Lecturenote 418728295text Book of Transformation Geometry
Uploaded by
Abenezer EndaleLecturenote 418728295text Book of Transformation Geometry
Uploaded by
Abenezer EndaleTEXT BOOK OF
TRANSFORMATION
GEOMETRY
Written by Begashaw Moltot (MED+MSC) 2007
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Contents
CHAPTER- 1
TRANSFORMATIONS
1.1 Revision on Mappings .............................................................................................4
1.2 Types of Mappings ...................................................................................................5
1.3 Composition of Transformations and Their Properties ..........................................12
1.4 Identity and Inverse Transformations ....................................................................15
1.5 Fixed Points of Mappings and Involution ..............................................................19
1.6 Collineations and Dilatations .................................................................................21
Problem Set 1.1 ............................................................................................................24
1.7 Definitions and examples of Transformation Groups ...........................................28
1.8 Criteria for Transformation Groups .......................................................................30
Problem Set 1.2 ............................................................................................................33
CHAPTER-2
AFFINE GEOMETRY
2.1 Introduction to Affine Spaces ................................................................................36
2.2 Geometry in Affine Space ......................................................................................39
2.3 Lines and Planes in Affine space ...........................................................................44
2.3.1 Lines in Affine Geometry ...................................................................................44
2.3.2 Planes in Affine Space ........................................................................................53
Problem Set 2.1 ............................................................................................................54
2.3.3 Collinearity in Affine Space................................................................................56
2.4 The Classical Theorems .........................................................................................59
Problem Set 2.2 ............................................................................................................66
Prepared by Begashaw M. 1
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
CHAPTER-3
ISOMETRIC (ORTHOGONAL) TRANSFORMATIONS
3.1 Introductions ..........................................................................................................68
3.2 Definition and Examples of Isometries ..................................................................69
3.3 Properties of Isometric (Orthogonal) Transformations ..........................................72
3.4 Fundamental Types of Isometric Transformations ................................................76
3.4.1 Translation...................................................................................................76
3.4.2 Reflection ....................................................................................................82
3.4.3 Rotation .......................................................................................................95
3.3.4 Glide Reflection ........................................................................................109
3. 5 Orientation and Orthogonal Transformations .....................................................116
3.5.1 Orientation of Vectors ..............................................................................116
3.5.2 Orientation of Plane Figures .....................................................................121
3.5.3 Orientation Preserving and Orientation Reversing Isometries.................123
3.6 Fixed Points of Isometries ....................................................................................130
3.7 Linear and Non-linear Isometries.........................................................................132
3.8 Representations of Orthogonal (Isometric) .........................................................136
Transformation as a Product of Reflections .............................................................136
3.8.1 Product of Reflections on Two Lines........................................................137
Case I: When the two Lines are Intersecting .....................................................137
Case II: When the two Lines are Parallel ...........................................................144
3.8.2 Product of Reflections on Three Lines......................................................152
Case-I: When the three lines are concurrent ......................................................152
Case-II: When the three lines are parallel ..........................................................152
Case-III: When the three lines are neither parallel nor concurrent ....................156
3.8.3 The Fundamental Theorems of Isometries ................................................161
3.9 Equations of Orthogonal Transformations in Coordinates ..................................165
3.10 Equations of Even and Odd Isometries ..............................................................166
3.10.1 Equations of Even Isometries .................................................................166
3.10.2 Equations of Odd Isometries ...................................................................166
3.11 Test for Type of Isometries ................................................................................171
Review Problems On Chapter-3.................................................................................175
Prepared by Begashaw M. 2
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
CHAPTER-4
SIMILARITY TRANSFORMATIONS
4.1 Introduction ..........................................................................................................178
4.2 Properties of Similarity Transformations .............................................................180
4.3 Common Types of Similarity Transformations ...................................................182
4.3.1 Isometries ..................................................................................................182
4.3.2 Homothety (Homothetic Transformations) ...............................................183
4.4 Representation of Similarity Transformations .....................................................184
4. 5 Equations of Similarity Transformations in Coordinates ...................................190
4.6 Direct and Opposite Similarities ..........................................................................191
Review Problems on Chapter-4 .................................................................................195
CHAPTER-5
AFFINE TRANSFORMATIONS
5.1 Introduction ..........................................................................................................198
5.2 Basic Properties of Affine Transformations ........................................................200
5.3 Types of Affine Transformations .........................................................................203
5.3.1 Line (Skew)-Reflections ...........................................................................203
5.3.2 Compressions ............................................................................................205
5.3.3 Shears ........................................................................................................206
5.3.4 Sililarities ..................................................................................................208
5.4 Affine Transformations and Linear Mappings .....................................................210
5.5 Matrix Representation of Affine Transformations ...............................................211
5.6 Orientation and Affine Transformations (Revisited) ...........................................217
5.7 Area and Affine Transformations ........................................................................220
5.8 Inverse of Affine Transformations .......................................................................225
Review Problems on Chapter-5 .................................................................................227
REFERENCES...........................................................................................................232
Prepared by Begashaw M. 3
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
CHAPTER- 1
TRANSFORMATIONS
1.1 Revision on Mappings
Definition: Let X and Y be nonempty sets. Then, a mapping f from X to Y is a
rule which assigns to every element x in X exactly one (unique) value f (x) in
Y , here, f (x) is called the image of x under f . The set X is said to be the
domain of f and Y is the co-domain of f . The set of all images of f is called
range of f . In this definition of mappings, the word unique (exactly one)
refers to the idea of well definedness. A rule which assigns to every element in
the domain (in X ) some value in the co domain (in Y ) is said to be a mapping
if it is well defined.
To show well-defined ness, it suffices to show that f ( x) y, f ( x) z y z .
Notation: The mapping f from X to Y is denoted symbolically by f : X Y .
Examples
1. Let g : R 2 R 2 be given by g ( x, y) (2 x,3 y). Show that g is a mapping.
Solution: Clearly g is a rule which assigns to each value in R 2 a value in R 2 .
Now, let’s show that g is well-defined. Suppose g ( x, y) (a, b) g ( x, y) (c, d )
g ( x, y ) (a, b) g ( x, y ) (c, d ) (2 x,3 y ) (a, b) (2 x,3 y ) (c, d )
2 x a , 3 y b 2 x c, 3 y d
a c b d ( a , b ) ( c, d )
This implies that the image of any point ( x, y) in R 2 is unique and hence g is
well defined and it is a mapping.
Prepared by Begashaw M. 4
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2*. Let Z be set of integers. Consider the set S { x Z : x 1 2} . Define
h : Z S by h( x) x 2 . Is h a mapping or not?
Solution: Here, S { x Z : x 1 2} { x Z : 2 x 1 2}
{ x Z : 1 x 3} { 1,0,1,2,3}
As we see, x 1 is in S . But there is no integer in the domain such that
h( x) 1. This means x 1 has no pre-image. Hence, h is not a mapping.
1.2 Types of Mappings
Definitions:
a) One-to-one (Injective) mapping: A mapping f : X Y is said to be a one-
to-one (injective) mapping if and only if f sends distinct elements of X in to
distinct elements of Y. This means x y f ( x) f ( y). In other words, f is
one to one if and only if f ( x) f ( y) x y .
b) Onto (Surjective) mapping: A mapping f : X Y is said to be onto
mapping if and only if for every point y in Y , there exists an element x in X
such that
y f (x) . Or if the image of f is the whole of Y . That is every element of Y
has at least one pre-image in X .
c) Bijective mapping: A mapping is said to be bijective if and only if it is both
one to one and onto mapping.
Examples:
1. Verify that the following mappings are one to one but not onto.
a) f : N N given by f ( x) 2 x
b) g : Z ZxZ given by g (n) (n,0)
c) h : NxN N given by h(m, n) 2 m.3n
d ) f : M 2 x 2 M 2 x 2 given by f ( A) A1
Prepared by Begashaw M. 5
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Solution:
a) i) One to one: Assume f ( x) f ( y) for any two numbers x, y in N . Then,
f ( x) f ( y) 2 x 2 y x y . So, f is one-to-one.
ii) Onto: Let x be in the co-domain of f . Particularly, y 3 (you can select
any other odd natural number as well). Then, if x N in the domain of f , such
3
that f ( x) y , then f is on to. But, f ( x) 3 2 x 3 x . The number
2
3 3
which maps to 3 under the given function is . But is not member of the
2 2
domain (natural number). So, the map is not onto.
b) i) One to one: Assume g (m) g (n) for any two integers m, n in Z . Then,
g (m) g (n) (m,0) (n,0) m n and thus m n (m,0) (n,0) .
So, f is one-to-one.
ii) Onto: For each (m, n), n 0 in the co-domain of g , we can not find a
preimage in Z , such that g (m) (m, n) because the second coordinates of all
image members is zero, or of the form (m,0) .So, the map is not onto.
c) i) One to one: Here, h(m, n) h(a, b) 2 m.3n 2 a.3b 2 ma 3bn .
Since the basis are different, we have
2 ma 3bn m a 0, b n 0 m a, b n (m, n) (a, b) .
So, f is one-to-one.
ii) Onto: For each y in the co-domain of h , we can not find a pre-image in
N N , such that h(m, n) y . For instance if we take any natural number which
is not a common multiple of 2 and 3, like 1,5,7 and so on we cannot find a pre-
image. So, the map is not onto.
d) i) One to one: Using property of inverse,
f ( A) f ( B) A1 B 1 ( A1 ) 1 ( B 1 ) 1 A B . So, f is one-to-one.
ii) Onto: For singular matrix, we cannot find a pre-image in the co-domain
because only non-singular matrices have inverses. So, the map is not onto.
Prepared by Begashaw M. 6
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2*. Let Z be set of integers. Consider the set T { x Z : x 2 3} . Define
h : T Z by h( x) x 2 . Is h a mapping? If it is a mapping, is it one to one?
Solution: Here, T { x Z : x 2 3} { x Z : 3 x 2 3}
{ x Z : 5 x 1} { 4,3,2,1,0 }
As we see, h( x) x 2 h(4) 16, h(3) 9, h(2) 4, h(1) 1, h(0) 0 .
This means for every element in T , we can get an image in Z .
Besides, every element in T , maps into a unique element in Z . Therefore, h is
a one to one mapping.
3. Verify that the following mappings are onto but not one to one.
a) f : R R given by f ( x) x b) f : ZxZ Z given by f ( x, y) x y
c) f : M 2 x 2 R given by f ( A) det( A) d ) h : R R given by h( x) e x 2
Solution:
a) Since f ( x) f ( x) , the map is not one to one.
For instance, f (2) f (2) 2 . But for every, positive real number x ,
f ( x) x x , That means there exist at least itself such that f ( x) x, x R .
Thus, the map is onto.
b) Since f (2 x, x) x, f ( x,0) x, and so on the map is not one to one.
For instance, f (4,2) f (2,0) 2 but (4,2) (2,0) . But for every, integer
x, ( x,0), f ( x,0) x 0 x .That means the map is onto.
1 2 4 3
c) For A , B , f ( A) f ( B) det A det B 1 but A B .
2 5 7 5
So. The map is not one to one. But for every, real number
x 0
x, A , f ( A) det A x .That means the map is onto.
0 1
d) For x 3, y 3, h(3) h(3) e 9 but 3 3 . So. The map is not one to one.
Now for every positive real number y if x such that h( x) y the map will
be onto. However, h( x) y e x y x 2 ln y x ln y .
2
So, y R , x ln y , h( x) e ( ln y ) 2
e ln y y . That means the map is onto.
Prepared by Begashaw M. 7
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3. Verify that the following mappings are bijective.
a) f : R 2 R 2 given by f ( x, y) (2 x, y 1) b) f : R2 R2 , f ( x, y) (3 x 1, y3 3)
c) f : R 2 R 2 given by f ( x, y) ( x y, x y) d ) f : R R given by f ( x) e x
Solution:
a) Assume f ( x, y) f ( z, w) for any two points ( x, y) and ( z, w) in R 2 . Then,
(2 x, y 1) (2 z, w 1) . But from equality of order pairs, this equality is true if and
only if 2 x 2 z
x z, y w ( x, y) ( z, w) . So, f is one-to-one.
y 1 w 1
ii) Let (a, b) R 2 be in the co-domain of f .Then, if ( x, y) R 2 in the domain of f ,
such that f ( x, y) (a, b) , then f is on to.
But, a
f ( x, y) (2 x, y 1) (a, b) 2 x a, y 1 b x , y b 1 .Thus, we can find
2
a a
( x, y ) ( , b 1) R 2 such that f ( x, y) f ( , b 1) (a, b), (a, b) R 2 .
2 2
So f is on to. Therefore, the given map is bijective.
b) Assume f ( x, y) f ( z, w) for any two points ( x, y) and ( z, w) in R2 . Then,
f ( x, y) f ( z, w) (3 x 1, y 3 1) (3 z 1, w3 1) . But from equality of order
pairs, this equality is true if and only if
3 x 1 3 z 1
3 x 1 z 1, y 3 w 3 x z, y w ( x, y ) ( z, w) . So, f is one-to-one.
y 1 w 1
3
ii) Let (a, b) R 2 be in the co-domain of f .Then, if ( x, y) R 2 in the domain of f ,
such that f ( x, y) (a, b) , then f is on to.
But, f ( x, y) (3 x 1, y 3 1) (a, b) 3 x 1 a, y 3 1 b x a 3 1, y 3 b 1 .
Since both the expressions x a 3 1, y 3 b 1 are always defined, we can
find ( x, y) (a 3 1, 3 b 1) R 2 such that
f ( x, y) f (a 3 1, 3 b 1) (a, b), (a, b) R 2 . So f is on to. Therefore, the
given map is bijective.
Prepared by Begashaw M. 8
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
c) Assume f ( x, y) f ( z, w) for any two points ( x, y) and ( z, w) in R2 .
Then, f ( x, y) f ( z, w) ( x y, x y) ( z w, z w) . But from equality of
order pairs, this equality is true if and only if x y z w
2x 2z x z .
x y z w
Again using x z in x y z w , we get y w .
Hence, f ( x, y) f ( z, w) ( x, y) ( z, w).
So, f is one-to-one.
ii) Let (a, b) R 2 be in the co-domain of f .Then, if ( x, y) R 2 in the domain of f ,
such that f ( x, y) (a, b) , then f is on to.
But, ab a b .
f ( x, y) ( x y, x y) (a, b) x y a, x y b x , y
2 2
a b a b
This means we can find ( x, y) ( , ) R 2 such that
2 2
a b a b
f ( x, y) f ( , ) (a, b), (a, b) R 2 . So f is on to.
2 2
Therefore, the given map is bijective.
d) i) Assume f ( x) f ( y) for any twonumbers x and y in R .
Then, f ( x) f ( y) e x e y x y . Hence, f ( x) f ( y) x y.
So, f is one-to-one.
ii) Now for every positive real number y if x such that h( x) y the map
will be onto. However, f ( x) y e x y x ln y .
So, y R , x ln y, f ( x) e ln y y . That means the map is onto.
Therefore, the given map is bijective.
Prepared by Begashaw M. 9
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Transformation Mappings :
Definition: Transformation is a one-to-one mapping from a set X onto itself.
In other words, the map f : X X is said to be a transformation if and only if it
is one to one and onto. This means that for every point P in the domain there is
a unique point Q such that f ( P) Q and conversely, for every point R in the
range there is a unique point S in the domain such that f (S ) R .
Examples
1. Let f : R R be given by f ( x) ax, a 0, a R . Show that f is a
transformation.
Solution: To show that f ( x) ax is a transformation, we need to show that it is
one to one and onto.
i) Assume f ( x) f ( y) . Then, ax ay x y , since a 0 . So, f is one-to-one
ii) Let y R be in the range of f . Then, if x R in the domain of f , such that
f ( x) y , then f is on to. But for a 0 , we can find x y / a R such that
f ( x) f ( y / a) y . So f is on to. Hence, f is a transformation.
2. Verify that the following mappings are transformations.
a) g : R 2 R 2 given by g ( x, y) ( x y 1, x y 1)
b) f : R 2 R 2 given by f ( x, y) ( x, x y)
Solution:
a) i) Assume g ( x, y) g ( z, w) for any two points ( x, y) and ( z, w) in R 2 . Then,
( x y 1, x y 1) ( z w 1, z w 1) .
But from equality of order pairs, this equality is true if and only if
x y 1 z w 1 . This gives ( x, y) ( z, w) So, g is one-to-
2 x 2 z x z, y w
x y 1 z w 1
one.
ii) Let (a, b) R 2 be in the co-domain of g .Then, if ( x, y) R 2 in the domain of
g , such that g ( x, y) (a, b) , then g is on to.
Prepared by Begashaw M. 10
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
But, g ( x, y ) ( x y 1, x y 1) (a, b)
x y 1 a ab a b2
2x a b x , y
x y 1 b 2 2
Thus, we can find a b a b2 such that g ( x, y) (a, b) . So g is on to.
( x, y) ( , ) R2
2 2
Therefore, the given map g is a transformation.
b) Assume f ( x, y) f ( z, w) for any two points ( x, y) and ( z, w) in R2 .
Then, f ( x, y) f ( z, w) ( x, x y) ( z, z w) .
But from equality of order pairs, this equality is true if and only if
x z
x z, y w ( x, y) ( z, w) .
x y z w
So, f is one-to-one.
ii) Let (a, b) R 2 be in the co-domain of f .Then, if ( x, y) R 2 in the domain of f ,
such that f ( x, y) (a, b) , then f is on to. But,
f ( x, y) ( x, x y) (a, b) x a, x y b y a b .
This means we can find ( x, y) (a, a b) R 2 such that
f ( x, y) f (a, a b) (a, b), (a, b) R 2 . So f is on to.
Therefore, the given map is a transformation.
Equality of Transformations: Two transformations f and g on the same set
from X to X are said to be equal if and only if they have the same value for
each x in X . That is, f g f ( x) g ( x), x X .
Examples:
1. Let f and g be transformations on R 2 given by f ( x, y) (2ax 5 3,4 y) and
g ( x, y) (6 x 5 2b,4 y) . If f g , find the constants a and b .
