | −𝟏] [𝟓 −𝟏 𝟐 −𝟖 | −𝟏]
 |
𝑹𝟐 − 𝟑 𝑹𝟏 𝟏 𝟏 −𝟏 −𝟏 −𝟏
𝟎 −𝟓 𝟒 −𝟑 | 𝟏
𝑹𝟑 − 𝟐 𝑹𝟏
~ 𝟎 −𝟏 𝟑 𝟎 | 𝟑
𝑹𝟒 − 𝟔 𝑹𝟏
𝟎 −𝟔 𝟕 −𝟑 | 𝟒
𝑹𝟓 − 𝟓 𝑹𝟏 [
𝟎 −𝟔 𝟕 −𝟑 | 𝟒]
𝟏 𝟏 −𝟏 −𝟏 | −𝟏 𝟏 𝟏 −𝟏 −𝟏 | −𝟏
𝟎 −𝟓 𝟒 −𝟑 | 𝟏 𝟎 −𝟏 𝟑 𝟎 | 𝟑
𝑹𝟓 − 𝑹𝟒 ~ 𝟎 −𝟏 𝟑 𝟎 | 𝟑 𝑹𝟐𝟑 ~ 𝟎 −𝟓 𝟒 −𝟑 | 𝟏
𝟎 −𝟔 𝟕 −𝟑 | 𝟒 𝟎 −𝟔 𝟕 −𝟑 | 𝟒
[𝟎 𝟎 𝟎 𝟎 | 𝟎 ] [ 𝟎 𝟎 𝟎 𝟎 | 𝟎]
𝟏 𝟏 −𝟏 −𝟏 | −𝟏 𝟏 𝟏 −𝟏 −𝟏 | −𝟏
𝟎 𝟏 −𝟑 𝟎 | −𝟑 𝟎 𝟏 −𝟑 𝟎 | −𝟑
𝑹𝟑 + 𝟓 𝑹𝟐
(−𝟏) 𝑹𝟐 ~ 𝟎 −𝟓 𝟒 −𝟑 | 𝟏 ~ 𝟎 𝟎 −𝟏𝟏 −𝟑 | −𝟏𝟒
𝑹𝟒 + 𝟔 𝑹𝟐
𝟎 −𝟔 𝟕 −𝟑 | 𝟒 𝟎 𝟎 −𝟏𝟏 −𝟑 | −𝟏𝟒
[𝟎 𝟎 𝟎 𝟎 | 𝟎 ] [ 𝟎 𝟎 𝟎 𝟎 | 𝟎 ]

𝟏 𝟏 −𝟏 −𝟏 | −𝟏 𝟏 𝟏 −𝟏 −𝟏 | −𝟏
𝟎 𝟏 −𝟑 𝟎 | −𝟑 𝟎 𝟏 −𝟑 𝟎 | −𝟑
𝑹𝟑
𝑹𝟒 − 𝑹𝟑 ~ 𝟎 𝟎 −𝟏𝟏 −𝟑 | −𝟏𝟒 ~ 𝟎 𝟎 𝟏 𝟑/𝟏𝟏 | 𝟏𝟒/𝟏𝟏
−𝟏𝟏
𝟎 𝟎 𝟎 𝟎 | 𝟎 𝟎 𝟎 𝟎 𝟎 | 𝟎
[𝟎 𝟎 𝟎 𝟎 | 𝟎 ] [𝟎 𝟎 𝟎 𝟎 | 𝟎 ]
We arrived at ROW ECHELN FORM (REF) and will find solution by backward
substitution.
𝟑 𝟏𝟒
𝒛+ 𝒘= … … … (𝟏)
𝟏𝟏 𝟏𝟏
𝒚 − 𝟑𝒛 = −𝟑 … … … (𝟐)
𝒙 + 𝒚 − 𝒛 − 𝒘 = −𝟏 … … … (𝟑)
3 equations and 4 unknowns (Unknown > Equatins) implies infinite many solutions
𝟏𝟒 𝟑
(𝟏) ⇒ 𝒛 = − 𝒘
𝟏𝟏 𝟏𝟏
𝟏𝟒 𝟑
Let 𝒘 = 𝒓 ∈ 𝑹, Then 𝒛 = 𝟏𝟏 − 𝟏𝟏 𝒓.