Solution: By definition of equality,
f g f ( x, y ) g ( x, y ), ( x, y ) R 2
(2ax 5 3,4 y ) (6 x 5 2b,4 y )
2ax 5 3 6 x 5 2b
2a 6,3 2b a 3, b 3 / 2
Prepared by Begashaw M. 11
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2. Let f and g be transformations on the whole R 2 such that f (1,3) (3,10)
and g ( x, y) (a x,2 by) . If f g , find the constants a and b .
Solution: By definition of equality, f g f ( x, y) g ( x, y), ( x, y) R 2 .
Partticularly, g (1,3) f (1,3) (a 1,2 3b) (3,10)
a 1 3, 2 3b 10 a 2, b 4
1.3 Composition of Transformations and Their Properties
Let f : X Y and g : Y Z be mappings. Then for each x X , f ( x) Y .
Thus, there exists y Y such that f ( x) y . Besides, as g is a mapping from
Y to Z for each y Y , there exists z Z such that g ( y) z . Thus,
g ( y) g ( f ( x)) z which makes sense to write g ( f ( x)) as a mapping from X to
Z . This mapping is evaluated by applying f first on the elements of X
followed by g . This is defined as g f ( x) g ( f ( x)) for each x X . So, the
mapping gof (" f follwed by g" ) is called the composition mapping. In such
cases, one has to remember that the range of the first mapping is a subset of the
domain of the second mapping.
In particular, composition of transformation is defined as ( g f )( x) g ( f ( x))
where f and g are transformations on the same set X .
Proposition 1.1: Composition of mappings is associative
If h : X Y , g : Y Z and f : Z W are mappings, then the compositions
( f g ) h and f ( g h) represent the same mapping from X in to W . That is
( f g ) h f ( g h) . Particularly, ( f g ) h f ( g h) holds if f , g and h
are transformations on the same set X .
Proof: Since the domain of h is X , by definition of composition of mappings
we can see that the domain of ( f g ) h is also X . But, the domain of
f ( g h) is the same as the domain of g h and the domain of g h is X .
Hence, the domain of ( f g ) h is the same as that of f ( g h) , that is, X .
Prepared by Begashaw M. 12
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
So, ( f g ) h and f ( g h) are mappings with the same domain.
Furthermore, for any x X ,
(( f g ) h)( x) ( f g )(h( x)) f ( g (h( x))), x X
Also, ( f ( g h))( x) f (( g h)( x)) f ( g (h( x))), x X
Hence, by definition of equality of mappings ( f g ) h f ( g h) .
Note: Composition of two transformations need not be commutative. That is
even though both f g and g f exists and have the same domain and co-
domain, f g and g f may not be equal.
Examples:
1. Let f : R R, g : R R be defined by f ( x) x 3 and g ( x) 2 x 5 .
Solution: Here, ( f g )( x) f ( g ( x)) f (2 x 5) 2 x 2 and
( g f )( x) g ( f ( x)) g ( x 3) 2 x 1.
Thus, 2 x 2 2 x 1 f g g f .
Hence, composition of mappings is not commutative.
2. Let f : R R, g : R R be defined by f ( x) 3x 8, g ( x) 2 x k . Find the
value of the constant k such that g f f g .
Solution: Here, by definition of equality of transformations,
g f f g ( g f )( x) ( f g )( x), x R
6 x 16 k 6 x 3k 8
2k 8 k 4
3. Let f ( x, y) (ax 8,3 y 5) and g ( x, y) (7 x,4 y b) be the transformations.
If ( g f )( x, y) (14 x 8,12 y 23 b) , then, find the constants a and b .
Solution:
Here, ( f g )( x, y ) (14 x 8,12 y 23 b)
f ( g ( x, y )) (14 x 8,12 y 23 b)
f (7 x,4 y b) (14 x 8,12 y 23 b)
(7ax 8,12 y 3b 5) (14 x 8,12 y 23 b)
7ax 8 14 x 8, 12 y 3b 5 12 y 23 b
7a 14, 3b 5 23 b
a 2, b 7
Prepared by Begashaw M. 13
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Proposition 1.2: The composition of transformations on the same set are again
transformations.
Proof: Let f : X X , g : X X be any two transformations on set X . We
need to show that f g is also a transformation. That means we need to verify
that f g is one to one and onto. By definition of composition f g is a
mapping from X into X . To show f g is one to one, let x and y be arbitrary
elements of X such that f g ( x) f g ( y).
Then, f g ( x) f g ( y) f ( g ( x)) f ( g ( y)) g ( x) g ( y) x y (because
both f and g are one to one). Thus, f g is one to one.
To show f g is onto, let t be any element in X (considering X as co-domain
of f ), since f is onto there exists an element y in X ( in the domain) such that
f ( y) t .
Again, g is onto corresponding to the element y in X there is an element x in
X suchthat g ( x) y . As a result, ( f g )( x) f ( g ( x)) f ( y) t .
Thus, f g is onto. Hence, we have got that f g is one to one and onto on the
set X . Therefore, f g is a transformation whenever f and g are
transformations on X .
Prepared by Begashaw M. 14
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1.4 Identity and Inverse Transformations
Definition: A transformation from a set X into X denoted by i is said to be
identity transformation if and only if i( x) x, x X .
Any two transformations f and g from X to X are said to be inverse of each
other if both g f and f g are identity transformations.
That is ( g f )( x) ( f g )( x) i( x) x, x X , then f is called the inverse of
g and g is called the inverse of f . We denote the inverse of a transformation f
by f 1 (Read as “the inverse of f ” or f inverse).
1
Example: Verify that f ( x) 5x 20 and g ( x) x 4 are inverse of each
5
other.
Solution: Using the definition, we have
1 1
( f g )( x) f ( g ( x)) f x 4 5 x 4 20 x 20 20 x and
5 5
1
( g f )( x) g ( f ( x)) g (5 x 20) (5 x 20) 4 x 4 4 x
5
Thus, ( g f )( x) ( f g )( x) i( x) x, x X . Hence, f 1 g and g 1 f .
Remark: Any mappings which have inverse are called invertible mappings.
Only these mappings which are bijective have inverses. That means a mapping
is invertible if and only if it is both one to one and onto. Thus, any
transformation has an inverse because transformations are one to one and onto
mappings on the same set.
Prepared by Begashaw M. 15
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Finding Inverse of a transformation: Now let’s see how can we find the
inverse of transformations. Since every transformation f is bijective, its
inverse denoted by f 1 always exists. But there is no hard and fast rule on how
to find f 1 from the formula of f . Any way, one can use the following hints on
how to find f 1 whenever the formula of f is given. Let f : S S be a
transformation such that Y f (X ) . Then, to find f 1 :
Step-1: Interchange X and Y in the formula of f
Step-2: Solve for Y (for coordinates of Y ) in terms of X (coordinates of X ).
Step-3: Equate f 1 ( X ) Y from Y f (X ) . That will be the formula of f 1 .
Examples:
1. Find the inverses of the following transformations
3e x 2
a) f : R R given by f ( x) 3 x 1 b) f : R R given by f ( x)
ex 4
1 x
c) g : R 2 R 2 , g ( x, y ) (2 x 1, y 5) d ) f : R R, f ( x) ln
1 x
Solution:
a) y f ( x) 3 x 1
Step-1: Interchange x and y .That is x 3 y 1
Step-2: Solve for y in terms of x .That is
3 y 1 x 3 y x 1 y ( x 1) 3 x 3 3x 2 3x 1
Step 3: Equate the value of y obtained in step 2 with f 1 ( x) .
That is f 1 ( x) x 3 3x 2 3x 1 .
b) Step-1: Interchange x and y .
3e x 2 3e y 2
That is y f ( x) x
ex 4 ey 4
Step-2: Solve for y in terms of x .That is
3e y 2 4x 2 4x 2
x 3e y 2 x(e y 4) e y (3 x) 4 x 2 e y y ln
e 4
y
3 x 3 x
Step-3: Equate the value of y in step-2 with f 1 ( x) .
4x 2
That is f 1 ( x) ln
3 x
Prepared by Begashaw M. 16
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
c) For g ( x, y) (2 x 1, y 5) ,let Y g ( X ) where X ( x, y), Y ( z, w) .
Step-1: Interchange coordinates of X and Y:
g (Y ) X g ( z, w) ( x, y) (2 z 1, w 5) ( x, y)
Step-2: Solve for coordinates of Y in terms of coordinates of X .That is
x 1
(2 z 1, w 5) ( x, y) 2 z 1 x, w 5 y z , w y 5
2 2
Step-3: Equate the coordinates of Y obtained in step 2 with f 1 ( x, y) .
x 1
Hence, g 1 ( X ) Y g 1 ( x, y) ( z, w) g 1 ( x, y) ( , y 5)
2 2
x 1
Thus, g ( x, y) (2 x 1, y 5) g 1 ( x, y) ( , y 5)
2 2
x3
2. Let f ( x) . Find x such that f 1 ( x) 2 .
x2 1
Solution: From the definition of inverse, we have that f 1 ( y) x y f ( x)
Thus, f 1 ( x) 2 x f (2) x 8 / 5 .
Preposition 1.3: The inverse of a transformation is unique. Besides,
( f 1 ) 1 f .
Proof: Let f be a transformation whose inverses are g and h . That is f 1 g
and f 1 h . We need to show g h . Here, f 1 g f g g f i and
f 1 h f h h f i . But, g i g (h f ) g h ( f g ) h i h
(Because composition is associative as well as g and h are inverses of f ).
From this we can conclude that the inverse of a transformation is unique.
Examples:
1 a
1. Let f ( x) x 2 . Find the value of a if h( x) x 6 is the inverse of f .
3 4
1
Solution: Here, f ( x) x 2 f 1 ( x) 3x 6 . Then, from the uniqueness of
3
a a
inverse, we have x 6 3x 6 3 a 12 .
4 4
Prepared by Begashaw M. 17
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2. Let ( g f ) 1 ( x, y) h( x, y) where g ( x, y) (1 x, y), h( x, y) ( x 1, y 3).
Then, find the formula for f ( x, y) .
Solution: Let’s apply the property ( f 1 ) 1 f .
Here, ( g f ) 1 ( x, y ) h( x, y )
[( g f ) 1 ]1 ( x, y ) h 1 ( x, y )
( g f )( x, y ) h 1 ( x, y ) [ Using the property[( g f ) 1 ]1 g f ]
g 1 ( g f )( x, y ) g 1 h 1 ( x, y ), [ Operate both sides by g 1 ]
f ( x, y ) ( g 1 h 1 )( x, y ) [ Using the property g 1 g i ]
But h( x, y) ( x 1, y 3) h 1 ( x, y) ( x 1, y 3) and
g ( x, y ) (1 x, y) g 1 ( x, y) (1 x, y)
Therefore, f ( x, y) ( g 1 h 1 )( x, y) g 1 (h 1 ( x, y)) g 1 ( x 1, y 3) (2 x,3 y
Preposition 1.4: The inverse of a transformation is again a transformation.
Proof: Let f : X X be any transformation on set X . Then, f 1 also exists as
1
f is bijective. Now, we need to show f is also a transformation.
(i ) One-to-ones: Let a and b be arbitrary elements in X . Since f is bijective,
there exists unique x, y X , f ( x) a, f ( y) b .
But, f ( x) a, f ( y) b x f 1 (a), y f 1 (b) .
Now, assume that f 1 (a) f 1 (b) .
But, f 1 (a) f 1 (b) x y f ( x) f ( y) a b
Thus, f 1 (a) f 1 (b) a b . Hence, f 1 is one to one.
(ii ) Onto ness: Let x be arbitrary element in X . Since f is onto, f ( x) X , so,
for every x X , f ( x) X , f 1 ( f ( x)) x . Hence, f 1 is onto. Therefore, from
1
(i ) and (ii ) , whenever f is a transformation on set X and so is f .
Proposition 1.5: (Reverse Law of Inverse)
For any two transformations f and g , ( f g ) 1 g 1 f 1 .
Prepared by Begashaw M. 18
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1.5 Fixed Points of Mappings and Involution
Definition: Let f : S T be a mapping. Then, a point x0 S (in the domain of
f ) is said to be a fixed point of f if and only if f ( x0 ) x0 . Generally, the set of
fixed points of a mapping f is the set given by S {x : f ( x) x} .
Examples: Determine the fixed points of the following mappings.
a) f : R R given by f ( x) x 3 3 x
b) g : R 2 R 2 given by g ( x, y ) ( x 3 , y 3 )
c) f : R 2 R 2 , f ( x, y ) ( x, x y 3)
d ) h : R 2 R 2 given by h( x, y ) ( x , 3 y )
e) t : R 2 R 2 , t ( x, y ) ( x 1, y 1)
f ) g : R 2 R 2 , g ( x, y ) ( y x,6 x)
Solution:
a) S {x : f ( x) x} {x : x 3 3x x} {x : x 3 4 x 0}
{x : x( x 2 4) 0} {x : x( x 2)( x 2) 0}
{x : x 0, 2, 2} {0, 2, 2}
b) S {( x, y ) : g ( x, y ) ( x, y )} {( x, y ) : ( x 3 , y 3 ) ( x, y )}
{( x, y ) : x 3 x, y 3 y} {( x, y ) : x 3 x 0, y 3 y 0}
{( x, y ) : x 0,1, 1, y 0,1, 1}
{(0,0)(0,1), (0,1), (1,0), (1,1), (1,1), (1,0), (1,1), (1,1)
c) S {( x, y ) : f ( x, y ) ( x, y )}
{( x, y ) : ( x, x y 3) ( x, y )}
{( x, y ) : x x, x y 3 y}
{( x, y ) : x 3, y R} {(3, y ) : y R}
d ) S {( x, y ) : h( x, y ) ( x, y )} {( x, y ) : ( x , 3 y ) ( x, y )}
{( x, y ) : x x, 3 y y}
{( x, y ) : x 0, y y 3 0}
{( x, y ) : x 0, y 0,1, 1}
{( x,0), ( x,1), ( x,1), x 0}
Here, for any x 0 and y 0, 1, 1, ( x, y) is the fixed point of h( x, y) ( x , 3 y ) .
Prepared by Begashaw M. 19
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
e) Here, t ( x, y ) ( x 1, y 1) ( x, y)
x 1 x, y 1 y 1 0, 1 0
But this is impossible. This means t has no fixed point.
f ) S {( x, y ) : g ( x, y ) ( x, y )}
{( x, y ) : ( y x,6 x) ( x, y )}
{( x, y ) : y x x, 6 x y} {( x, y ) : y 2 x, 6 x y}
{( x, y ) : 2 x 6 x} {( x, y ) : x 2, y 4} {(2,4)}
Note: From the above examples, we can conclude that a given mapping can
have exactly one fixed point, finitely many fixed points or infinitely many
fixed points. On the other hand, part (f), shows that there are mappings that
have no fixed points.
Involution: A non-identity transformation is said to be an involution if and
only if 2 i . That means 2 ( x) ( )( x) ( ( x)) i( x) x for all
x in the domain of .
Examples: Verify whether the following transformations are involution or not.
a) : R R given by ( x) 1 x
b) h : R 2 R 2 , h( x, y ) ( x 7, y 2)
c) : R R given by ( x) x 3
d ) g : R 2 R 2 , g ( x, y ) ( x 3, y 5)
Solution:
a) 2 ( x) ( x) ( ( x)) (1 x) x i( x) 2 i .
So, is an involution.
b) h 2 ( x, y) h(h( x, y)) h( x 7, y 2) ( x, y) i( x, y) h 2 i .
So, h is an involution.
c) 2 ( x) ( ( x)) ( x 3) x 6 x i( x) 2 i is not an involution .
d ) g 2 ( x, y) g ( g ( x, y)) g ( x 3, y 5) ( x 6, y 10) ( x, y) i( x, y) g 2 i .
Hence, g is not an involution.
Prepared by Begashaw M. 20
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1.6 Collineations and Dilatations
Definition: A transformation f is said to be a collineation if and only if the
image of any line l under f is again a line. In other words, for any point P l
the image f ( P) f (l ) . Further more; f is said to be a dilatation if and only if
the image of any line l under f is a line parallel to l . That is f (l ) // l whenever
f is a collineation then f is said to be a dilatation.
Examples:
1. Let : R 2 R 2 be given by ( x, y) ( x 1, y 2) . Show that is a dilatation.
Solution: First let’s show that is a transformation. But, to show that is a
transformation, we need to show that it is one to one and onto.
One- to- one: Assume ( x, y) ( z, w) for any two points ( x, y) and ( z, w) in R 2 .
Then, ( x 1, y 2) ( z 1, w 2) . But from equality of order pairs, this equality is
true if and only if x 1 z 1 and y 2 w 2 . This gives x z and y w which
implies ( x, y) ( z, w) . So, is one-to-one.
On to ness: Let (a, b) R 2 be in the co-domain of . Then, if ( x, y) R2 in the
domain of , such that ( x, y) (a, b) , then is on to. But,
( x, y) ( x 1, y 2) (a, b) x 1 a, y 2 b x a 1, y b 2. Thus, we can find
( x, y) (a 1, b 2) R 2 such that ( x, y) (a, b) . So is onto. Therefore, the given
map is a transformation. To show that is a collineation we need to show
the image of an arbitrary line l : ax by c 0 is again a line.
Let ( x, y) be any point on l . Then, the image ( x' , y' ) ( x, y) ( x 1, y 2) .
Solving this for x and y we get x x'1, y y'2 . So,the image line will be
l ': a( x'1) b( y'2) c 0 ax'by'2b c a 0 and this is equation of a line.
Hence we can say that is a collineation. Besides, l ' has the same slope to
that of l which means l ' is parallel to l . In other words, l // (l ). Therefore, is
a dilatation.
Prepared by Begashaw M. 21
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2. Let : R 2 R 2 be given by ( x, y) ( y, x 1) . Then, find
a) The image of the line l : x 3 y 6 under .
b) The pre- image of the line l ': 2 x y 3 0 under .
Solution: Images and pre-images: Let be a transformation. Then, for any
two objects P and Q , if ( P) Q , the object Q is called the image of the
object P under and the object P is called the pre-image of Q under .
Besides, to find the image of any line l under any coollination :
First: Select any two points P and Q from the given line (the selection is
arbitrary you can select in any way you want !) and find their images P' ( P)
and Q' (Q).
Second: Form the equation of a line using the image points P' and Q' . This is
the required image line under .
a) Using the above steps, select P (0,2) and Q (6,0) on l : x 3 y 6 (you
can select any other points). Now find their images.
That is P' ( P) (0,2) (2,1), Q' (Q) (6,0) (0,7)
Hence, the image line is the line through P' (2,1) and Q' (0,7) .
y 6
That is the slope is m 3.
x 2
Then, using slope intercept form l ': y mx b 3x b y 3x 7 .
b) In this case, we are asked the pre-image, that means what is the line whose
image is given. To, do so, select two points P' and Q' on the image line
l ': 2 x y 3 0 . Say P' (0,3) and Q' (2,1) . Then, find two points P (a, b)
and Q (c, d ) such that ( P) P' and (Q) Q' .
Thus, ( P) P' (a, b) (b, a 1) (0,3) a 4, b 0 P (4,0) and
(Q) Q' (c, d ) (d , c 1) (2,1) d 2, c 0 Q (0,2)
Hence, the pre-image of the line l ': 2 x y 3 0 is a line through P (4,0)
y 2 1
and Q (0,2) . That is the slope is m . Then, using slope intercept
x 4 2
1 1
form l : y mx b x b y x 2 x 2y 4 0 .
2 2
Prepared by Begashaw M. 22
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Remarks:
1. The main difference between collineation and dilatation is that any
collineation maps a pair of parallel lines to a pair of parallel line but a dilatation
maps every line to a line parallel to the given line. This means a transformation
is a collineation if and only if for any two lines m and n,
m // n (m) // (n) . A transformation is a dilatation if and only if for any
line m , m // (m).
2. If ( x, y) ( x' , y' ) where x' ax by h, y' cx dy k , then the necessary
and sufficient conditions on the coefficients of x and y such that to be a
transformation is that ad bc 0 .(This is known as transformation test).
Examples:
1. Define : R 2 R 2 by ( x, y)) ( y, x) . Show that is a collineation but not
a dilatation.
Solution: Clearly, is a transformation. Besides, for any two arbitrary parallel
lines m : ax by c 0 and n : ax by k 0 (Parallel lines differ by a constant),
m' (m) : ay bx c 0, n' ( n) ay bx k 0 . Still, m' and n' have the
same slope which means they are parallel. i.e m // n (m) // (n). Thus, is a
collineation. But, if we consider only the single line m : ax by c 0
separately, m' (m) : ay bx c 0 . In this case, slope of line m is ba while
that of m' is b
a which gives the product of their slope is 1. This means
i.e m (m) m'. In other words, m and (m) m' are not parallel lines.
Particularly, take the line m : 6 x 2 y 5 0 . Then, its image under is
(m) m': 2 x 6 y 5 0 . Consequently, is a collineation but not a dilatation.
2. Find the value of the constant t for which ( x, y) (4x ty 7,8x 6 y 2) will
not be a transformation.
Solution: By the second part of the above ramark, will bot be a
transformation if and only if 4.(6) 8.(t ) 0 24 8t 0 8t 24 t 3 .
Prepared by Begashaw M. 23
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 1.1
1*. Let Z, N and W be the set of integers, natural numbers and whole numbers.
a) Define f : Z N by f ( x) x 2 . Is f a mapping? If so, is it one to one?
b) Define g : N Z by g ( x) x 2 . Is g a mapping? If so, is it one to one?
c) Consider the set T { x Z : x 1 2} . Define h : T Z by h( x) x 3 . Is h
a mapping? If so, is it one to one? is it onto?
d) Let S { x Z : x 2 3}, T { 0, 1, 4, 9, 16, 25} .
Define r : S T by r ( x) x 2 . Is this relation a mapping? If not, justify! If it is
a mapping, is it bijective?
Answer : a) It is not a mapping because for x 0 Z , f (0) 0 2 0 N .
This mean there is no image in the set of natural numbers N for x 0.
b) It is a mapping but neither one to one nor onto.
c) It is a mapping, it is one to one but not onto.
d ) It is a mapping, it is onto but not one to one.
2. Determine whether the following mappings are transformations or not on R 2
a ) f ( x, y ) (e x , y 3 ) b) g ( x, y ) (2 x, y 3) c) h( x, y ) (3 y, x 2)
d ) ( x, y ) (3 y 2,2 x 1) e) ( x, y ) ( x 3 x, y ) f ) ( x, y ) (3 x 1, y 2)
g ) ( x, y ) (2 y x, x 2) h ) ( x, y ) ( x 3 , y 3 ) i ) h( x, y ) ( y 1, x 3 )
j ) ( x, y ) ( x y 3,2 x 3 y 8) k ) ( x, y ) ( x y 1, x y 1)
Answer : Only b, c, d, f, g, h, j, k are Transformations
3. Find the fixed points (if any) of each of the following mappings
a) : R 2 R 2 , given by ( x, y ) ( y 3 , 5 x ) b ) f : R 2 R 2 , f ( x, y ) ( x , y x )
c) : R 2 R 2 , ( x, y ) ( x y 1, x y 1) d ) : R 2 R 2 , ( x, y ) ( x 3 x, y )
e) k : R 2 R 2 , k ( x , y ) ( x , y ) f ) f : R 2 R 2 , f ( x, y ) ( x y , x y )
g ) f : R 2 R 2 , f ( x, y ) ( x y 2, y 3) h ) k : R 2 R 2 , k ( x, y ) ( x, y x )
Answer : a) S {(0,0), (1,1), (1,1)} b) S {(0, y ) : y R} c) S {(1,1)}
d ) S {(0, y ), ( 2 , y ), ( 2 , y ) : y R} e) S {( x, y ) : x, y 0}
f ) S {(0,0)} g ) It has no fixed points h) S {(0, y) : y R}
Prepared by Begashaw M. 24
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
4. Let fn : Z Z be a map given by f n ( x) nx, x Z . For which values of n , f n
is one to one? Onto? Answer : One to one n 0 and onto n 1
5. Let f : R R, g : R R be defined by f ( x) kx 8, g ( x) 2 x 4 . Find the
value of the constant k such that g f f g Answer : k 3
6. Consider the transformations f ( x, y) (ax,3 y 2) and g ( x, y) (5x,4 y 2b) .
If ( g f )( x, y) (10 x,12 y 40) , then, find a and b . Answer : a 2, b 7
7. If S is a finite set, show that f : S S is one-to-one if and only if it is onto.
8. Find the formula for the inverse of the following functions.
1 x
a) f ( x) ln b) h( x, y ) (2 x,6 y ) c) ( x, y ) (3 x 1, y 2)
1 x
d ) ( x, y ) ( x y , x y ) e ) ( x, y ) ( x, y x )
e x 1
1
Answer : a) f ( x) x b) f 1 ( x, y ) (2 x,6 y )
e 1
x 10 x
9. Let : R 2 R 2 be given by . What value of t makes
y y 2t 6
an involution? Answer :t 3
10. *Consider the line L : y x 3 such that f : L R is given by
f ( x, y) 2 x, ( x, y) L . Show that f is bijective. If f ( P) 3 , find the
coordinates of P . Answer : P (3 / 2,9 / 2)
11. Let h be any transformation from R to R defined by h( x) 3x 6 . If
a is the inverse of h(x) , then what must be a ? Answer : a 2 / 3
g ( x) x2
2
12. Let and be any two transformations with ( P) 2 ( P) P, P R 2 . If
(7,3) (9,9) , then calculate (7,3) . Answer :(8,6)
13. If ( x, y) ( x' , y' ) where x' ax by h, y' cx dy k , then find the
necessary and sufficient conditions on the coefficients of x and y such that to
be a transformation. Answer : ad bc 0
14. Find the constant c for which ( x, y) (5x 3 y 2, cx 6 y 5) will not be a
transformation. Answer : c 10
Prepared by Begashaw M. 25
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
15. Let : R 2 R 2 be given by ( x, y) ( x y 7,2 x 3) . Show that is a
y 3 y 2 x 17
transformation and find 1 . Answer : 1 ( x, y) ( , )
2 2
16. Define g : R 2 R 2 by g ( x, y)) (ax c, by d ) where a and b are non-zero.
Show that a) g is a transformation b) g is a dilatation if and only if a b
17. Show that : R2 R2 given by ( x, y) (3 y, x y) is a collineation and find
the pre-image of the line l : y 3x 2 under . Answer : x 10 y 2 0
18. Define : R 2 R 2 by ( x, y) (5 y, x 1) . Show that is a
collineation but not a dilatation.
19. Show that : R 2 R 2 given by ( x, y) ( x 2, y 3) is a dilatation.
20. Given ( x, y) (3mx 2 x 1, my 6 y 1) . What should be the value of m so
that is a dilatation? Answer : m 4
21. Prove: If is an involution, then for any transformation , 1 is
also an involution.
22. If f an involution such that f (1,5) (2,1) , find f 1 (1,5) and f 1 (2,1) .
23*. Show that any collineation sends a pair of parallel lines to a pair of
parallel lines.
x
24. Find the condition on a and b such that ( x, y) (ay, ) is an involution.
b
Answer : a b
3
25. For what value (s) of k is ( x, y) (3k 2 1) y, x an involution?
22
Answer : k 5 / 3
1 2t
26*. Let : S R be a mapping given by (t ) where the set S is
2t 1 1
given by S t : 0 t 1.
Show that is a bijective mapping but not a transformation.
Prepared by Begashaw M. 26
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
27. Consider the napping f ( x) sin x . Define an equivalence relation
xRy sin x sin y . Find the equivalence class determined by y
2
Answer : R / 2 { / 2 n ; n Z}
28. Let : M 2 x 2 M 2 x 2 be a mapping given by ( A) A 1 , A M 2 x 2 where M 2 x 2 is
the set of all invertible 2x 2 matrices. Show that is an involution and
calculate all the fixed points of .
28. Show that the map : C C given by ( x yi ) x yi is a
transformation with infinite number of fixed points.
29. For each order pair (a, b) of integers define a,b : Z Z by a,b (n) an b .
For which pair (a, b) is a,b injective (one to one)? For which pair (a, b) is a,b
surjective (onto)?
Answer : a,b is injective for a 0, b Z , a,b is surjective only for a 1
30. Let M (a, b) and : R 2 R 2 be a mapping with the property that for any
point P in R 2 the mid point of P and (P) is M . Prove that is a
transformation with fixed point M itself.
31*. Let L be the set of all lines in the plane and let be a collineation in R 2 .
Then, is a transformation from L to L .
Prepared by Begashaw M. 27
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1.7 Definitions and examples of Transformation Groups
Definition: Let G be the set of all transformations on a non empty set S . Then
the system (G,) is said to be a transformation group if and only if the following
conditions are satisfied:
i) For all f , g in G , g f is in G .
ii) For all f in G , g G f g g f i and denoted by f 1 g.
iii) For all f in G , i G f i i f f .
Examples
1. Let f ab : R R be defined by f ab ( x) ax b, a 0.
Let G { f ab / a, b R, a 0}. Show that (G,) forms a transformation group.
Solution: To check whether the elements of G are transformation or not, for
any f ab in G. f ab ( x) f ab ( y) ax b ay b x y and for each r R , there
r b
exists x , a 0 in R such that
a
r b r b
f ab ( x) f ab ( ) a( ) b r b b r . Thus, any element f ab in G is a
a a
transformation.
i) Taking a 1, b 0, we get f10 ( x) x, x R . So, i f10 which is the identity
transformation contained in G .
ii) For any f ab and f cd in G ,
f ab f cd ( x) f ab ( f cd ( x)) f ab (cx d ) acx ad b f ac,ad b ( x)
Since f ab and f cd in G , a 0, c 0 ac 0 . Thus, f ab f cd f ac,ad b G . This
shows closure property holds true.
iii) To have inverse for any f ab in G , we have to find f cd in G such that
f ab f cd f cd f ab i f10 .
But f ab f cd ( x) f ac,ad b ( x) f10 ( x) x acx ad b x .
1 b
So, acx ad b x ac 1, ad b 0 c a 1 , d a 1b .
a a
Since a 0 , c and d are defined and hence ( f ab ) 1 f a 1
, a 1b
because
f ab f a 1 a 1b ( x) f ab (a 1 x a 1b) a(a 1 x a 1b) b x b b x f10 ( x) .
Hence, ( f ab ) 1 f a 1
, a 1b
is the required inverse of f ab in G . Therefore, by
definition (G,) forms a transformation group.
Prepared by Begashaw M. 28
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1 1
2. Consider the transformations f ( x) x, g ( x) x, h( x) , k ( x) defined
x x
on the domain R /{0} . Show that (G,) forms abelian group of transformation
where G { f , g , h, k} .
Solution: To show whether a finite set together with a binary operation forms a
group or not is simple by making a table called Cayley table. Each cell is filled
using the calculation as follow:
f f ( x) f ( f ( x)) f ( x) x f f f ,
f g ( x) f ( g ( x)) f ( x) x g ( x) f g g
1 1
g h( x) g (h( x)) g ( ) k ( x) and so on.
x x
From this table, we can infer the following results.
i) Existence of Identity: In the table, f g g f g and so on for all elements
of G . Thus f is the identity transformation in G .
ii) Existence of inverse: As we see from the table,
f f f , g g f , h h f , k k f where f is identity. So every element is
its own inverse.
iii) Closure property: If all the cells in the body of Cayley table is filled with
elements from the set, then closurity holds which is true in our case. Hence,
(G,) forms transformation group.
x' x
3. Let k be a mapping for k 0 given by k ( x, y) ( x' , y' ) where .
y ' ky
Show that the set G { k : k 0} forms abelian group of transformations with
composition. Particularly, give the identity element and the inverse of
3 ( x, y) ( x,3 y) in this group.
Proof: Apply the definition.
4. Show that the set of all transformations on the plane forms a transformation
group.
Proof: Apply the definition.
Prepared by Begashaw M. 29
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1.8 Criteria for Transformation Groups
So far we discussed about transformation groups and we saw how to show a
given set of transformations forms a transformation group using definitions and
or using Cayley table. But, some times definitions alone are not easy or
efficient to use in any cases. That is why mathematicians are looking for a short
cut method to use.
In transformation geometry, they developed the following theorem as a test for
a transformation groups.
Theorem 1.1 ( Test for a transformation groups):
Let G be a nonempty set of transformations on a set S . Then, G with
composition is a transformation group if and only if the following conditions
are satisfied:
a) f G f 1 G, f G
b) f , g G f g G, f , g G
Proof: Suppose (G,) is a transformation group. We need to show conditions
(a ) and (b) hold true. Since (G,) is a transformation group, from the definition
1
f , g G , f G and f g G whish implies that conditions (a) and (b) are
true. Conversely, suppose conditions (a) and (b) hold true. We need to show
(G,) is a transformation group.
i) Existence of Inverse: Since G , f G but f G, f 1 G from
condition (a) . So, G contains inverse transformation.
ii) Closure property: From condition (b) , f , g G f g G, f , g G.
iii) Existence of Identity: f G , f 1 G from condition (a) and from
condition (b) , f 1 f G and f f 1 G .
But, f 1 f i G and f f 1 i G .
Thus, G contains identity transformation. Therefore, by definition, (G,) forms
a transformation group.
Examples:
1. Let g a : R R be defined by g a ( x) ax, a 0. Let G {g a / a R, a 0} .
Show that (G,) forms a transformation group.
Solution: So far, we have seen that g a is a transformation. Using the above test
for transformation group, we will verify the problem as follow:
Prepared by Begashaw M. 30
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1
i) For any g a G, g a ( x) ax g a1 ( x) x a 1 x g a1 G.
a
ii) For any g a , g b G, g a g b ( x) g a ( g b ( x)) abx g ab ( x) g a g b g ab
Since g ab G ,and g a g b g ab , we get g a g b G . Hence, by test of
transformation group, (G,) forms a transformation group.
2. Let f a : R 2 R 2 ; f a ( x, y) (2a x, y), a R such that G { f a :a R} .
Using criteria of transformation group, determine whether (G,) is
transformation group or not.
Solution: For any f a G, f a ( x, y) (2a x, y) . First of all, we need to show
that every element of G is transformation.
Now suppose, f a ( x, y) f a ( z, w) .
f a ( x, y ) f a ( z, w) (2a x, y ) (2a z, w)
2a x 2a z , y w
x z, y w ( x, y ) ( z, w)
This shows that each f a is one to one.
Besides, to each ( z, w) R 2 (in the co-domain), ( x, y) (2a z, w) R 2 such
that f a ( x, y) f a (2a z, w) ( z, w).
So, each f a is also onto.
Therefore, from these explanations every element of G is a transformation.
For any f a G,
f a ( x, y) (2a x, y) f a1 ( x, y) (2a x, y) f a1 f a f a1 G .
But for any two elements f a , f b G,
f a f b ( x, y) f a (2b x, y) ( x 2a 2b, y) f r ( x, y), r R f a f b G.
This means the second condition of the above theorem (test for a
transformation group) fails.
As a result, (G,) does not form transformation group.
Prepared by Begashaw M. 31
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Theorem 1.2 (Cancellation Laws on Transformation Groups):
Let G be a transformation group. Then, for , , in G
a) (This is called Left Cancellation Law)
b) (This is called Right Cancellation Law)
Proof: Let G be a transformation group. Then, for , , in G
a) 1 ( ) 1 ( )
b) ( ) 1 ( ) 1
Theorem 1.3: In any transformation group G , for any , in G , the equation
has a unique solution for in G which is given by 1 .
Proof: For , in G , 1 ( ) 1 1 G
Hence, the equation has a solution in G . For the uniqueness of this solution,
assume there are two different solutions say , , then
, (by Left cancellation)
Example: Let (G,) be a transformation group such that , , in G . If
( x, y) (3 x,5 y) , ( x, y) (3 2 x,5 3 y) . Find such that
Solution: By the above theorem, has a unique solution for in G
which is given by 1 .
But after some ups and downs we get 1 ( x, y) (3 x,5 y) .