Put value of 𝒛 in (𝟐) we get

𝟒𝟐 𝟗 𝟗 𝟗
𝒚 = −𝟑 + 𝟑𝒛 = −𝟑 + − 𝒓 ⇒𝒚= − 𝒓
𝟏𝟏 𝟏𝟏 𝟏𝟏 𝟏𝟏
Put value of 𝒚, 𝒛 and 𝒘 in (𝟑) we get
𝟗 𝟗 𝟏𝟒 𝟑
𝒙 + 𝒚 − 𝒛 − 𝒘 = −𝟏 ⇒ 𝒙 = −𝒚 + 𝒛 + 𝒘 − 𝟏 ⇒ 𝒙 = − + 𝒓+ − 𝒓+𝒓−𝟏
𝟏𝟏 𝟏𝟏 𝟏𝟏 𝟏𝟏
−𝟔 𝟏𝟕
𝒙= + 𝒓
𝟏𝟏 𝟏𝟏
𝟏𝟕
−𝟔
+
𝟏𝟕
𝒓
−𝟔 𝟏𝟕 −𝟔 𝒓
𝒙 𝟏𝟏 𝟏𝟏 𝟏𝟏
𝒓 𝟏𝟏
𝟏𝟏
𝟏𝟏 −𝟗
𝒚 𝟗
−
𝟗
𝒓
𝟗 −𝟗 𝟗 𝒓
Additional Solution Set: 𝑿 = [ ] = 𝟏𝟏 = 𝟏𝟏 + 𝟏𝟏 𝒓 = 𝟏𝟏 +
𝟏𝟏
𝟏𝟏
𝒛 𝟏𝟒 𝟑
𝟏𝟒 −𝟑 𝟏𝟒
−𝟑
𝒓
𝒘 − 𝒓 𝒓 𝟏𝟏
𝟏𝟏 𝟏𝟏 𝟏𝟏 𝟏𝟏 𝟏𝟏 𝟏𝟏𝒓
[ 𝟎+𝒓 ] [𝟎] [ 𝒓 ] [𝟎] [ ]
𝟏𝟏

−𝟔 𝟏𝟕
𝟏 𝟗 𝒓 −𝟗
= [ ]+ [ ]
𝟏𝟏 𝟏𝟒 𝟏𝟏 −𝟑
𝟎 𝟏𝟏
A variation of Question 6(c): CONTINUED

Now we will work for Reduced Row Echelon Form (RREF). Proceed part (a) as
follows.
𝟏 𝟏 −𝟏 −𝟏 | −𝟏 𝟏 𝟎 𝟐 −𝟏 | 𝟐
𝟎 𝟏 −𝟑 𝟎 | −𝟑 𝟎 𝟏 −𝟑 𝟎 | −𝟑
[𝑨|𝒃]~ 𝟎 𝟎 𝟏 𝟑/𝟏𝟏 | 𝟏𝟒/𝟏𝟏 𝑹𝟏 − 𝑹𝟐 ~ 𝟎 𝟎 𝟏 𝟑/𝟏𝟏 | 𝟏𝟒/𝟏𝟏
𝟎 𝟎 𝟎 𝟎 | 𝟎 𝟎 𝟎 𝟎 𝟎 | 𝟎
[𝟎 𝟎 𝟎 𝟎 | 𝟎 ] [𝟎 𝟎 𝟎 𝟎 | 𝟎 ]
𝟏 𝟎 𝟎 −𝟏𝟕/𝟏𝟏 | −𝟔/𝟏𝟏
𝟎 𝟏 𝟎 𝟗/𝟏𝟏 | 𝟗/𝟏𝟏
𝑹𝟏 − 𝟐𝑹𝟑
~ 𝟎 𝟎 𝟏 𝟑/𝟏𝟏 | 𝟏𝟒/𝟏𝟏
𝑹𝟐 + 𝟑 𝑹𝟑
𝟎 𝟎 𝟎 𝟎 | 𝟎
[𝟎 𝟎 𝟎 𝟎 | 𝟎 ]
𝟑 𝟏𝟒 𝟏𝟒 𝟑
𝒛+ 𝒘= ⇒ 𝒛= − 𝒓, 𝒘𝒉𝒆𝒓𝒆 𝒘 = 𝒓 ∈ 𝑹
𝟏𝟏 𝟏𝟏 𝟏𝟏 𝟏𝟏
𝟗 𝟗 𝟗 𝟗
𝒚+ 𝒘= ⇒ 𝒚= − 𝒓
𝟏𝟏 𝟏𝟏 𝟏𝟏 𝟏𝟏
 𝟏𝟕 −𝟔 −𝟔 𝟏𝟕
𝒙− 𝒘= ⇒ 𝒙= + 𝒓
𝟏𝟏 𝟏𝟏 𝟏𝟏 𝟏𝟏