Therefore, ( x, y) 1 ( x, y) 1 ( ( x, y ))
1 (3 2 x,5 3 y) (2 x,3 y)
Prepared by Begashaw M. 32
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 1.2
1. Consider the following transformations on R /{1,0) .
1 1 x 1 x
f ( x) x, g ( x) , h( x) 1 x, k ( x) , m( x) , n( x ) .
x 1 x x x 1
If G { f , g , h, k , m, n}, show that (G,) forms a transformation group.
2. Consider the set G { f , g , h, k} where f , g , h, k are transformations from
R 2 to R 2 given by:
f ( x, y ) ( x, y), g ( x, y) (2 y, x)
h( x, y ) ( y,2 x), k ( x) (2 x,2 y)
Show that (G,) forms an abelian transformation group.
3. Prove that if a is a fixed element of a group G , and : G G is defined by
( x) ax, x G , then is a transformation.
4. Let f : R R be defined by f ab ( x) ax b, a 0. Let
G { f ab / a, b R, a 0} , H { f1b : b R} . Show that
a) ( H ,) is a subgroup of the transformation group (G,)
b b
b) R {(a, b) : f ( x)dx g ( x)dx, f , g G} is an equivalence relation
a a
x' kx
5. Let k be a mapping for k 0 given by k ( x, y) ( x' , y' ) where .
y' y
Show that G { k : k 0} forms abelian group of transformations with
composition. Give the identity element and the inverse of 5 ( x, y) (5x, y) in
this group. List the involution elements (if any).
1
Answer : Identity element, i 1 , Inverse of 5 is 1 ( x, y ) ( x, y ),
5
5
The set of all involution elements : 1 , 1
Prepared by Begashaw M. 33
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
6. *Let a , b be maps for a, b 0 given by a ( x, y) (ax, y), b ( x, y) ( x, by) .
Consider the set G { a b : a, b 0} . Show that G, forms abelian group of
transformations. Give the identity element and the inverse of 1 2 in this
3
group. List all the involution elements .
1
Answer : Identity element, i 1 1 , Inverse of 1 2 is 3 1 ( x, y ) (3x, y ),
3 2
2
The set of all involution elements : 1 1 , 1 1 , 1 1 , 1 1
7. Let G be a nonempty set of transformations on a set S . Then, (G,) forms a
transformation group if and only if f , g G f g 1 G, f , g G.
8. Let G be any group. Define f : G G by f (a) a 1 , a G . Show that f is
a transformation.
9. Suppose , , are transformation such that ( X ) ( X ) where
( x, y) ( y, x) and ( x, y) (1 x, y 10) . Find equation of .
Answer : ( x, y) ( y 10, x 1)
10. In a transformation group G , if f f 1 , f G , then prove that G is
abelian.
11. Let G be a finite transformation group. Show that in a Cayley table of G ,
each element appears exactly once in each raw and in each column.
12. Let G be a transformation group. If f g g , then show that f i .
13. If G is the set of transformations given by G { f , g , h} , then complete the
following Cayley table of G .Which element of G is the identity? Give the
inverse of each element. Which element (s) of G are involutions?
Prepared by Begashaw M. 34
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
14. In a transformation group G , if ( ) 2 2 2 for each , in G , prove
that G is abelian group of transformation.
15. Let (G,) be a transformation group with the property that
f G f f f 2 i. Prove that; every element of G is its own inverse.
16. In a transformation group G , if all the non-identity elements are involution,
then show that G is abelian.
17. Let , , be elements of a transformation group G such that
( x, y) ( x, y), ( x, y) R 2 where ( x, y) (3 y,2 x 1) and
( x, y) (9 y,4 x 1) . Find the equation of .
18. Let (G,) be arbitrary group. Prove that there is a group T of transformation
on G and a one to one mapping g : G T which assigns to each a G , a
transformation f a T such that f ba f b f a , a, b G .
Prepared by Begashaw M. 35
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
CHAPTER-2
AFFINE GEOMETRY
2.1 Introduction to Affine Spaces
Affine geometry is a type of geometry that can be used to make analysis in a
plane and space whose major entries or elements are vectors. The properties
and conditions of affine geometry work in any dimensional space (finite or
infinite). As a result, affine geometry is treated as dimension free. The space
that we will work using this geometry is called affine space (which is formally
defined below). In this space, there is no specified origin.
As vectors are the main elements in this geometry, most of the theorems and
problems that we know in Linear Algebra courses are treated in this unit only
by using vectors and vector analysis rather than using coordinates and
Synthetic analysis. Now, we will prove and solve different theorems and
problems using vector method in order to understand the notion of affine
geometry and affine spaces. Finally, you may appreciate to what extent vectors
are useful in analysis of theorems and problems that we already know in
different branches of Mathematics.
Definition: A non-empty set W is said to be an affine space associated with the
vector space V if and only if the following three conditions are satisfied:
i. To every two points A, B in W , there exists a unique vector u AB B A in
V , and u is the zero vector if and only if A B .
ii. To any point A in W and any vector u in V there exists a unique point B
such that u AB .
iii. For any three points A, B, C in W , we have AB BC AC .
Prepared by Begashaw M. 36
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
In any affine space W , its elements are called vectors. Most of the concepts
like norm, projection, cross product and other concepts of vectors in Euclidean
Geometry are similarly defined in affine geometry.
Examples:
1. Let V R 3 be the vector space with the usual operations. Determine whether
each of the following sets are affine spaces associated with V or not
a) W {( x, y, z ) : x y z}
b) W {( x, y, z ) : x y z 0}
c) W {( x,0,0) : x R}
d ) W {( x, y, z ) : x 2 y z 7}
Solution: Follows directly from the definition.
2. Let V M 22 be the vector space of 2 2 matrices with the usual operations.
Verify that W {A M 22 : A At } forms an affine spaces associated with V .
Perpendicular and Parallel vectors:
Definition: Two non-zero vectors a and b are said to be orthogonal or
perpendicular if their dot product is zero.
Denoted by a b . Thus, a b a.b 0 .
Note that if a is perpendicular to b and k is any scalar, then a is also
perpendicular to k b because a k b k (a.b) k.0 0 a k b .
Example: Show that the vectors a (3,1,5) and b (1,3,2) are orthogonal.
Solution: Here, a.b 3 13 10 0 a b .
Definition: Any two non-zero vectors are said to be parallel if and only if one
is the scalar multiple of the other.
Let a and b be any two non-zero vectors. Then, a // b t 0 R, a t b and
note that if a // b, a t b we always have b // a, b 1t a .
a
The scalar t is said to be norm (length) ratio and given by t .
b
Prepared by Begashaw M. 37
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Examples
1. Verify that the vectors a (2,1,3) and b (6,3,9) are parallel.
Solution: Here, b (6,3,9) 3(2,1,3) 3a t 3, b 3a .
Therefore, the two vectors are parallel. That is a // b .
2. Let a (m,3,4) and b (2,n,8) . Find the values of m and n so that the
vectors a and b are parallel.
Solution: Here,
a // b t 0 R, b t a (2,n,8) t (m,3,4)
2 m.t , n 3t , 8 4t t 2 m 1, n 6
Proposition 2.1:
i) Parallelism relation denoted by // on vectors is an equivalence relation.
ii) If a and b are not parallel, then r , t R , the equation r a t b has a unique
solution r t 0
Proof:
i) For any vector a 1.a a // a and thus // is reflexive. Suppose a // b. Then,
t R, a t b b 1t a b // a .Thus, // is symmetric. If a // b and
b // c , then a r b and b t c . Therefore a rtc a // c so // is transitive.
From the three conditions, // is an equivalent relation on vectors.
ii) Suppose r 0 such that r a t b , then a rt b a // b which contradicts with
the assumption that a and b are not parallel. Therefore r 0 and similarly by
assuming t 0 it can be shown t 0 .
Prepared by Begashaw M. 38
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2.2 Geometry in Affine Space
Since the elements of affine space are vectors, the geometric theorems that we
know are analyzed using vector analysis. Now, let’s analyze some theorems
and problems using vectors in order to understand affine geometry and affine
space.
Proposition 2.2:
1. The line segment joining the mid-points of two sides of a triangle is parallel
to the third side and its length is half of that side.
Proof: Consider ABC below (figure 2.1a) where M and N are the mid-points
of AC and BC .
From the third condition in the definition of affine space, we have
AC AB BC 2MC AB 2 NC
2MC 2 NC AB
2(C M ) 2(C N ) AB
2MN AB MN 12 AB
Besides, since MN 12 AB , MN is a scalar multiple of AB with scalar t 12 .
Hence, MN // AB .
2. The diagonals of a rhombus are perpendicular.
Proof: Let ABCD be a rhombus with diagonals AC and BD as shown in figure
2.1b above. To show that the diagonals AC and BD are perpendicular, it
suffices to show that their dot product is zero. Let AB a, BC b such that
AC a b, DB a b .
Prepared by Begashaw M. 39
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Besides, since the four sides of a rhombus are congruent AB a b .
2 2
Thus, AC.DB (a b) (a b) a a a b b a b b a b 0 .
Hence the diagonals are perpendicular to each other.
3. The cosine laws: In ABC , suppose the lengths of the sides of the triangle
opposite to the angles A, B, C respectively are a, b, c . Then,
a 2 b 2 c 2 2bc cos A
b 2 a 2 c 2 2ac cos B
c 2 a 2 b 2 2ab cos C
Proof: Consider figure 2.2b. In this diagram, the arrow indicates in which
direction we considered the vectors along the sides so that to determine what
angle between the vectors to be taken. Because angle between two vectors is
the angle formed by the vectors when they share (made to share) the same
initial points or tails. For instance, the angle between the vectors AC and CB is
given by C
Here, we have
AC CB AB ( AC CB ).( AC CB) AB. AB
2 2 2
AC CB 2 AC .CB AB
b 2 a 2 2 AC CB cos c 2
b 2 a 2 2ab cos( C ) c 2
c 2 a 2 b 2 2ab cos C , cos( C ) cos C
Similarly, the other identities can also be derived.
Prepared by Begashaw M. 40
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
4. Sum and Difference Formula of Trigonometric: Let and be any two
angles. Then,
i ) sin( ) sin cos cos sin
ii ) cos( ) cos cos sin sin
iii ) sin( ) sin cos cos sin
iv ) cos( ) sin cos cos sin
Proof: Let OA and OB be two unit vectors along the lines OL and OM making
angles of and with the x-axis respectively as shown in the diagram below.
Here;
a) OA cos .i sin .j, OB cos .i sin .j
b) MOL MOC COL
Now, to prove (i), let’s proceed as follow.
OA OB (cos .i sin .j) (cos .i sin .j)
cos cos (i i) cos sin (i j) sin cos ( j i) sin sin ( j j)
cos sin (i j) sin cos ( j i), [ i i j j 0]
cos sin ( j i) sin cos ( j i), [ i j j i]
[cos sin sin cos ]( j i)
[cos sin sin cos ]k , [ j i k ]..............................................(i )
On the other hand, using OA OB 1, MOL , we have
OA OB OA . OB sin MOL.k sin( )k........................................................(ii )
Thus, from (i) and (ii), we have
OA OB OA . OB sin MOL.k sin( )k [cos sin sin cos ]k
sin( ) cos sin sin cos sin cos cos sin
Prepared by Begashaw M. 41
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
To prove part (ii), let’s appy dot product.
That is OA.OB (cos .i sin .j).(cos .i sin .j) cos cos sin sin
On the other hand, OA.OB OA . OB cos MOL cos( )
Combining the two equations, gives us cos( ) cos cos sin sin
5. The sine Law: Suppose the sides of ABC are represented by the vectors
a b c
a, b and c where a a, b b, c c . Then, .
sin A sin B sin C
Proof: Consider ABC with A , B , C as shown in the diagram.
Now, observe that:
i) The angle between the vectors a and b is . (Do you see why?)
ii) The angle between the vectors a and c is . (Do you see why ?)
iii) The angle between the vectors b and c is . (Do you see why?)
iv) For any angle , sin( ) sin (Difference formula of angles)
Besides, by parallelogram law a b c 0 (zero vector).
But a b c 0 a b c . Here, if we cross by a both sides, we get
a ( a b ) a c a a a b a c
a b c a, [because a a 0, a c c a]
Similarly, if we cross by b both sides, we get
b (c) b (a b) b c b a b b b c a b b c a b
Prepared by Begashaw M. 42
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
From the two results, we get that b c c a a b .
But if two vectors are equal, then their norms are equal and if the angle
between any two non-zero vectors a and b is , then
a b a b sin , 0 .
Hence,
bc ca ab bc ca ab
b . c sin( ) c . a sin( ) a . b sin( )
b.c. sin A c.a sin B a.b. sin C
1
(b.c. sin A c.a sin B a.b. sin C )
a.b.c
sin A sin B sin C a b c
a b c sin A sin B sin C
a b c
In any ABC , the result is known as The Sine Law.
sin A sin B sin C
6. Using vectors, prove Pythagoras Theorem.
Proof: Consider ABC with right angle at vertex A . We need to show
c2 a2 b2 .
From figure 2.2a, AB a, AC b, BC c, AC. AB 0 and
AB BC AC BC AC AB . Thus,
2
c 2 BC BC .BC
( AC AB ).( AC AB )
2 2
AC 2 AB. AC AB
2 2
AC AB , AC . AB 0
a b
2 2
Prepared by Begashaw M. 43
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2.3 Lines and Planes in Affine space
From Euclidean geometry, we know that any two distinct points determine a
unique line whose equation can be determined using the concepts in coordinate
geometry. But, in affine space lines and planes are defined using the concept of
vectors. So our next discussion focuses on analysis of lines and planes in affine
space using vectors.
2.3.1 Lines in Affine Geometry
Definition: Let W be an affine space. Then, a line in W through two different
points A and B is defined as the set A, B {X W : AX t AB, t R}
or l : X A t AB, t R . Here, the vector AB is called direction vector of the line
and the scalar t is called parameter. Now, let l be any line through the points A
and B in affine space. Then, l : A, B {X W : AX t AB, t R} .
Letting X ( x, y, z ), A ( x0 , y0 , z 0 ), d AB (a, b, c) , we get that
l : ( x, y, z) ( x0 , y0 , z 0 ) t (a, b, c) (This is called vector equation of the line).
Parametric and Symmetric Equations of a line:
x x0 at
Equating corresponding components from this equation, we have y y 0 bt
z z ct
0
(This is called Parametric Equation).
Again, from the parametric equation (whenever a, b, c are all non zero), solving
x x0
x x0 ta t a
for the parameter t gives us, y y tb t y y 0
0
b
z z0
z z 0 tc t c
x x0 y y 0 z z 0
Equating these values of t implies (This is known as the
a b c
Symmetric Equation).
Prepared by Begashaw M. 44
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Remarks: The case when one of a, b, c is zero gives the following equations.
y y0 z z0
i) If a 0, x x0 ,
b c
x x0 z z 0
ii) If b 0, y y 0 ,
a c
x x0 y y 0
iii) If c 0, z z 0 ,
a b
Examples: Give the vector, parametric and symmetric equations of the line
that passes:
a) Through the points A(2,1,1) and B(1,3,2)
b) Through the point A(1,3,0) and parallel to the vector u 2i 5 j 7k
c) Through the point A(2,3,5) and parallel to the vector u 4 j 8k
Solution:
a ) VE : l : X A t AB (2,1,1) t (1,4,3),
x 2 t
x 2 y 1 z 1
Parametric Equations : y 1 4t , and Symmetric Equation :
z 1 3t 1 4 3
b) Vector equation (VE) : l : X (1,3,0) t (2,5,7)
x 1 2t
x 1 y 3 z
Parametric Equations (PE) : y 3 5t and Symmetric equation (SE) :
z 7t 2 5 7
x 2
y 3 z 5
c) VE : l : X (2,3,5) t (0,4,8), PE : y 3 4t , SE : x 2;
z 5 8t 4 8
Note that the vector and or the parametric equations of a line are not unique.
That means we can find many parametric equations for the same line.
Proposition 2.3:
a) There is a unique line through any two points in affine space.
b) Any two different direction vectors of a line are parallel.
Proof: (Follows from the definition)
Prepared by Begashaw M. 45
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Parallel and Perpendicular Lines in Affine Space:
Definition: Any two lines A, B and C, D are said to be parallel if their
direction vectors AB and CD are parallel and they are said to be
perpendicular if their direction vectors AB and CD are perpendicular.
Besides, the angle between these lines is the same as the angle between their
direction vectors.
Examples:
1. Determine whether the following pairs of lines are parallel, perpendicular or
neither and for these which are neither find the angle between them.
a) l : X (1,2,3) t (1,2,1) and m : X (1,0,1) t (3,6,3)
b) The lines through A (1,3,5),B (4,7,5) and C (5,2,2),D (1,1,7)
c) The lines through A (2,1,4),B (2,2,5) and C (3,4,3),D (3,5,3)
Solution:
a) Here, u (1,2,1), v (3,6,3) v 3u u //v .
Hence, from the definition, the lines are also parallel.
b) In this case, u B A (3,4,0), v D C (4,3,5) v.u 0 u v .
Hence, the lines are also perpendicular.
c) u B A (0,1,1), v D C (0,1,0) . But those vectors are neither parallel
nor perpendicular and so are the lines through these points. Let be the angle
u.v 1 1 3
between the lines. Then, cos cos 1 ( ) .
u v 2 2 4
3
Or it could be .
4 4
2. For what value of k are the line l : x 2t , y 1 3t , z 2 7t and the line
m : X (2,3,5) r (3, k , 3) perpendicular in Affine space?
Solution: The direction vectors of the lines are u 2i 3 j 7k , v 3i kj 3k .
Then, the two lines will be perpendicular if their direction vectors are
perpendicular. So, u v u.v 0 6 3k 21 0 k 9 .
Prepared by Begashaw M. 46
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3. Find the value of a if the cosine value of the angle between the lines
1
l : x y z and m : x 1 t , y at 5, z 4 2t is .
3
Solution: The angle between the two lines is the same as the angle between
their direction vectors u i j k , v i aj 2k
1 u.v 1 a3 1
So, cos a3 a2 5
3 u v 3 3 a 52
3
(a 3) 2 a 2 5 a 2 6a 9 a 2 5 a 2 / 3
4. Let l : A, B and m : C, D be two non-perpendicular intersecting lines in
AB xCD
affine space. Then, verify that tan where is the angle between
AB CD
the lines.
Proof: From the above definition, the angle between two lines in Affine Space
is the same as the angle between their direction vectors. Thus, as is the angle
between the lines, it is also is the angle between their direction vectors AB
and CD . Besides, from vector analysis we know that AB CD AB CD sin
and AB CD AB CD cos . Thus dividing the first relation by the second
AB CD AB CD sin sin
yields, tan .
AB CD AB CD cos cos
Proposition 2.4: Two Lines l : X A ru and m : X A tv passing through
the same point P are equal (identical) if and only if u and v are parallel.
Proof: Suppose l m and let B l m, B A . Then,
B A ru, B A tv A ru A tv
ru tv u rt v, [ B A r 0 ]
u // v
Conversely, let u and v be parallel. Then, there exists a scalar k 0 such that
u kv . Let X be arbitrary point on line l .
Thus, X l X A ru X A r (kv)
X A tv, t rk X m l m
Prepared by Begashaw M. 47
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Proposition 2.5: Given a line l and a point Q not on l . Then, there is exactly
one line m through Q parallel to l .
Proof: Let l : X P tu , then m : X Q tu for any point Q is parallel to l
(because by definition any two lines with the same direction vectors are
parallel). This proves the existence of parallel line to l .
Now suppose there is another line n through Q parallel to l . In this case, it is
given by n : X Q rv, m // n u // v .
Therefore, from m : X Q tu and n : X Q rv with u parallel to v we get
m n by the previous proposition.
Theorem 2.1(Length Ratio Theorem): Let l be a line passing through two
different points A and B given by l : P A r AB, r R . Then,
AP r AP
a) r b) , if P B
AB 1 r PB
c) P is between A and B if 0 r 1
d ) B is between A and P if r 1
Proof: a) From vector equation of the line l : P A r AB, r R , we have
AP AP
P A r AB, r R AP r AB AP r AB r AB r
AB AB
b) For any point P on a line determined by two distinct points A and B ,
AB AP PB . Besides, from equation of the given line we have AP r AB
which gives us AB 1r AP . From these relations, we get
1
AP AP PB (1 r ) AP r PB
r
(1 r ) AP r PB
AP r
1 r AP r PB
PB 1 r
c) Here, we use the fact that for any vector a and a positive constatnt k , a and
k a have the same direction and k a k a .
For 0 r 1 , AP r AB AP r AB AB AP rAB AB .
This shows P is between A and B . The rest parts of the theorem can be derived
from what we did here.
Prepared by Begashaw M. 48
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Examples:
1. If A (5,0,7) and B (2,3,6) , then find a point P on the line A, B which
AP
satisfies 3.
PB
Solution: From part (b) of Length Ratio Theorem, on the line l : P A r AB
r AP
if P B , we have .
1 r PB
AP
But we need to find a point P which satisfies 3.
PB
r AP r 3 3
Thus, 3 3 r or r . Besides, any point P on the
1 r PB 1 r 4 2
line determined by A and B is given by l : P A r AB, r R.
3 3 3 11 9 25
Using r , we get P A AB (5,0,7) (3,3,1) ( , , ).
4 4 4 4 4 4
3 3 3 1 9 11
Using r , we get P A AB (5,0,7) (3,3,1) ( , , ) .
2 2 2 2 2 2
2. Application of Line Equation in Affine Space: Suppose an Engineer
wants to paint a rectangular portion ABCD of a building using white and
yellow colors where two of the vertices are A (2,3,1) and B (3,7,4) . If he
divides the region into two sub-rectangles by a line through P (on the segment
AB ) and Q (on the segment CD ) so that to paint white color the region APQD
and yellow color the region PBCQ where the ratio of the length painted white
AP 3
to the length painted yellow is known to be . Find the coordinates of P
PB 2
where he should draw the dividing line to Q .
r AP
Solution: From Length Ratio Theorem, we have . But we need to
1 r PB
AP 3
find a point P which satisfies .
PB 2
Prepared by Begashaw M. 49
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
r AP 3 r 3
Thus, 2r 3(1 r )
1 r PB 2 1 r 2
2r 3 3r or 2r 3r 3 r 3 / 5 or r 3
Since we need a point P between A and B , only r 3 / 5 is valid by part (c)
of
Length Ratio Theorem. Besides, any point P on the line determined by A and
B is given by l : P A r AB, r R.
3 3 3 13 27
Hence, using r , we get P A AB (2,3,1) (1,4,5) ( , ,2).
5 5 5 5 5
Here, if we use r 3 , we get P A 3 AB (2,3,1) 3(1,4,5) (5,15,14). But we
didn’t do so. (Verify why this value of P is not used!).
Theorem 2.2 (Distance in affine space)
Let l be a line through A and B . If C is any point not on the given line l , then
there is exactly one point P on the line l such that CP is perpendicular to AB
AC . AB
given by P A t AB, where t .
AB
Moreover, if Q is any other point on the line, then CQ CP and hence P is the
point on l closest to C . Besides, the shortest distance from the line to the point
AC 2 AB 2 ( AC . AB ) 2 AB AC
C is given by d CP
AB AB
Proof: Let P A t AB and assume CP is perpendicular to AB as shown in
figure 2.3 below.
Prepared by Begashaw M. 50
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Then, CP. AB 0 ( P C ). AB 0
( A t AB C ). AB 0
(CA t AB ). AB 0
CA. AB t ( AB. AB ) 0
AC . AB t ( AB. AB ) 0
AC . AB
t ( AB 2 AB )
2
,
AB 2
The uniqueness of point P follows from the unique value of the parameter t .
Now for any other point Q , the inequality CQ CP follows from Pythagoras
theorem. Because if Q is any other point on the line (on the either side of P ),
then CQ 2 CP 2 PQ 2 CQ 2 CP 2 CQ CP. Equality holds when P Q .
Finally, from the diagram as the vector CP is perpendicular to PA , using
Pythagoras theorem we get,
Prepared by Begashaw M. 51
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2
CP 2 AC 2 AP 2 AC 2 t AB , P A t AB AP t AB
2
AC . AB
AC 2 AB 2
AB 2
AC 2 . AB 2 ( AC . AB ) 2
AB 2
AB ( AC AB cos ) 2
2 2
AC
AB 2
AB [1 cos 2 ]
2 2
AC
AB 2
AC 2 AB sin 2
2
, 1 cos 2 sin 2
AB 2
AC AB sin
2
AB 2
2
AC AB
AB 2
Taking square root both sides gives the required result.
Examples:
1. Let l be a line through the points A (2,1,3) and B (1,2,4) . Find a point P
on this line which is closest to the origin and calculate the shortest distance
from the line to the origin.
Solution: From the above theorem, the closest point on the line to a given point
AC . AB
C (not on the line) is given by P A t AB , where t .
AB 2
In our case, A (2,1,3) , B (1,2,4) and C (0,0,0) . Thus,
2,1,3 . 3,1,1 2
AC 2,1,3 , AB 3,1,1 , AB 11 t
11 11
2 16 13 35
P (2,1,3) 3,1,1 ( , , )
11 11 11 11
AC 2 AB 2 ( AC . AB ) 2 154 4 150
d CP d
AB 11 11
Prepared by Begashaw M. 52
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2. Find a point on the line l : X (1,2,3) r (2,3,4) closest to Q (5,9,3) .
Solution: From the above theorem, the closest point on the line to a given point
AQ. AB
Q is given by P A t AB , where t and AB is any direction vector of
AB 2
the line. In our case, A (1,2,3) and AB (2,3,4) .
4,7,0 . 2,3,4 29
Thus, t 1.
29 29
Therefore, P A t AB P (1,2,3) (2,3,4) (3,5,7).
2.3.2 Planes in Affine Space
Definition: Let W be an affine space.
Then, a plane passing through three non-collinear points A, B, C is the set
given by A, B, C {X : X A r AB t AC , r, t R}
From this equation, if we let A ( x 0 , y0 , z 0 ), AB (a, b, c), AC (d , e, f ) , then
any arbitrary point X ( x, y, z) on this plane is given by
X ( x, y, z) ( x 0 , y0 , z 0 ) r (a, b, c) t (d , e, f ).
This is known as vector equation of the plane. Now by equating components
x x0 ra td
from the vector equation, we get y y 0 rb te . This is called parametric
z z rc tf
0
equation of the plane where r, t are parameters.
Example: Find vector and parametric equations of a plane passing through the
points A (1,2,3), B (4,1,2), C (1,0,1) .
Solution: First let’s determine the vectors using the given points.
AB (3,3,5), AC (0,2,4) . Thus, the vector equation becomes,
: A, B, C {X : X ( x, y, z) (1,2,3) r (3,3,5) t (0,2,4)}
x 1 3r
The parameter equation with parameters r, t becomes y 2 3r 2t
z 3 5r 4t
Prepared by Begashaw M. 53
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 2.1
1. Let A (3,10,9), B (7,5,1), D (8,16,8) . Show that AB and AD are
perpendicular. Find a fourth point D such that ABCD forms a rectangle.
2. Show that R is an affine space over R .
3. Using vector method, prove that:
a) An angle inscribed in a semicircle is right angle.
b) The altitudes of a triangle are concurrent.
c) The sum of the squares of the diagonals of a parallelogram is
equal to twice the sum of the squares of its sides.
4. Using vector method, show that:
a) cos( ) cos cos sin sin
b) sin( ) sin cos cos sin
AP 1
5. If A (0,4), B (8,2) , find a point P on the line AB such that .
PB 3
Answer : P (2,5 / 2)
6. Show that a point P on the line A, B is uniquely determined by the
AP
ratio r . If A (2,3,1), B (3,7,4) , then find the coordinates of
PB
AP 2
point P on AB such that .
PB 5
7. Let A, B, C be points in affine space. Show that the set A, B, C is either
a plane, a line or a single point.
Particularly, if A B C , what is the geometric figure represented by
the set A, B, C ?
8. If P is any point inside ABC , prove that P rA sB tC where
r s t 1.
Prepared by Begashaw M. 54
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
9. If P is any point inside ABC such that P 2rA 34 B 12 C , then what
must be the value of r ?
10. Give the centroid of DEF whose vertices are D (3,5,6), E (2,7,9)
and F (2,1,7) .
11. Let l be a line through A (7,1,9), B (1,5,8) . Show that the length of the
17
projection of the segment CD is L where C (1,2,3), D (5,2,6) .
3
12. Find a point on the line l : X (1,2,3) r (2,3,4) which is closest to
Q (5,9,3) . Answer : (3,5,7)
13. Let l : A, B and m : C, D be any two intersecting lines in affine
AB CD
space. Then, verify that cot where is the angle between
AB xCD
the lines.
Prepared by Begashaw M. 55
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2.3.3 Collinearity in Affine Space
Definition: (Co-linearity in affine space): Any three points A, B, C in affine
space are said to be collinear if and only if A aB bC where a b 1.
Proposition 2.6: Suppose A, B, C, D are distinct points such that any three of
them are not collinear. Then, AB CD if and only if AB // CD and AC // BD .
Proof: If AB CD , then B A D C C A D B AC BD . But any
two equal vectors are parallel and thus AB // CD and AC // BD .
Conversely, suppose that AB // CD and AC // BD .Then, for some scalars r and t ,
AB r CD and AC t BD . This in turn implies, B A r ( D C ) and
C A t ( D B) . Here, if we can show that either r 1 or t 1 we are done.
Now, subtracting the second relation from the first gives
B C r ( D C) t ( D B) (1 t ) B (1 r )C (r t ) D
1 r r t
If we assume t 1 , we have (1 t ) B (1 r )C (r t ) D B C D
1 t 1 t
1 r r t
This shows that B aC bD with a b 1 where a , b .
1 t 1 t
But this mean that B is on the line CD which implies that B, C, D are collinear
but this is a contradiction with the hypothesis that any three of the points are
not collinear. Hence, t 1 . Similarly one can get r 1 .
Therefore, AC t BD C A D B B A D C AB CD .
Prepared by Begashaw M. 56
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Theorem 2.3 (The intercept Theorem):
Let A, B, C, D be non-collinear points such that AB // CD, AB CD . Then, the
OA OB AB
lines A, C and B, D intersect in a points O for which
OC OD CD
AC BD
and . Conversely, if the lines A, C and B, D intersect in a point
OA OB
OC OD
O such that , then AB // CD, AB CD .
OA OB
Proof: Consider figure 2.4.
Since AB // CD, AB CD the lines A, C and B, D intersect in a point.
Otherwise, if they don’t intersect, ABDC becomes a parallelogram such that
AB CD which is a contradiction. Now suppose they intersect at some point O
(refer the figure above)and then consider OAB and OCD .In these triangles,
OAB OCD, OBA ODC (corresponding angles) as AB // CD.
Thus, OAB ~ OCD by Angle-Angle similarity theorem.
OA OB AB
As a result, .
OC OD CD
Besides, from this result,
Prepared by Begashaw M. 57
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
OA OB OC OD
OC OD OA OB
OC OD
1 1
OA OB
OC OA OD OB
OA OB
AC BD
OA OB
Conversely, suppose the lines A, C and B, D intersect at some point O
OC OD OC OD
such that . Let k . Then, OC k OA, OD k OB .
OA OB OA OB
Thus, OC CD OD CD OD OC k OB k OA k (OB OA) .
On the other hand, OA AB OB AB OB OA.
So, from these two results, CD k (OB OA) and AB OB OA , we get
CD k AB . Therefore, AB // CD besides AB CD .
Otherwise, if we assume that AB CD , then by Proposition 2.6, we get
AC // BD which in turn implies that the lines A, C and B, D are
parallel. But this contradicts with the hypothesis.
Then, AB CD if and only if AB // CD and AC // BD .
Prepared by Begashaw M. 58
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2.4 The Classical Theorems
Definition: Three points A, B, C are said to be collinear if and only if AB // BC .
That means if all the lines lie on the same line it means. Two or more lines are
said to be concurrent if and only if they pass through the same point. The
common point at which the lines meet is called concurrency point or point of
concurrency.
Now let’s see the most useful theorems usually called Classical Theorems.
Menelao’s Theorem 2.4: Let D, E and F be points on the lines
B, C , A, C and A, B of triangle ABC respectively. Then, the points
BD CE AF
D, E and F are collinear if and only if . . 1
DC EA FB
Proof: Suppose D, E and F are points on the lines B, C , A, C and
A, B respectively such that D, E and F are collinear.
BD CE AF
We need to show that . . 1
DC EA FB
Draw a line through B parallel to AE that intersects DF at a point P as shown
in the figure 2.5 below.
Since BP // AE , DPB ~ DEC, FAE ~ FBP by Angle- Angle similarity
theorem.
Prepared by Begashaw M. 59
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Using this result together with the intercept Theorem,
DB PB FA EA
we get and .
DC EC FB PB
Multiplying these two equations, we get
DB FA PB EA DB FA EC
. . . . 1
DC FB EC PB DC FB EA
DB AF CE
. . 1, FA AF , EC CE
DC FB EA
BD AF CE
. . 1, BD DB
DC FB EA
BD CE AF
Hence, co linearity of D, E and F implies that . . 1 .
DC EA FB
Suppose BD CE AF
. . 1.....................................................................................(i)
DC EA FB
We need to prove that the points D, E and F are collinear. Consider line
D, E and suppose that F is not on this line.
Assume that F ' is another point on D, E such that D, E and F ' are collinear.
Then, by the forward part
BD CE AF '
. . 1................................................................(ii )
DC EA F ' B
So, from (i ) and (ii ) we get,
BD CE AF ' BD AF CE
. . . .
DC EA F ' B DC FB EA
AF ' AF
F ' B FB
F ' A FA
1 1
B F' BF
BA BA
B F' B F F' F
B F' B F
BD CE AF
Therefore, if . . 1, then the points D, E and F are collinear.
DC EA FB
Prepared by Begashaw M. 60
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Ceva’s Theorem 2.5: Let D, E and F be points on the lines B, C , A, C
and A, B of triangle ABC respectively. Then, the lines A, D , B, E
BD CE AF
and C, F are concurrent if and only if . . 1.
DC EA FB
Proof: Suppose the lines A, D , B, E and C, F are concurrent at a
point P . Draw lines through B and C which are parallel to the line A, D .
Extend the line B, E to a point Q and
C, F to a point R (Refer figure 2.6 below).
CE CQ AF PA
Since AP // QC, AP // RB and BR // CQ , we have , and
EA PA FB BR
BD RP BR
.
DC PC CQ
BD CE AF BR CQ PA
Now, multiplying the three equations yields, . . . . 1.
DC EA FB CQ PA BR
BD CE AF
Conversely, suppose . . 1.........................................(i)
DC EA FB
Let the point P be the intersection of the lines A, D and B, E . Assume
the line C, F does not pass through this point P . Then, we have another
point F ' on line A, B such that the line C, F ' passes through the point P .
BD CE AF '
Then, by the forward part . . 1................(ii ) .
DC EA F ' B
Prepared by Begashaw M. 61
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
So, from (i ) and (ii ) we get,
BD CE AF ' BD AF CE AF ' AF AF ' AF
. . . . 1 1
DC EA F ' B DC FB EA F ' B FB F'B FB
AF ' F ' B AF FB AB AB
, AF ' F ' B AB
F'B FB F'B FB
F ' B FB B F ' B F F ' F
Therefore, the lines A, D , B, E and C, F are concurrent whenever
BD CE AF
. . 1.
DC EA FB
Corollary 2.1: The three medians of any triangle are concurrent.
Proof: Consider ABC below (figure 2.7a) where AD, BE , CF are the medians.
From the definition of medians, we know that AF FB, BD DC, CE EA
BD CE AF
Thus, ( AF )( BD )(CE ) ( FB )( DC)( EA) . . 1.1.1 1
DC EA FB
Therefore, by Ceva’s theorem the medians are concurrent.