Another variation of Question 6(c):

𝟐𝒙 + 𝒚 + 𝒛 − 𝒘 = 𝟏
𝟑𝒙 − 𝟐𝒚 + 𝒛 − 𝟔𝒘 = −𝟐
𝒙 + 𝒚 − 𝒛 − 𝒘 = −𝟏
𝟔𝒙 + 𝒛 − 𝟗𝒘 = −𝟐
𝟓𝒙 − 𝒚 + 𝟐𝒛 − 𝟖𝒘 = −𝟏

OR

(DO YOURSELF, IMPORTANT)

Solution is given here.( Unique solution)
𝒙 −𝟔/𝟏𝟏
𝒚 𝟗/𝟏𝟏
𝑿=[ ]=[ ]
𝒛 𝟏𝟒/𝟏𝟏
𝒘 𝟎

𝒙
Solution: 𝑨𝑿 = 𝟑𝑿 where 𝑿 = [𝒚] =?
𝒛
Consider 𝑨𝑿 = 𝟑𝑿 𝒈𝒊𝒗𝒆𝒔 𝑨𝑿 − 𝟑𝑿 = 𝑶𝟑×𝟏 implies 𝑨𝑿 − 𝟑𝑿 = 𝑶𝟑×𝟏
 (𝑨 − 𝟑𝑰)𝑿 = 𝑶𝟑×𝟏

𝟏 𝟐 −𝟏 𝟑 𝟎 𝟎 𝒙 𝟎
([𝟏 𝟎 𝟏 ] − [𝟎 𝟑 𝟎]) [𝒚] = [𝟎]
𝟒 −𝟒 𝟓 𝟎 𝟎 𝟑 𝒛 𝟎
−𝟐 𝟐 −𝟏 𝒙 𝟎
𝒚
([ 𝟏 −𝟑 𝟏 ]) [ ] = [𝟎] − − − (𝟏) 𝑨𝑿 = 𝑶
𝟒 −𝟒 𝟐 𝒛 𝟎
We will work on augmented matrix to find solution of above homogeneous
system
−𝟐 𝟐 −𝟏 | 𝟎 𝟏 −𝟑 𝟏 | 𝟎
[𝑨|𝑶] = [ 𝟏 −𝟑 𝟏 | 𝟎] 𝑹𝟏𝟐 ~ [−𝟐 𝟐 −𝟏 | 𝟎]
𝟒 −𝟒 𝟐 | 𝟎 𝟒 −𝟒 𝟐 | 𝟎