Corollary 2.2: In any triangle, the three altitudes are concurrent
Proof: The proof is given for the case when triangle ABC is acute angled
triangle. The case for obtuse follows the same reasoning and is left as exercise.
Consider AFC and AEB (refer figure 2.7 b above). In these triangles
BAE BAE (it is common angle) and BEA CFA (Both are right
angles). Hence, AFC ~ AEB by AA similarity theorem.
AF AC
But AFC ~ AEB
EA AB
Prepared by Begashaw M. 62
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Similarly in BDA and BFC , ABD ABD (it is common angle) and
BDA BFC (Both are right angles). Hence, BDA ~ BFC by AA similarity
BD AB BD AB
theorem. But BDA ~ BFC . Finally, by considering
BF BC FB BC
CE BC
CEB and CDA , we get CEB ~ CDA . CEB ~ CDA
DC AC
BD CE AF
Multiply theses three ratios above will lead to . . 1.
DC EA FB
Therefore, by Ceva’s theorem the three altitudes are concurrent.
Corollary 2.3: The angle bisectors of a triangle intersect at a common point.
Proof: Use the same reasoning as above.
Desargue’s Theorem 2.6: Let ABC and DEF be given where their
respective (corresponding) sides are parallel. Then, the lines A, D , B, E
AB AC BC
and C, F are concurrent and .
DE DF EF
Proof: Given ABC and DEF with AB // DE, AC // DF and BC // EF (refer
figure 2.8 below). Since AB // DE, AC // DF and BC // EF , then there exist
AB AC BC
scalars k , r , t such that k, r, t. Now we need to show that
DE DF EF
these scalars are equal in order to prove that the sides are proportional.
Using vector addition,
AB BC AC k DE t EF r DF r ( DE EF )
(k r ) DE (r t ) EF
Prepared by Begashaw M. 63
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Since DE is not parallel to EF , then by proposition 2.1, the equation
(k r ) DE (r t ) EF has a unique solution given by k r r t 0 k r t.
AB AC BC
Therefore, k where k is a constant. Here, there are two cases
DE DF EF
to be considered depending on the values of the constant k .
Suppose k 1 . In this case, assume the lines A, D and B, E intersect at
point O . Thus, AOB ~ DOE by AA similarity theorem (because AB // DE
OA AB
implies ODE OAB, OED OBA ). So, . Similarly, suppose
OD DE
the lines A, D and C, F intersect at point P . This also gives
PA AC
PAC ~ PDF by AA similarity theorem which in turn implies .
PD DF
AB AC BC
But from the first result, we have .
DE DF EF
OA AB BC PA OA PA
Hence, .
OD DE EF PD OD PD
Besides,
OA PA OA PA
1 1
OD PD OD PD
OA OD PA PD
OD PD
AD AD
PD OD
OD PD
D P DO P O
Hence, the lines A, D , B, E and C, F intersect at the same point and
then they are concurrent.
AB AC BC
Finally, if the ratio k 1 , then from k , we
DE DF EF
get AB DE, AC DF . Besides from the hypothesis, AB // DE, AC // DF .
Prepared by Begashaw M. 64
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Hence, ABED and ACFD become parallelograms (A quadrilateral having a pair
of parallel and congruent sides is said to be a parallelogram).
Thus, the lines A, D , B, E and C, F are all parallel.
Therefore, the lines A, D , B, E and C, F are concurrent.
Papu’s Theorem 2.7: Let A, C, E be points on a line k and B, D, F on a line l . If
AB // DE, BC // EF , then CD // FA.
Proof: Consider the diagram below. Suppose the lines k and l intersects at
some point O (Refer figure 2.9b).
Then, using the hypothesis AB // DE, BC // EF , and by The intercept Theorem,
OE OD OC OB
and .
OA OB OE OF
OC OD
Multiplying these equations, gives .
OA OF
Therefore, by The intercept Theorem, CD // FA. Again, suppose k // l (Refer
figure 2.9a). In this case, AE // BD and AB // ED .
Thus, AEDB is a parallelogram. So, AE BD and BF CE.
On the other hand,
AC AE EC AC AE CE BD BF FB BD FD
AC FD AC // FD
Thus, AC and FD are congruent and parallel.
Hence, AFDC is a parallelogram.
Since opposite sides of a parallelogram are parallel CD // FA.
Prepared by Begashaw M. 65
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 2.2
1. In Affine Space, suppose the lines l : X t (1,1,3), m : X (1,0,1) r (2,3,8) and
n : X (a,1, b) k (0,1,2) are concurrent. Find a and b . Answer : a 3, b 5
2. Find the value of a for which the lines l : y 3x 2, m : y ax 7 and
n : y 2 x 3 are concurrent.
3. The sign Law: Given ABC with AB c, BC a, AC b as shown below.
a b c
Then, using vector method, prove that .
sin A sin B sin C
4. Find the vector, parametric and symmetric equations of a line through the
2 2
intersection of m : y x and n : y 2 x which is parallel l : X t .
4 6
5. Let D, E and F be points on the sides B, C , A, C and A, B of
ABC respectively where AB 12, EA 3, BC 15, FB 8, BD 6 . If the lines
A, D , B, E and C, F are concurrent, then calculate the length CE .
Answer :CE 9
6. Let D, E, F be points on the sides B, C , A, C and A, B of triangle ABC
DB 3 CE 12 AF
such that and . Then, find
DC 4 EA 5 FB
a) If the points D, E, F are collinear.
b) If A, D , B, E and C, F are concurrent.
7. Suppose D, E and F are collinear points on the lines B, C , A, C and
AG BD CE
A, B of ABC . If for any point G , . . 1 , prove that F G .
GB DC EA
Prepared by Begashaw M. 66
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
8. Let k // l and let A, C, E be on k and B, D, F be on l such that the lines
A, B , C, D and E, F <E,F> are concurrent.
Then, prove that AC / CE BD / DF .
9. Prove that the angle bisectors of a triangle are concurrent at a point.
10. Consider the vertices A (1,2,5), B (1,8,1), C (7,2,1) of an equilateral
triangle ABC . Suppose the points D (4,5,1), E (4,2,2), F (1,5,2) are on the
sides B, C , A, C and A, B of ABC respectively. Prove that the lines
A, D , B, E and C, F are concurrent.
11. Suppose the medians AP, BQ, CR of ABC intersect at point T . Then,
TP TB TR
show that 1
AP BQ CR
12. Let ABC and DEF be triangles with the condition AB / DE AC / DF .
Show that BC || EF .
13 .Let D, E, F be points on sides BC , CA, AB of A BC respectively such that
the Cevians AD, BE , CF are concurrent. Show that if M , N , P are points on
EF , FD, DE respectively, then the lines AM , BN , CP are concurrent if and only
if the lines DM , EN , FP are concurrent.
14. If a circle is inscribed in ABC such that M , N and P are points of tangency
on the sides BC, CA and AC , prove that AM , BN and CP are concurrent.
15. The bisector of any interior angle of a non-isosceles and the bisectors of the
two exterior angles at the other vertices are concurrent.
16. In ABC , let P AB and Q AC such that PQ // BC . Then, prove that PC
and QB concur (intersect ) at a point on the median AM .
17. Let P, Q, R be points on A BC distinct from A, B, C respectively. Then
sin CAP sin ABQ sin BCR
AP, BQ, CR are concurrent if and only if 1.
sin APB sin QBC sin RCA
18. Prove that the angle bisectors of a triangle are concurrent at a point.
Prepared by Begashaw M. 67
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
CHAPTER-3
Orthogonal (Isometric)Transformations
3.1 Introductions
In order to introduce the concept of isometrics, first let’s consider the notion of
distance. Distance is a real valued non-negative function denoted by d ( P, Q)
which assigns to any pair of points in the plane or space a non-negative real
number satisfying the following conditions:
i) d ( P, Q) d (Q, P)
ii ) d ( P, Q) 0d ( P, Q) 0 P Q
iii ) d ( P, R) d ( P, Q) d (Q, R)
Here, the third property is known as triangle inequality and equality occurs if
and only if the points P, Q, R are collinear points.
Note: The notation d ( P, Q) stands to mean the distance from P to Q and
equivalently denoted by d ( P, Q) PQ PQ . It is the length of the line segment
between P and Q which shows that line segment is the shortest path between
two points.
In Euclidean geometry, distance between two points P and Q in a plane is given
by d ( P, O) PQ ( x1 x2 )2 ( y1 y2 )2 , where P ( x1 , y1 ) and Q ( x2 , y2 )
It can be easily verified that this distance formula satisfies the above three
conditions. Throughout this text, the writer uses f ( P) f (Q) to mean the
distance between f ( P) and f (Q) , PQ to mean the distance between P and Q .
In a plane, it is common to see the case where the norm of a vector or the
distance (length) between two points to have the same measure after the vector
or the points have been mapped to other vectors or points by some mappings.
Prepared by Begashaw M. 68
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
In such cases, whenever f (v) v (That is the norm of the image vector is
equal to the norm of the vector before mapping) for any vector v and
f ( P) f (Q) PQ (That is the distance between the images of two points is equal
to the distance between the points) for any two points P and Q , we say that f is
norm, length or distance preserving. That is really what we mean isometric or
orthogonal transformations. Now let’s have the formal definition.
3.2 Definition and Examples of Isometries
Definition: An isometric transformation of a plane is a transformation from a
plane on to itself which preserves distances.
That is, f is an isometric transformation if for any two points P and Q in the
plane, d ( f ( P), f (Q)) d ( P, Q) .
Isometrics in Euclidean space are sometimes named as orthogonal
transformations or rigid motions.
Examples:
1. Verify whether the following transformations are isometries or mot.
a) f : R R given by f ( x) x 5
b) f : R2 R2 given by ( y 1, x 7)
c) f : R2 R2 given by f ( x, y) (3x,3 y)
d) f : R2 R2 given by f ( x, y) (2 y 9,2 x 9)
Solution:
a) Since the distance between any two points x and y in R is given by
d ( x, y) x y and d ( f ( x), f ( y)) x 5 ( y 5) x y d ( x, y) .
Hence, f is an isometry.
b) Here, if we take any two points P ( x, y), Q ( z, w) , we have
d ( P, Q) ( z x) 2 ( y w) 2 .
Prepared by Begashaw M. 69
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Besides, P' f ( P) ( y 1, x 7), Q' f ( Q) ( z 1, w 7) and thus
d ( P' , Q' ) d ( f ( P), f (Q))
[( z 1) ( y 1)]2 [( w 7) ( x 7)]2
( z y ) 2 ( w x) 2
This shows that d ( P' , Q' ) d ( f ( P), f (Q)) d ( P, Q) .
Hence, it is an isometry.
c) Here, for any two points P ( x, y), Q ( z, w) , we have
d ( P, Q) ( z x) 2 ( y w) 2 .
Besides,
P' f ( P) (3x,3 y), Q' f ( Q) (3z,3w) and thus
d ( P' , Q' ) d ( f ( P), f (Q))
(3z 3z ) 2 (3w 3 y ) 2
9( z y ) 2 9( w x) 2
3 ( z y ) 2 ( w x) 2
This shows that d ( P' , Q' ) d ( f ( P), f (Q)) 3d ( P, Q) d ( P' , Q' ) d ( P, Q) .
Hence, it is not an isometry.
d) Using similar procedures as in part (c), it is not isometric transformation.
2. If f is an isomteric transformation and v (a,6) is a vector such that
f ( a,3) (3,4) , find a .
Solution: Since f is an isomtery, for any vector v we have f (v) v .
Thus, f ( a,3) (3,4) f (a,3) (a,3)
a 2 9 9 16
a 2 9 25 a 2 16 a 4
Prepared by Begashaw M. 70
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3. Prove that any isometry preserves dot product and conclude that an isometry
preserves angle.
Proof: Let v and w be any two vectors and f be an isometry. Since an isometry
preserves length (norm), we have f (v) v , f (w) w and
f (v) f ( w) v w .
But f (v) f ( w) v w
2 2
f (v) f ( w) vw
( f (v) f ( w)).( f (v) f ( w)) (v w).(v w)
f (v). f (v) f (v). f ( w) f ( w). f (v) f ( w). f ( w) v.v v.w w.v w.w
2 2 2 2
f (v) 2 f (v). f ( w) f ( w) v 2v . w w
2 2 2 2
v 2 f (v). f ( w) w v 2v.w w
2 f (v). f ( w) 2v . w
f (v). f ( w) v . w
Hence, f (v). f (w) v . w shows that f preserves dot product. Besides, if is the
angle between the non-zero vectors v , w and is the angle between the image
vectors f (v), f (w) , then we have
f (v ). f ( w) v.w
cos cos . (Why ?)
f (v) f ( w) vw
Prepared by Begashaw M. 71
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3.3 Properties of Isometric (Orthogonal) Transformations
Propositions 3.1: The inverse of an isometry is an isometry.
Proof: Let f be an isometry. Now let P and Q be any two points. We need to
show f 1 ( P) f 1 (Q) P Q . Since f is an isometry,
1 1 1 1
f ( P) f (Q) f ( f ( P)) f ( f (Q))
( f f 1 )( P)) ( f f 1 )(Q)) i ( P) i (Q) P Q
f 1 ( P) f 1 (Q) P Q
Hence, for any isometry f , f 1 is also an isometry.
Proposition 3.2: The composition of any two isometries is again an isometry.
Proof: Let f and g be any two isometries. We need to show their composition
f g is also an isometry. Let P and Q be any two points.
Since f is an isometry, f ( P) f (Q) P Q .
But g is also an isometry, so g ( f ( P)) g ( f (Q)) f ( P) f (Q) .
Combining these two results together, we get that
g f ( P) g f (Q) g ( f ( P)) g ( f (Q)) f ( P) f (Q) P Q .
Hence, the composition f g is an isometry.
Proposition 3.3: An isometry maps distinct points into distinct points.
Proof: Let f be an isometry. We need to show A B f ( A) f ( B).
But, A B A B 0 d ( A, B) 0 . Besides, since f is an isometry
d ( A, B) d ( f ( A), f ( B)) .
So, d ( A, B) 0 d ( f ( A), f ( B)) 0 f ( A) f ( B) 0 f ( A) f ( B).
Proposition 3.4: Any isometry preserves between ness. That means if the
points A' , B' , C ' are the image of the distinct points A, B, C under a given
isometry such that B is between A and C , then B' is also between A' and C ' .
Proof: Let f be an isometry such that A' f ( A), B' f ( B), C' f (C). Then,
A' B' AB, B' C ' BC , A' C ' AC .
Prepared by Begashaw M. 72
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Besides, as B is between A and C , AB AC (from definition of betweeness).
Now suppose B' is not between A' and C ' (It may be either to the left of A' or
to the right of C ' refer figure 3.1).
Thus, if we assume B' is to the right of C ' , then A' C' A' B' AC AB
(Because A' B' AB, A' C' AC ).
But this contradicts with the hypothesis AB AC . Similar argument holds if
B' is assumed to be on the left of A' .
Hence, any isometry preserves betweeness.
Proposition 3.5: If three points A, B, C are collinear, then their images A' , B' , C '
are also collinear. This means any isometry preserves co linearity.
Proof: Apply Triangle inequality and proposition 3.4 to arrive at a
contradiction.
Proposition 3.6: An isometry maps lines into lines and parallel lines into
parallel lines.
Proof: Let l be any line and f be an isometry. Take any two distinct points
A, B on l such that f ( A) A' and f ( B) B'.
But A B f ( A) f ( B) A' B'. Hence, A' and B' are distinct points and
determine a unique line l '. Besides, since f preserves co-linearity for any point
P on l , f ( P) P' is on l '. Moreover; if l // m , then f (l ) // f (m).
Otherwise, if f (l ) and f (m) are not parallel, they intersect at a point Q .
So, there exist two different points A l and B m such that f ( A) Q and
f ( B) Q which implies f ( A) f ( B) . But this contradicts with proposition 3.3,
A B f ( A) f ( B) . Thus, l // m f (l ) // f (m).
Prepared by Begashaw M. 73
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Theorem 3.1 (The Three-Point Theorem):
Two isometrics with same image at three non-collinear points are equal.
Proof: Suppose f and g are two isometries such that
f ( A) g ( A), f ( B) g ( B), f (C) g (C), where A, B and C are non-collinear
points. We need to show that f g , i.e f ( P) g ( P) for any point P in the
plane. Now, suppose f g . Then, there exists at least one point P such that
f ( P) g ( P) . Let f ( P) P' , g ( P) P" as shown in figure 3.2.
Since f and g are isometrics, A' P' AP, A' P" AP.
From these two equations, we get A' P' A' P", which shows that A' is on the
perpendicular bisector of P' P" (If a point is at equal distance from the end
points of a line segment, then it lies on the perpendicular bisector of that line
segment).
Similarly, if we consider points B and C separately in the plane of A we get
that B' and C ' are also in the perpendicular bisector of P' P".
That means they all lie on the same line which yields that A' , B' and C ' are
collinear. But, this in turn implies that A, B, C are collinear because any
isometrics preserves co-linearity.
However, this contradicts the hypothesis that A, B, C are non-collinear.
Consequently, f g.
Prepared by Begashaw M. 74
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 3.1
4 3 3 4
1. Show that the transformation given by f ( x, y) x y, x y is an
5 5 5 5
isometry.
4 4 3
2. If f ( x, y) ax y, x y is an isometry, find the value of a.
5 5 5
Answer : a 3 / 5
3. Let P and Q be two points and be an isometry with ( P) P, (Q) Q .
Show that (M ) M where M is the mid point of P and Q .
4. Suppose is a mapping with the property that ( P) (Q) PQ for all points
P and Q . Show that is an isometry. (Hint: Simply show that is a
transformation)
5. Prove that any isometry preserves the Cauchy- Schwarz Inequality.
6. Using triangle inequality, prove that an isometry maps lines in to lines.
7. Let C be a circle with center O and radius r . Prove that if f is an isometry,
then f (C ) is a circle with center f (O) and radius r.
9. Prove that every isometry f preserves norm of a vector.
10. Prove that every isometry f preserves dot product and conclude that it also
preserves angle measures.
Prepared by Begashaw M. 75
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3.4 Fundamental Types of Isometric Transformations
3.4.1 Translation
Definition: A mapping T : R 2 R 2 is called translation if there exists a vector
v such that T ( P) P v for every point P in R 2 .
x' x a
In other word, for P ( x, y), v (a, b), T ( P) T ( x, y) ( x' , y' ) where
y' y b
The vector v is called translator vector. The translation with translator vector
v is sometimes denoted by Tv . In translation problem, whenever any two of
( x, y), ( x' , y' ), or (a, b) (the pre image, the image or the translator vector) are
given, the third can be uniquely determined from the translation equation.
Examples:
1. Let T be a translation that takes the point (3,4) to (5,3). Find the equation for
the translation T.
Solution: Let v (a, b) be the translator vector. Then for any point P( x, y) , we
x' x a
have T ( x, y) ( x' , y' ) where .
y' y b
Particularly, for (3,4) , we have
T (3,4) (3 a,4 b) (5,3) 3 a 5,4 b 3 a 2, b 1.
Therefore, T ( x, y) ( x 2, y 1) .
2. For any two translations Tv and Tw , if Tv (1,2) 2Tw (4,7) and
Tw (9,2) 3Tv (1,3), find the equations of these translations.
Solution: Let v (a, b) and w (c, d ) . Then, for any object ( x, y),
Tv ( x, y) ( x a, y b) , Tw ( x, y) ( x c, y d ). But,
Tv (1,2) 2Tw (4,7) (1 a,2 b) (8 2c,14 2d )
a 2c 7
.......................................................(i )
b 2d 12
Tw (9,2) 3Tv (1,3) (9 c,2 d ) (3 3a,9 3b)
c 3a 6
.............................................................(ii )
d 3b 11
Prepared by Begashaw M. 76
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Collecting like terms together from (i ) and (ii ) , we get the following systems;
a 2c 7 b 2d 12
a 1, c 3 , b 2, d 5
c 3a 6 d 3b 11
v (a, b) (1,2), w (c, d ) (3,5)
This implies that v (a, b) (1,2), w (c, d ) (3,5)
Therefore, Tv ( x, y) ( x 1, y 2) and Tw ( x, y) ( x 3, y 5).
For instance, Tv (1,2) (2,4), Tw (4,7) (1,2) Tv (1,2) 2(1,2) 2Tw (4,7).
Proposition 3.7 (Properties of Translations):
a) The translator vector of a translation is unique.
b) The composition of translation Tv and Tw is again a translation by v w .
c) The inverse of a translation is a gain a translation with opposite vector.
d) The image of a line under a translation is a line parallel to the given line.
e) The image of a vector under a translation is an equal vector.
Proof: (Follows from the definition)
Examples:
1. If Tv (1,2) (3,1), (Tv Tw )(1,5) (3,7) , then find the equations of Tv and Tw .
Solution: Here,
Tv (1,2) (3,1) v (1,2) (3,1) v (2,3)
Tv ( x, y) ( x 2, y 3)
A gain , from part (b) of the above properties,
(Tv Tw )(1,5) v w (1,5) (3,7)
(2,3) w (1,5) (3,7) w (0,5)
Therefore, Tw ( x, y) ( x, y 5) .
2. If T 2 (1,3) (7,1) , then find the equation of T .
Solution: Here, T 2 ( P) T T ( P) 2v P . In general, T n ( P) nv P .
Hence, T 2 (1,3) (7,1) T T (1,3) (7,1) 2v (1,3) (7,1)
v (3,1) T ( x, y) ( x 3, y 1)
Prepared by Begashaw M. 77
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Remark: The important concept that students should bear in mind about
translation is that translation Tv can be applied to arbitrary shapes of object
point –by-point. Each point of a given shape S is translated by v and the
collection of these translated points gives the translated image S ' of S. This is
denoted by S ' Tv (S ). So, to determine the image S ' when S is a polygon it is
sufficient to translate its vertices. For instance, if PQRT is an arbitrary four
side polygon, where P, Q, R, T are its vertices, to find the image polygon
under a translation Tv , it suffices to determine only the images of the vertices.
i.e. Tv ( P) P' , Tv (Q) Q' , Tv ( R) R' , TV (T ) T ' . So that the image
quadrilateral is P' Q' R'T ' . This means that other points of the quadrilateral
between the vertices are assumed to be translated (moved) equal distance in the
same direction.
Examples: Let T be a translation by the vector (1,2) . Find the image of
a) ABC whose vertices are A(0,0), B(3,0) and C (0,4)
b) the line l : x 2 y 6 .
Solution: Let P ( x, y) by any object in the plane containing ABC . Then,
T ( P) T ( x, y) ( x 1, y 2) by definition of translation. Thus,
a) A' T ( A) (1,2), B' T ( B) (4,2), C' T (C) (1,6) . Hence, the image of
ABC under T will be A' B' C ' with vertices A' (1,2, ), B' (4,2) and C ' (1,6) .
b) This problem can be done using two methods.
Method I: Take two points A and B on l : x 2 y 6 .
Say A (0,3), B (6,0) .Then, find the images of these points under T .
T ( A) A' (1,1), T ( B) B' (7,2) .
Now form the equation of a line l ' passing through A' and B' . This line will be
the image of l under T . This is because from the previous explanation all
other points between A and B or on the line l will be translated the same
2 1 3 1
distance in the same direction. Hence, l ': y mx b , where m
7 1 6 2
1
l ': y x b . Taking either A' or B' on l ' we can determine b .
2
1 3 1 3
Using A' ,1 .1 b b . Thus, l ': y x x 2 y 3 .
2 2 2 2
Prepared by Begashaw M. 78
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Method II: Take any arbitrary point P ( x, y) on l . Then, find P' . Finally,
substitute in the equation of l . P' ( x' , y' ) T ( x, y) ( x 1, y 2) . From the
general equation of translation, x' x 1, y' y 2 x x'1, y y'2 .
Subsisting these values in equation of l , we have
l : x 2 y 6 l ': ( x'1) 2( y'20 6 l ': x'2 y' 3
Note that it is advisable to use Method II. Because most of the time Method I
is subjected to arithmetic errors and is a bit longer than Method II. In this
1
example one can observe m' m which shows that l ' is parallel to l .This
2
will enable us to state the following proposition in general.
Proposition 3.8: Any translation is a dilatation.
Proof: In chapter 1, we have seen that to show a given transformation is a
dilatation, first we have to show it is a collineation and then it maps any line to
parallel line. Let Tv be arbitrary translation with translator vector v (h, k ) and
let l : ax by c 0 be any line. Now, for any point ( x, y) on this line,
x' x h
Tv ( x, y ) ( x' , y ' ) where . Solving for x and y in terms of x' and y '
y' y k
yields x x'h, y y'k .
Then, substitute these values in the equation of l ( Just using Method II)
l : ax by c 0 a( x'h) b( y'k ) c 0 ax'by'(c ah bk ) 0
But this is equation of a line which means that Tv is a collineation. Thus, for
any line l : ax by c 0 its image under a translation Tv is a line given by
Tv (l ) l ': ax'by'(c ah bk ) 0 .
Besides, comparing the equations of l and l ' , we observe that they do have the
same slope. This shows they are parallel which in turn implies that Tv maps any
line into a parallel line. Consequently, any translation is a dilatation.
Prepared by Begashaw M. 79
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Proposition 3.9: For two given points, P and Q there is a unique translations
that take P to Q this is usually denoted by TP ,Q . Besides, the translator vector
is given by v Q P PQ .
Proof: Let T1 ( x, y) ( x a, y b), T2 ( x, y) ( x m, y n) be any two translations
such that T1 ( P) Q and T2 ( P) Q . We need to show T1 and T2 are identical.
But to show this it suffices to show that their translator vectors are equal. Let
P (c, d ), Q (e, f ) . Then, T1 ( P) (c a, d b) (e, f ) and
T2 ( P) (c m, d n) (e, f ) . Thus, (c a, d b) (c m, d n) which implies,
(a, b) (m, n) . So, T1 ( P) T2 ( P) for any point P . Hence, T1 T2 .
Therefore, TP ,Q is unique. Besides, Tv ( P) v P Q v Q P PQ .
Corollary 3.1:
For any four points P, Q, R and S , if TP,Q ( R) S , then TP,Q TR,S .
Proof: Let v be a translator vector of the translation T . Then,
Tv ( P) v P Q v Q P PQ
Tv ( R) Q P R S
Q P S R PQ RS TPQ TRS
Prepared by Begashaw M. 80
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 3.2
1. Let Tv be a non-identity translation in R 2 . Prove that Tv (l ) l l // v .
2. For any four non-collinear points A, B, C, D , TAB TDC ABCD is a
parallelogram.
3. Let f be a translation with translator vector v (a, b) . If f 4 (2,7) (6,5) ,
find the translator vector v and f (4,3)
4. Find the image of the circle C : x 2 y 2 4 y under a translation T which maps
(1,2) to (0,7) .
5. Let Tv be a translation with translator vector v . Then, show that
n
Tv ( X ) X nv and Tv ( X ) X nv for any object X .
n
6. For any two translations, Tv and Tw , their composition is commutative. i.e
Tv .Tw Tw Tv .
Prepared by Begashaw M. 81
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3.4.2 Reflection
When we look at our selves through a mirror, it is common to see our face. At
this time, our true face is displayed through the mirror and the displayed face is
called our image under the mirror. But in the theory of Optics one can find that
this image is found to be at equal distance from the mirror as our real face. This
means that the distance between the image and the mirror is the same as the
distance between our face and the mirror and this is what we mean by the
concept of reflection.
Definition: Given a line l and a point P . Then P ' is said to be the reflection
image of P on the line l if and only if PP ' is perpendicular to l and
PM P ' M , where M is the point of intersection of PP ' and the line l . In
other words, P and P ' are located on different sides of l but at equal distances
from the line l . In this case, P ' is said to be the mirror image of P and the line
l is said to be line of reflection or axis of symmetry.
Notation: Reflection on l is usually denoted by Sl .
P, if P l
So, Sl ( P)
P ', if P l and l is the perpendicular bisector of PP '
Examples: Common reflection equations.The following simple equations of
reflections can be easily derived using the definition.As they are most
coomonly, used in this chapter, please bear them in mind. Let ( x, y) be any
point.
a) Reflection on the x-axis: S x ( x, y) ( x, y)
b) Reflection on the y-axis: S y ( x, y) ( x, y)
c) Reflection on a vertical line x a : S a ( x, y) (2a x, y)
d) Reflection on a horizontal line y b : S b ( x, y) ( x,2b y)
e) Reflection on the line y x : Sl ( x, y) ( y, x)
Prepared by Begashaw M. 82
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Example: If the image of the point (3,5) on a vertical line is (7,5) , find the
equation of the line.
Solution: Let the line be l : x a .Then, reflection on this line will have
equation S a ( x, y) (2a x, y) .
Particularly, S a (2,5) (7,5) (2a 3,5) (7,5) 2a 3 7 a 2 .
Hence, the line is l : x 2 .
Problematic situations: In solving reflection problems, there are three
possible situation:
I) Given the point P and line l , we need to find the image point P ' under Sl .
II) Given the image point P ' and the line l , we need to find the object point P .
III) Given and the objetc point P and its image point P ' , we need to find the
equation of l .
But, in each case, the basic definition of reflection can be used to determine
any of the required values.
Examples:
1. Find the image of the point (4,6) by a reflection on the line l : y 2 x .
Solution: Let the image point be P' ( x' , y' ) . Then, by definition, the midpoint of
the P and P’ is on the line of reflection.
x'4 y'6 y'6
That is ( , ) l x'4 y' 2 x'2..................(i)
2 2 2
A gain the line through P and P’ is perpendicular to the given line. Hence, its
1
slope must be m . Thus,usin P and P’, we have
2
y'6 1 x'
y' 8.....................................................(ii )
x'4 2 2
x' 12 34
From the two equations, we get 2 x'2 8 x' , y ' .
2 5 5
12 34
Therefore, the image point is ( x' , y' ) ( , ).
5 5
Prepared by Begashaw M. 83
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2. Let the image of the point (2,5) by a reflection Sl be the point (5, 2) . Find
the equation of the line of reflection l .
Solution: Let P (2,5) , P ' (5, 2) and l : y mx b. Since PP ' is
perpendicular to l , the product of the slopes of l and the line through P and
P ' must be 1 . But the slope of the line through P and P ' is 1 . So, slope m
of l is 1 . Thus, l : y mx b y x b.
Again, since l is the perpendicular bisector of PP ' , l passes through the mid
P P ' (2,5) (5, 2) 7 7 7 7
point of P and P ' . That is, ( , ) l . Thus, ( , )
2 2 2 2 2 2
7 7
satisfies the equation of l . So, ( ) b b 0 .
2 2
Therefore, equation of l is y x.
Proposition 3.10: Properties of Reflection
a) For any two reflections, Sl and Sm , Sl Sm Sm Sl
b) For any reflection Sl , Sl1 Sl , the inverse of any reflection is itself.
c) Reflection is an involution.
Proof: (Follows from the definition)
Part (a) of the above proposition shows that composition of reflections in
general is not commutative.
However, there is a special case in which S m S l S l S m . In
general, “Under what condition, product (composition) of reflections will be
commutative?” is basic question to be answered. Since it needs some basic
theorems, it is not easy to answer this question for the moment. We will get the
full answer to this question later on in section 3.5 of this chapter.
Proposition 3.11: Let l and k be any two lines. Then, the following conditions
are equivalent.
a) S l S k S k S l b) (S l S k ) 2 i c) (S l S k ) 1 S l S k
Proof: We need to show a b .
Prepared by Begashaw M. 84
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
(S l S k ) 2 (S l S k ) (S l S k ) (S l S k ) (S k S l )
S l (S k S k ) S l S l i S l S l S l i
We need to show b c
(S l S k ) 2 i (S l S k ) (S l S k ) i
( S l S k ) ( S l S k ) ( S l S k ) 1 ( S l S k ) 1
( S l S k ) 1 S l S k
We need to show c a
(S l S k ) 1 S l S k S l S k S k S l , because
1 1 1
(S l S k ) 1 S k S l , S l S l , S k1 S k
The generalized Analytic Equation of Reflection
Theorem 3.2: (The Generalized Reflection Theorem): Let l : ax by c 0
be any line and S l be a reflection on line l .
2a(ax by c)
x ' x
a2 b2
Then, for any point ( x, y) , S l ( x, y) ( x' , y' ) where
y ' y 2b(ax by c)
a2 b2
Proof: From the definition of reflection, the line through P( x, y) and P' ( x' , y' ) is
perpendicular to the given line and the midpoint of ( x, y) and ( x' , y' ) is on the
line l. Refer the figure 3.3.
a
As the slope of the given line is m , the slope of the line through P( x, y)
b
b
and P' ( x' , y' ) is m' .
a
Prepared by Begashaw M. 85
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Thus, the equation of the line through P( x, y) and P' ( x' , y' ) is given by
y' y b
a( y' y) b( x' x)......................................................................(i)
x' x a
x x' y y '
Now, the midpoint of P( x, y) and P' ( x' , y' ) is on l means , is on l .
2 2
x x y y
So, a b c 0 ax'by' 2c ax by...............(ii )
2 2
Combining these two equations gives us
bx ay bx ay
...........................................................(iii )
ax by 2c ax by
Now, solve these equations for x' and y ' . In this equation, by multiplying the
first equation by 'b' , the second by 'a' and adding them we obtain,
a 2 x'b 2 x' b(bx ay a(2c ax by )
b(bx ay ) a(2c ax by )
x'
a2 b2
b 2 x a 2 x 2a 2 x 2aby 2ac
a2 b2
x(a 2 b 2 ) 2a(ax by c)
a2 b2
2a(ax by c)
x
a2 b2
Similarly, multiplying the first equation by 'a' , the second by 'b' and adding
the result gives,
b(2c ax by ) a (bx ay )
y'
a2 b2
a 2 y b 2 y 2b 2 y 2abx 2bc
a2 b2
y (a 2 b 2 ) 2b(ax by c)
a2 b2
2b(ax by c)
y
a2 b2
Prepared by Begashaw M. 86
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Examples
1. Find the image of the point (2,3) by a reflection on the line l : 3x 2 y 5 0
Solution: Given ( x, y) (2,3) and from l : 3x 2 y 5 0, a 3, b 2, c 5 .
Then,
2a(ax by c) 6(6 6 5) 4
x ' x 2
a b
2 2
94 13
y ' y 2b(ax by c) 3 4(6 6 5) 59
a2 b2 94 13
4 59
Therefore, S l (2,3) ( , ).
13 13
2. Given S l (a, b) (2,5) where l : x y 1 0 . Find the value of the point (a, b).
Solution: Using the generalized reflection equation derived in the above
theorem,
2(a b 1)
2 a a (a b 1) b 3
1 1
S l (a, b) (2,5)
5 b 2(a b 1) b a b 1 a 4
11
3. Given the lines m : y 2 x 1 and n : y 2 x 3 . Find the image of the point
(1,1) by a product of reflection on line m followed by line n .
Solution: We need to find S n S m (1,1)
First calculate S m (1,1) using reflection equation as
4(2 1 1) 3
x' 1 4 1 5
S m (1,1) ( x' , y ' )
y ' 1 2(2 1 1) 9
4 1 5
6 9
4( 3)
3 5 5 21
x' '
Now, S n S m (1,1) S n ( S m (1,1)) S n ( 3 , 9 ) 5 4 1 5
5 5 6 9
2( 3)
9 5 5 3
y ' '
5 4 1 5
21 3
Therefore, S n S m (1,1) ( , ) .
5 5
Prepared by Begashaw M. 87
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
4 Given the line l : y (tan ) x .
x' x cos 2 y sin 2
a) Show that S l ( x, y) ( x' , y' ) where
y ' x sin 2 y cos 2
b) Calculate the image of the point (5,5) by a reflection on a line l : y 3x.