𝑹𝟐 + 𝟐 𝑹𝟏 𝟏 −𝟑 𝟏 | 𝟎 𝑹 𝟏 −𝟑 𝟏 | 𝟎
𝟐
𝑹𝟑 − 𝟒 𝑹𝟏
~ [𝟎 −𝟒 𝟏 | 𝟎] ~ [𝟎 𝟏 −𝟏/𝟒 | 𝟎]
−𝟒
𝟎 𝟖 −𝟐 | 𝟎 𝟎 𝟖 −𝟐 | 𝟎
𝟏 −𝟑 𝟏 | 𝟎 𝟏 𝟎 𝟏/𝟒 | 𝟎
𝑹𝟑 − 𝟖 𝑹𝟐 ~ [𝟎 𝟏 −𝟏/𝟒 | 𝟎] 𝑹𝟏 + 𝟑 𝑹𝟐 ~ [𝟎 𝟏 −𝟏/𝟒 | 𝟎] − −(𝟐)
𝟎 𝟎 𝟎 | 𝟎 𝟎 𝟎 𝟎 | 𝟎
Now read rows of last matrix (RREF) and write equivalent system as
𝟎𝒙 + 𝟎𝒚 + 𝟎𝒛 = 𝟎
𝟏 𝒛 𝟏 −𝒛
𝒚− 𝒛=𝟎 → 𝒚= ; 𝒙+ 𝒛=𝟎 → 𝒙= ; where 𝒛 = 𝒓 ∈ 𝑹 is an
𝟒 𝟒 𝟒 𝟒
arbitrary or free variable.
−𝒓
−𝒓
𝒙 𝟒
𝟒
𝒓
−𝟏
𝒓
𝒚
𝑿 = [ ] = [ ] = 𝟒 = [ 𝟏 ] Non trivial solution.
𝒓
𝟒
𝒛 𝟒 𝟒𝒓 𝟒
𝒓 [ ] 𝟒
 𝟏 𝟏 −𝟏 | 𝟐 𝑹 − 𝑹 𝟏 𝟏 −𝟏 | 𝟐
𝟐 𝟏
Solution: [𝑨|𝒃] = [𝟏 𝟐 𝟏 | 𝟑] ~ [𝟎 𝟏 𝟐 | 𝟏 ]
𝟐 𝑹 𝟑 − 𝑹𝟏 𝟐
𝟏 𝟏 𝒂 −𝟓 | 𝒂 𝟎 𝟎 𝒂 −𝟒 | 𝒂−𝟐
Case1: If 𝒂𝟐 − 𝟒 = 𝟎 and 𝒂 − 𝟐 ≠ 𝟎 implies No solution;
Now 𝒂𝟐 = 𝟒 → 𝒂 = ±𝟐; when 𝒂 = −𝟐, then
𝟏 𝟏 −𝟏 | 𝟐
[𝟎 𝟏 𝟐 | 𝟏 ]
𝟎 𝟎 𝟎 | −𝟐 − 𝟐
𝟎𝒙 + 𝟎𝒚 + 𝟎𝒛 = −𝟒, not acceptable. Hence No solution at 𝒂 = −𝟐
Now take 𝒂 = 𝟐, then
𝟏 𝟏 −𝟏 | 𝟐
[ 𝟎 𝟏 𝟐 | 𝟏]
𝟎 𝟎 𝟎 | 𝟎
𝟎𝒙 + 𝟎𝒚 + 𝟎𝒛 = 𝟎, (This expression signals about “Infinite many solutions”)
Case2: Infinite many solution:
𝟏 𝟏 −𝟏 | 𝟐
[𝟎 𝟏 𝟐 | 𝟏 ]
𝟐
𝟎 𝟎 𝒂 −𝟒=𝟎 | 𝒂−𝟐=𝟎
Implies 𝒂𝟐 − 𝟒 = 𝟎 and 𝒂 − 𝟐 = 𝟎 both should be zero at the same time.
For 𝒂 = 𝟐 given system has Infinite many solution.
Case3: Unique Solution: For all values of 𝒂 ∈ 𝑹 other than ±𝟐 system has
Unique solution.
 𝟏 𝟏 𝟏 | 𝟐 𝟏 𝟏 𝟏 | 𝟐
𝑹 𝟐 − 𝑹𝟏
Solution: [𝑨|𝒃] = [𝟏 𝟐 𝟏 | 𝟑] ~ [𝟎 𝟏 𝟎 | 𝟏 ]
𝟐 𝑹 𝟑 − 𝑹𝟏 𝟐
𝟏 𝟏 𝒂 −𝟓 | 𝒂 𝟎 𝟎 𝒂 −𝟔 | 𝒂−𝟐
Case1: If 𝒂𝟐 − 𝟔 = 𝟎 and 𝒂 − 𝟐 ≠ 𝟎 implies No solution; 𝒂𝟐 = 𝟔 → 𝒂 = ±√𝟔
When 𝒂 = −√𝟔
𝟏 𝟏 𝟏 | 𝟐
[𝟎 𝟏 𝟎 | 𝟏 ]
𝟎 𝟎 𝟎 | −√𝟔 − 𝟐
𝟎𝒙 + 𝟎𝒚 + 𝟎𝒛 = √𝟔 − 𝟐 Invalid, No solution.
When 𝒂 = √𝟔
𝟏 𝟏 𝟏 | 𝟐
[𝟎 𝟏 𝟎 | 𝟏 ]
𝟎 𝟎 𝟎 | √𝟔 − 𝟐
𝟎𝒙 + 𝟎𝒚 + 𝟎𝒛 = √𝟔 − 𝟐, Invalid, No solution.
for 𝒂 = ±√𝟔 we have NO SOLUTION.
𝟏 𝟏 𝟏 | 𝟐
Case2: Infinite many solution: [𝟎 𝟏 𝟎 | 𝟏 ]
𝟐
𝟎 𝟎 𝒂 −𝟔=𝟎 | 𝒂−𝟐=𝟎
Implies 𝒂𝟐 − 𝟔 = 𝟎 and 𝒂 − 𝟐 = 𝟎 both should be zero at the same time.
There will be no value of 𝒂 for which given system has Infinite many solution.
Case3: Unique Solution: For all values of 𝒂 ∈ 𝑹 other than ±√𝟔 system has
Unique solution.
 𝟏 𝟐 𝟑 𝒙 𝟐
Solution: [−𝟑 −𝟐 −𝟏] [𝒚] = [𝟐] → 𝑨𝑿 = 𝒃
−𝟐 𝟎 𝟐 𝒛 𝟒
𝟏 𝟐 𝟑 | 𝟐
[𝑨|𝒃] = [−𝟑 −𝟐 −𝟏 | 𝟐] (I am going to give its solution via linear
−𝟐 𝟎 𝟐 | 𝟒
algebra toolkit) VERIFY IT
𝒛 = 𝒓 ∈ 𝑹, 𝒚 = −𝟐𝒓 + 𝟐, 𝒙 = 𝒓 − 𝟐

𝟒 𝟏 𝟑 | 𝑎 𝟐 𝟐 𝟎 | 𝑐
Solution: [𝑨|𝒃] =[A|b]= [𝟐 −𝟏 𝟑 | 𝑏 ] 𝑹𝟏𝟑 ~ [𝟐 −𝟏 𝟑 | 𝑏]
𝟐 𝟐 𝟎 | 𝑐 𝟒 𝟏 𝟑 | 𝑎

𝑹𝟏 𝟏 𝟏 𝟎 | 𝑐/2 𝑹 − 𝟐𝑹 𝟏 𝟏 𝟎 | 𝑐/2
𝟐 𝟏
~ [𝟐 −𝟏 𝟑 | 𝑏 ] ~ [𝟎 −𝟑 𝟑 | 𝑏 − 𝑐 ]
𝟐 𝑹𝟑 − 𝟒𝑹𝟏
𝟒 𝟏 𝟑 | 𝑎 𝟎 −𝟑 𝟑 | 𝑎 − 2𝑐
𝟏 𝟏 𝟎 | 𝑐
𝑹𝟑 − 𝑹𝟐 ~ [𝟎 −𝟑 𝟑 | 𝑏−𝑐 ]
(𝑏
𝟎 𝟎 𝟎 | 𝑎 − 2𝑐 − − 𝑐)
 𝟏 −𝟑 𝟐 | 𝑐 𝑹𝟐 𝟏 −𝟑 𝟐 | 𝑐
= [𝟎 −𝟑 𝟑 | 𝑏 − 3𝑐 ] ~ [𝟎 𝟏 −𝟏 | (3𝑐 − 𝑏)/3 ]
−𝟑
𝟎 𝟎 𝟎 | 𝑎−𝑏−𝑐 𝟎 𝟎 𝟎 | 𝑎−𝑏−𝑐 =0
For system to be consistent (either unique solution OR infinite many
solutions), our expression 𝑎 − 𝑏 − 𝑐 appeared in Row Echelon Form must be
zero, i.e.
𝑎−𝑏−𝑐 =0

𝟐 𝟐 𝟑 | 𝑎 𝟏 −𝟑 𝟐 | 𝑐
Solution: [𝑨|𝒃] = [𝟑 −𝟏 𝟓 | 𝑏 ] 𝑹𝟏𝟑 ~ [𝟑 −𝟏 𝟓 | 𝑏 ]
𝟏 −𝟑 𝟐 | 𝑐 𝟐 𝟐 𝟑 | 𝑎

𝑹𝟐 − 𝟑𝑹𝟏 𝟏 −𝟑 𝟐 | 𝑐
~ [𝟎 𝟖 −𝟏 | 𝑏 − 3𝑐 ]
𝑹𝟑 − 𝟐𝑹𝟏
𝟎 𝟖 −𝟏 | 𝑎 − 2𝑐
𝟏 −𝟑 𝟐 | 𝑐
𝑹𝟑 − 𝑹𝟐 ~ [𝟎 𝟖 −𝟏 | 𝑏 − 3𝑐 ]
𝟎 𝟎 𝟎 | 𝑎 − 2𝑐 − (𝑏 − 3𝑐)
𝟏 −𝟑 𝟐 | 𝑐
= [𝟎 𝟖 −𝟏 | 𝑏 − 3𝑐 ]
𝟎 𝟎 𝟎 | 𝑎−𝑏+𝑐
𝟐 | 𝑐
𝑹𝟐 𝟏 −𝟑
~ [𝟎 𝟏 −𝟏/𝟖 | (𝑏 − 3𝑐)/8 ]
𝟖
𝟎 𝟎 𝟎 | 𝑎−𝑏+𝑐 =0
For system to be consistent (either unique solution OR infinite many
solutions), our expression 𝑎 − 𝑏 + 𝑐 appeared in Row Echelon Form must be
zero, i.e. 𝑎−𝑏+𝑐 =0
For instance, take 𝑎 = 5, 𝑏 = 4, 𝑐 = −1; Also when 𝑎 = 9, 𝑏 = 3, 𝑐 = −6
Recall, we defined three elementary row operations on a matrix 𝐴:

1. Interchange two rows.

2. Multiply a row by a nonzero constant 𝑐.
3. Add a constant 𝑐 times one row to another.

𝟏 𝟐 −𝟏𝟓 −𝟏𝟎
Example: 𝑨 = [ ] 𝑹𝟏 − 𝟒𝑹𝟐 ~ [ ]
𝟒 𝟑 𝟒 𝟑
𝟏 𝟎 𝟏 −𝟒
𝑰=[ ] 𝑹𝟏 − 𝟒𝑹𝟐 𝑬 = [ ]
𝟎 𝟏 𝟎 𝟏
 𝟏 −𝟒 𝟏 𝟐 −𝟏𝟓 −𝟏𝟎
𝑬𝑨 = [ ][ ]=[ ]
𝟎 𝟏 𝟒 𝟑 𝟒 𝟑

𝑬𝒌 𝑬𝒌−𝟏 . . . 𝑬𝟑 𝑬𝟐 𝑬𝟏 𝑨 = 𝑰 (if A is non-singular)

𝑬𝒌 −𝟏 𝑬𝒌 𝑬𝒌−𝟏 . . . 𝑬𝟑 𝑬𝟐 𝑬𝟏 𝑨 = 𝑬𝒌 −𝟏 𝑰

𝑬𝒌−𝟏 −𝟏 𝑬𝒌−𝟏 . . . 𝑬𝟑 𝑬𝟐 𝑬𝟏 𝑨 = 𝑬𝒌−𝟏 −𝟏 𝑬𝒌 −𝟏 𝑰

𝑨 = 𝑬𝟏 −𝟏 𝑬𝟐 −𝟏 𝑬𝟑 −𝟏 … 𝑬𝒌−𝟏 −𝟏 𝑬𝒌 −𝟏 𝑰 − − − (𝟐. 𝟑. 𝟏) 𝑵𝒆𝒙𝒕 𝑪𝒍𝒂𝒔𝒔

Question: Write a matrix A as product of elementary matrices.

Result: Every matrix 𝑨 can be written as a product of elementary matrices

(Important question). Use (𝑨𝑩)−𝟏 = (𝑩)−𝟏 (𝑨)−𝟏

−𝟏
𝑨−𝟏 = (𝑬𝟏 −𝟏 𝑬𝟐 −𝟏 𝑬𝟑 −𝟏 … 𝑬𝒌−𝟏 −𝟏 𝑬𝒌 −𝟏 𝑰 ) = 𝑰𝑬𝒌 𝑬𝒌−𝟏 … 𝑬𝟑 𝑬𝟐 𝑬𝟏
= 𝑬𝒌 𝑬𝒌−𝟏 … 𝑬𝟑 𝑬𝟐 𝑬𝟏 𝑰 − − − (𝟐. 𝟑. 𝟐) 𝑻𝒐𝒅𝒂𝒚′ 𝒔𝒄𝒍𝒂𝒔𝒔

2.3.2 gives rise to idea of inverse of matrix using row operations.

Inverse of Matrix using row operations.
[𝑨 | 𝑰] 𝑹𝒐𝒘 𝒐𝒑𝒆𝒓𝒂𝒕𝒊𝒐𝒏𝒔 𝒕𝒐 𝒈𝒆𝒕 𝒓𝒆𝒅𝒖𝒄𝒆𝒅 𝒆𝒄𝒉𝒆𝒍𝒐𝒏𝒇𝒐𝒓𝒎 𝒐𝒇 𝑨
[𝑰 | 𝑨−𝟏 ]
Exercise 2.3
Question 11(b): Find the inverse of matrix 𝑨, using row operations.
𝟏 𝟏 𝟏 𝟏
𝟏 𝟐 −𝟏 𝟐
𝑨=[ ]
𝟏 −𝟏 𝟐 𝟏
𝟏 𝟑 𝟑 𝟐
𝟏 𝟏 𝟏 𝟏 | 𝟏 𝟎 𝟎 𝟎
𝟐 −𝟏 | 𝟎 𝟏 𝟎
[𝑨 | 𝑰] = [𝟏 𝟐 𝟎
]
𝟏 −𝟏 𝟐 𝟏 | 𝟎 𝟎 𝟏 𝟎
𝟏 𝟑 𝟑 𝟐 | 𝟎 𝟎 𝟎 𝟏
𝟏 𝟏 𝟏 𝟏 | 𝟏 𝟎 𝟎 𝟎
𝑹𝟐 − 𝑹𝟏
|
𝑹𝟑 − 𝑹𝟏 ~ [𝟎 𝟏 −𝟐 𝟏 −𝟏 𝟏 𝟎 𝟎
]
𝟎 −𝟐 𝟏 𝟎 | −𝟏 𝟎 𝟏 𝟎
𝑹𝟒 − 𝑹𝟏
𝟎 𝟐 𝟐 𝟏 | −𝟏 𝟎 𝟎 𝟏
 𝟏 𝟏 𝟏 𝟏 | 𝟏 𝟎 𝟎 𝟎
𝑹𝟑 + 𝟐𝑹𝟐 𝟎 𝟏 −𝟐 𝟏 | −𝟏 𝟏 𝟎 𝟎
~
𝑹𝟒 − 𝟐𝑹𝟐 𝟎 𝟎 −𝟑 𝟐 | −𝟑 𝟐 𝟏 𝟎
[𝟎 𝟎 𝟔 −𝟏 | 𝟏 −𝟐 𝟎 𝟏]

𝟏 𝟏 𝟏 𝟏 | 𝟏 𝟎 𝟎 𝟎
𝟎 𝟏 −𝟐 𝟏 | −𝟏 𝟏 𝟎 𝟎
𝑹𝟒 + 𝟐𝑹𝟑 ~ [ ]
𝟎 𝟎 −𝟑 𝟐 | −𝟑 𝟐 𝟏 𝟎
𝟎 𝟎 𝟎 𝟑 | −𝟓 𝟐 𝟐 𝟏
𝑹𝟑 𝟏 𝟏 𝟏 𝟏 | 𝟏 𝟎 𝟎 𝟎
−𝟑 ~ 𝟎 𝟏 −𝟐 𝟏 | −𝟏 𝟏 𝟎 𝟎
𝑹𝟒 𝟎 𝟎 𝟏 −𝟐/𝟑 | 𝟏 −𝟐/𝟑 −𝟏/𝟑 𝟎
𝟑 [𝟎 𝟎 𝟎 𝟏 | −𝟓/𝟑 𝟐/𝟑 𝟐/𝟑 𝟏/𝟑]

𝟏 𝟎 𝟑 𝟎 | 𝟐 −𝟏 𝟎 𝟎
𝟎 𝟏 −𝟐 𝟏 | −𝟏 𝟏 𝟎 𝟎
𝑹𝟏 − 𝑹𝟐 ~ [ ]
𝟎 𝟎 𝟏 −𝟐/𝟑 | 𝟏 −𝟐/𝟑 −𝟏/𝟑 𝟎
𝟎 𝟎 𝟎 𝟏 | −𝟓/𝟑 𝟐/𝟑 𝟐/𝟑 𝟏/𝟑

𝟏 𝟎 𝟎 𝟐 | −𝟏 𝟏 𝟏 𝟎
𝑹𝟏 − 𝟑𝑹𝟑 𝟎 𝟏 𝟎 −𝟏/𝟑 | 𝟏 −𝟏/𝟑 −𝟐/𝟑 𝟎
~
𝑹𝟐 + 𝟐𝑹𝟑 𝟎 𝟎 𝟏 −𝟐/𝟑 | 𝟏 −𝟐/𝟑 −𝟏/𝟑 𝟎
[𝟎 𝟎 𝟎 𝟏 | −𝟓/𝟑 𝟐/𝟑 𝟐/𝟑 𝟏/𝟑]
𝑹𝟏 − 𝟐𝑹𝟒 | 𝟕/𝟑 −𝟏/𝟑 −𝟏/𝟑 −𝟐/𝟑
𝟏 𝟏 𝟎 𝟎 𝟎
𝑹𝟐 + 𝑹𝟒 ~ 𝟎 𝟏 𝟎 𝟎 | 𝟒/𝟗 −𝟏/𝟗 −𝟒/𝟗 𝟏/𝟗
𝟑 = [𝑰 | 𝑨−𝟏 ]
𝟐
𝟎 𝟎 𝟏 𝟎 | −𝟏/𝟗 −𝟐/𝟗 𝟏/𝟗 𝟐/𝟗
𝑹𝟑 + 𝑹𝟒 [𝟎 𝟎 𝟎 𝟏 | −𝟓/𝟑 𝟐/𝟑 𝟐/𝟑 𝟏/𝟑 ]
𝟑
 𝟏 𝟐 | 𝟏 𝟎 𝟏 𝟐 | 𝟏 𝟎
Solution: [𝑨|𝑰] = [ ] 𝑹𝟐 − 𝟑𝑹𝟏 ~ [ ]
𝟑 𝟒 | 𝟎 𝟏 𝟎 −𝟐 | −𝟑 𝟏
𝑹𝟐 𝟏 𝟐 | 𝟏 𝟎
~ [ ]
−𝟐 𝟎 𝟏 | 𝟑/𝟐 −𝟏/𝟐

𝟏 𝟎 | −𝟐 𝟏
𝑹𝟏 − 𝟐𝑹𝟐 ~ [ ] = [𝑰 | 𝑨−𝟏 ]
𝟎 𝟏 | 𝟑/𝟐 −𝟏/𝟐
Matrix 𝑨 is non singular.
𝟏 𝟎 𝟏 𝟎 𝟏 𝟎
𝑹𝟐 − 𝟑𝑹𝟏 𝒐𝒏 𝑰 = [ ] 𝒈𝒊𝒗𝒆𝒔 𝑬𝟏 = [ ] 𝒕𝒉𝒆𝒏 𝑬𝟏 −𝟏 = [ ]
𝟎 𝟏 −𝟑 𝟏 𝟑 𝟏
𝑹𝟐 𝟏 𝟎 𝟏 𝟎 𝟏 𝟎
𝒐𝒏 𝑰 = [ ] 𝒈𝒊𝒗𝒆𝒔 𝑬𝟐 = [ ] 𝒕𝒉𝒆𝒏 𝑬𝟐 −𝟏 = [ ]
−𝟐 𝟎 𝟏 𝟎 −𝟏/𝟐 𝟎 −𝟐
𝟏 𝟎 𝟏 −𝟐 𝟏 𝟐
𝑹𝟏 − 𝟐𝑹𝟐 𝒐𝒏 𝑰 = [ ] 𝒈𝒊𝒗𝒆𝒔 𝑬𝟑 = [ ] 𝒕𝒉𝒆𝒏 𝑬𝟑 −𝟏 = [ ]
𝟎 𝟏 𝟎 𝟏 𝟎 𝟏
𝟏 𝟎 𝟏 𝟎 𝟏 𝟐
Verify 𝑨 = 𝑬𝟏 −𝟏 𝑬𝟐 −𝟏 𝑬𝟑 −𝟏 = [ ][ ][ ]
𝟑 𝟏 𝟎 −𝟐 𝟎 𝟏
Extra work
−𝟐 𝟏 𝟏 𝟎 𝟏 𝟎 𝟏 −𝟐 𝟏 𝟎
𝑨−𝟏 = [ ] = 𝑬𝟑 𝑬𝟐 𝑬𝟏 𝑰 = [ ][ ][ ][ ]
𝟑/𝟐 −𝟏/𝟐 −𝟑 𝟏 𝟎 −𝟏/𝟐 𝟎 𝟏 𝟎 𝟏