Solution:
a) Here, l : y (tan ) x (tan ) x y 0 . So, S l ( x, y) ( x' , y' ) can be
calculated using direct formula
2a (ax by c)
x' x , a tan , b 1, c 0
a2 b2
2 tan (tan x y )
x
tan 2 1
x tan 2 x 2 y tan
, tan 2 1 sec 2
sec
2
x sin x cos 2 2 sin cos
2
x(cos 2 sin 2 ) y (2 sin cos )
x cos 2 y sin 2
Similarly, y ' can be calculated as follow:
2b(ax by c)
y' y , a tan , b 1, c 0
a2 b2
2(tan x y )
y
tan 2 1
y tan 2 y 2 x tan
, tan 2 1 sec 2
sec 2
y sin 2 y cos 2 2 x sin cos
y (cos 2 sin 2 ) x(2 sin cos )
x sin 2 y cos 2
b) Here, y 3x tan 3. Then, using trigonometric relation, we can find
sin and cos as follow.
Prepared by Begashaw M. 88
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3 1
tan 3 r 10 sin , cos
10 10
4 3
cos 2 cos 2 sin 2 , sin 2 2 sin cos
5 5
x' x cos 2 y sin 2 5( 54 ) 5( 53 ) 1
Thus, ( x' , y ' ) (1,7)
y ' x sin 2 y cos 2 5( 5 ) 5( 5 ) 7
3 4
Here, one can check the correctness of the answer using direct formula of
reflection.
6(15 5)
x' 5 1
9 1
S l (5,5) ( x' , y ' ) ( x' , y ' ) (1,7) which agrees with
y ' 5 2(15 5) 7
9 1
our previous result.
Reflecting Geometric Figures: So far we saw how to find the image of a point
by reflecting on a line. But it is also possible to find the images of different
figures (like lines, circles, ellipse, rectangles) under a reflection. So, here under
let’s see some examples.
Reflecting a circle or an ellipse on a line:
To find the image of a circle under a reflection on a line l , we follow the
following procedures.
First: Identify the center O and radius r of the given circle C .
Second: Find the image O' of the center O of the circle by a reflection on l .
Third: Write the equation of the image circle using the center O' and the same
radius r . (Why we use the same radius? We use the same radius because
reflection is an isometry that preserves length)
We also use the same procedure to reflect an ellipse. That is first identify the
center of the ellipse and its minor and major axes, then reflect the center of the
ellipse, finally write the equation of the ellipse using the image center and the
same axes as the given ellipse. Please bear in mind that these procedure works
for all types of isometries.
Prepared by Begashaw M. 89
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Examples: Find the images of the circle C : x 2 y 2 2 x 6 y 6 0 and the
ellipse E : 4 x 2 9 y 2 36 under a reflection on the line l : y x 1 .
Solution: For clarity, let’s follow the above procedure directly.
First: Identify the center and radius of the circle C : x 2 y 2 2 x 6 y 6 0 .
By completing square, we get x 2 y 2 2 x 6 y 6 0 ( x 1) 2 ( y 3) 2 4 .
Hence, the center is O (1,3) and its radius is r 2 .
Second : Find the image of the center O (1,3) by a reflection on l : y x 1
Using, reflection formula we get the image of the center to be
O' S l (1,3) (2,0)
Third: Write the equation of the image circle using the image center and the
radius of the given circle.
That is the image circle is C ': ( x 2) 2 ( y 0) 2 4 x 2 y 2 4 x
x2 y2
In standard form, the ellipse is written as E : 4 x 2 9 y 2 36 1.
9 4
Thus, the major axis is a 9 , the minor axis is b 2 and the center is C (0,0) .
Besides, the image of the center is C ' (1,1) . Therefore, the image of the ellipse
( x 1) 2 ( y 1) 2
becomes E ': 1 4( x 1) 2 9( y 1) 2 36.
9 4
Reflecting a line on a line:
To find the image of a line m under a reflection on another line l , we follow
the following procedures.
First: Select any two points P and Q from line m .
Second: Find the image P' and Q ' of P and Q by a reflection on l .
Third: Write the equation of the line through P' and Q ' which is image of m .
These procedures work for any pairs of line m and l , even when they are
intersecting. Besides, these procedure works for all types of isometries.
Prepared by Begashaw M. 90
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Example: Find the images of the lines m : y 2 x 3 and n : y x 2 under a
reflection on the line l : y x 1 .
Solution: For clarity, let’s follow the above procedure directly.
First: Select two points P and Q from line m . Say, P (0,3), Q (1,1) .
Second: Find the image P' and Q ' of P and Q by a reflection on l .
Using, the generalized reflection formula we get the images t be
P' (2,1), O' (0,0)
Third: Write the equation of the line through P' and Q ' which is image of m .
1
That is the image line is m': y x . Use similar argument to find image of n
2
Application (Shortest Path Problem): Suppose Kumlachew wants to fetch
water from a river on his way while he is going from his farm land (found at
location A ) to his home (found at location B ) as shown in figure 3.4 below.
Suggest the shortest path that takes kumlachew from his farm land to the river
and then to his home and prove that the path you suggested is really the
shortest path.
Solution: In this problem, any one can guess that Kumlachew should go
straight to some point on the river bank, fetch water and then go to his house.
But the problem is to what point of the river should he go so as to minimize the
length of the road. First of all assume that the river bank to which Kumlachew
wants to go forms straight line say L.
Prepared by Begashaw M. 91
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Claim: Let B' be the reflection of B in L (straight side of the river bank) and
let P A B' L. Then, A P B is the shortest path.
Proof: It suffices to show that if Q P is any other point on L, then
AP PB AQ QB. Since B' is the reflection of B in L and reflection is
an isometry, QB QB'. Thus, AQ QB' AQ QB. By the same argument,
PB PB' implies AP PB AP PB'.
Besides, as A, P, B' are collinear (do you see why?), we have
AP PB' AB'. But, by Triangle Inequality, AB' AQ QB'.
Therefore, AP PB' AQ QB' AP PB AQ QB.
This proves our claim.
Example (Application): Suppose an ant moves in straight lines from position
A (1,5) to position B (2,4) by touching the line L : y x at some point P .
Find the coordinate of P that minimizes the total path of the ant.
Solution: This is a particular case of the shortest path problem. As in the
procedures of the above proof;
First: Find the reflection image of B (2,4) on the given line L : y x . Say, B'
Second: Form the equation of the line through A (1,5) and B' . Say,
m : y ax b
Third: Find the intersection of the line L : y x and m : y ax b .This is the
required point P .
If we reflect the point B (2,4) on L : y x , we get B' (4,2) . Next find the
equation of the line through A (1,5) and B' (4,2) .It is found to be m : y x 6
. Finally, determine the intersection of this line with the given line. That is
L : y x, m : y x 6 x x 6 2 x 6 x 3, y 3 .
Therefore, the required point is P (3,3) . (If you use Calculus, you will get the
same result)
Prepared by Begashaw M. 92
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 3.3
1. Find the image of a circle C : x 2 y 2 2 x 6 y 5 by a reflection on the line
l : y x 2. Answer : ( x 1) 2 ( y 1) 2 15
2. Let l : x a and m : y b be two lines. If S m S l (2,3) (4,7) , find the
values of a and b . Answer : a 3, b 5
3. Find the image of
a) the circle C : x 2 y 2 1 by a reflection on the line y x 7 .
b) the ellipse E : 9 x 2 4 y 2 36 by a reflection on the line l : y x 3
4 3 3 4
4. The equation of a reflection is given by S l ( x, y) x y, x y .
5 5 5 5
Find the equation of the line of reflection. Answer :l : y 3x
5. Suppose an ant moves from the position A (0,2) to the position B (6,1) by
touching the x-axis at some point P . Find the coordinate of P that minimizes
the total path of the ant. Answer : (1 / 4,0)
6. If the image of ABC by a reflection on the line l : y x 1 is DEF where
the vertices of the image are D (1,1), E (5,1), F (1,2) .
Find the vertices of ABC .
Answer : A (0,0), B (0,4), C (3,0)
7. If a line y mx b is reflected along the line y mx c , show that its image
is y mx 2c b .
8. Suppose l is a line through the origin which makes an angle of from the
positive x-axis with tan 3 . Find the image of the point (7,1) by a reflection
on this line. Answer : (5,5)
9. Let S l be a reflection on a line l in R 2 . Prove that S l (v) 2(u.v)u v for all
vectors v in R 2 where u is a unit vector along l .
Prepared by Begashaw M. 93
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
10. Let l be a line along the vector u and let Pu (X ) be the projection of X on l .
7 8 7 7 9
If Pu , find the reflection image S l on l . Answer : S l
3 6 3 3 9
11. What is the minimum length of a flat-against the wall, full- length mirror
for the Smiths who ranges in height from 170cm to 182cm , if you assume eyes
are 10cm below the top of the head. Answer : 72cm
12. Prove that every reflection is its own inverse.
13. Suppose m and n are two distinct lines. Then, show that
S n S m ( P) P P m n .
14. Can the product of two reflections ever be a reflection? Explain your
answer!
Prepared by Begashaw M. 94
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3.4.3 Rotation
Definition: A rotation is a transformation in which a figure is turned about a
fixed point through an angle of in a specific direction. In other words,
rotation about a point C through directed angle is a transformation that fixes
the point C and sends every other point P to P ' such that P and P ' have the
same distance from the fixed point C . Here, the fixed point C is called the
center of rotation and the angle measured from CP to CP ' is called
direction of the rotation. The rotation may happen either clockwise or counter
clockwise direction, usually clockwise rotation will have negative measure of
angle, whereas counter clockwise rotation will have positive measure of angle.
Rotation with center C through an angle of is usually denoted by C , .
So, the image of any point P under C , is given as:
C , if P C
C , ( P)
P ', if P C , s.t. CP CP '
Theorem 3.3: A rotation through an angle of , about the origin which takes
each point P( x, y) in to P '( x ', y ') is given by O, ( x, y) ( x ' , y' ) , where
x ' x cos y sin
y ' x sin y cos
Proof: Let o , be a rotation through an angle of , about the origin O and let
P( x, y) be any point such that O, ( x, y) ( x ' , y' ) . As shown in figure 3.5,
suppose is the angle from the positive x-axis to the segment OP .
Prepared by Begashaw M. 95
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Then the angle from the positive x-axis to the segment OP ' will be . Let
r OP . From the definition of rotation OP OP ' , so we have r OP ' .
Consequently, using polar coordinates for x ' and y ' , we get
x ' r cos( ) y ' r sin( )
r cos cos r sin sin , r sin cos r cos sin
x cos y sin x sin y cos
Examples:
1. Find the image of the point (1,1) and the line l : 3x 5 y 7 by a a counter
clockwise rotation through 90 about the origin.
Solution: From the above theorem, for any point P( x, y) , O, ( x, y) ( x ' , y' )
where x ' x cos y sin and y ' x sin y cos . So, using P(1,1) ,
90 , and C (0,0) , we get O, ( x, y) ( x ' , y' ) (1,1) .
To find the image of the line l , take arbitrary point ( x, y) on l .
x ' x cos 90 y sin 90 y
So O, ( x, y) ( x , y' ) such that
'
y ' x sin 90 y cos 90 x
Now, solving these equations for x and y and substituting in the equation of l ,
we get l ' : 3 y ' 5( x ') 7 5x ' 3 y ' 7 .
2. Find the image of the point ( 3,1) by a counterclockwise rotation about the
origin with 105 .
Solution: Here, we use angle sum formula to determine cos(105 ), sin(105 ) .
That is
cos(105 ) cos(60 45 ) cos 60 cos 45 sin 60 sin 45
1 2 3 2 2 6
. .
2 2 2 2 4 4
sin(105 ) sin(60 45 ) sin 60 cos 45 cos 60 sin 45
3 2 1 2 6 2
. .
2 2 2 2 4 4
Prepared by Begashaw M. 96
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2 6 6 2 4 2
x' 3 cos(105 ) sin(105 ) 3.( )( ) 2
4 4 4 4 4
Thus,
y ' 3 sin(105 ) cos(105 ) 3. ( 6 2
)(
2
6
)
4 2
2
4 4 4 4 4
Therefore, ( x' , y' ) ( 2 , 2 )
2
3. Suppose is a counterclockwise rotation about the origin with tan .
3
Find the image of (5 13,2 13 ) under .
2 2 3
Solution: Here, using right angle triangle, tan sin , cos .
3 13 13
3 2
x' x cos y sin x' 5 13. 13 2 13. 13 15 4 11
( x' , y ' ) (11,16)
y ' x sin y cos y ' 5 13. 2 2 13. 3 10 6 16
13 13
Rotating a line and a circle : To find the image of a line or a circle under a
given rotation we use the same procedure as we did in reflection section. For
instance to find the image of a circle under a given rotation, first identify the
center and radius of the given circle, second find the image of the center under
the given rotation, finally write the circle using the image center and the same
radius as the given circle. (Please refer the procedures we use for reflecting a
line and a circle)
Theorem 3.4 (Generalized Rotation Theorem):
The image of any point P( x, y) under a rotation about arbitrary center C h, k
through an angle of is given by O, ( x, y) ( x ' , y' ) where
x' ( x h) cos ( y k ) sin h
y' ( x h) sin ( y k ) cos k
Outline of the proof: The proof of this theorem is not difficult. It is simply
accomplished by performing the following coordinate transformations step by
step.
Prepared by Begashaw M. 97
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Step-1: Translate the center C h, k to the origin by a vector v (h, k ) so
that any point P( x, y) is translated to the point ( x h, y k ) .
Step-2: Rotate the result obtained in step one about the origin by an angle .
Step-3: Translate the result obtained in step two by a vector v (h, k ) to take
every point back to the original position. This will complete the proof.
The equations of the generalized theorem are too complicated to memorize.
Subsequently, the readers of this material are advised to remember the methods
in the outline of the proof how the formulas are obtained (Translate- Rotate-
Translate) rather than memorizing the formulas.
Example: The image of the point (1,2) by a counter clockwise rotation about
the center C (2,3) is (2,3 2 ). Find the angle of rotation.
Solution: By the generalized rotation theorem,
x' ( x h) cos ( y k ) sin h
y' ( x h) sin ( y k ) cos k
So, for ( x, y) (1,2) , C (h, k ) (2,3) and ( x' , y' ) (2,3 2 ) , we get
x' cos sin 2 2 x' cos sin 0
y ' sin cos 3 3 2 y ' sin cos 2
cos sin 2
sin cos
sin cos 2 2
Here, both sin and cos are positive.
But this is true if and only if is in the first quadrant.
2
Thus, an angle in the first quadrant with sin cos is .
2 4
Theorem 3.5 (Formula for center of a rotation):
Prepared by Begashaw M. 98
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Suppose the general equation of a rotation about any center C (h, k ) with
angle of rotation is given by C , ( x, y) ( x' , y' ) where
x' x cos y sin r
with r ,t , being real numbers.
y ' x sin y cos t
Then, the center C (h, k ) of this rotation is given
r t
h
2 2 tan 2
k t r
2 2 tan 2
x' ( x h) cos ( y k ) sin h
Proof: From the general equation,
y' ( x h) sin ( y k ) cos k
h(1 cos ) k sin r
we have
h sin k (1 cos ) t
Solve these equations for h and k whenever r ,t , are given real numbers.
This can be solved using Cramer’s rules with h and k as variables.
1 cos sin
1 2 cos cos 2 sin 2 2(1 cos ) 4 sin 2
sin 1 cos 2
r sin
h r (1 cos ) t sin
t 1 cos
1 cos r
k t (1 cos ) r sin
sin t
h r (1 cos ) t sin r sin r 2 sin 2 cos 2 r t
h t t
2(1 cos ) 2 2(1 cos ) 2 4 sin 22
2 2 tan 2
k t (1 cos ) r sin t sin t 2 sin 2 cos 2 t r
k r r
2(1 cos ) 2 2(1 cos ) 2 4 sin 22
2 2 tan 2
Example: Suppose RC , is a counterclockwise rotation with center C (h, k )
1 3
x' x y24 3
2 2
whose equations are given by . Find the angle and
y' 3 1
x y42 3
2 2
center of this rotation.
Prepared by Begashaw M. 99
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Solution: Let’s use the formula for center of rotation given in the above
theorem.
1 3
Here, r 2 4 3, t 4 2 3 . Besides, cos , sin 60 .
2 2
r t 24 3 42 3
h
1 2 3 3 (2 3 ) 4
2 2 2 tan 30
2 tan
2
Hence,
k t r 42 3 24 3
2 3 3 (1 2 3 ) 8
2 2 2 tan 30
2 tan
2
Therefore, the center of the rotation is C (h, k ) (4,8) .
Theorem 3.6: Let R be a counter clock wise rotation by a given angle about
the origin. Then,
a) R R R , for any two angles
1
b) R R , the inverse of a rotation by is a rotation by -
c) R i 2n , n Z (Where i is identity rotation)
Proof: a) Let P( x, y) be any point. Then,
R R ( x, y ) R ( R ( x, y ))
R ( x cos y sin , x sin y cos ) ( x ', y '), where
x ' ( x cos y sin ) cos ( x sin y cos ) sin
y ' ( x cos y sin ) sin ( x sin y cos ) sin
Rearranging these equations and using angle sum theorem, we get
x ' x cos( ) y sin( )
...................................................(i)
y ' x sin( ) y cos( )
On the other hand, using as angle of rotation, we get
R ( x, y) ( x cos( ) y sin( ), x sin( ) y cos( )) ( x ', y ')...................(ii )
Comparing equations (i) and (ii) , one can conclude that R R R .
The proof of part (b) and (c) is left as an exercise.
Prepared by Begashaw M. 100
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
The above theorem reveals that the composition of two rotations R and R
about the point C is the same as a single rotation by an angle of about the
same point C . Besides, the order of the rotation is immaterial and we write this
relation as Rc, Rc, Rc, Rc, Rc, . On the other hand part (b) of the
theorem tells us that the inverse of a rotation about center C by an angle of is
the same as a rotation about the same center but by an angle of and this is
written as Rc,1 Rc, . This means that the inverse of a counter clock wise
rotation is the same as a rotation in clock wise direction by the same angle
through the same center. Besides, part (c)
shows that a rotation is identity rotation if and only if the angle of rotation is a
multiple of 2 , that is 2n , n Z .
Examples:
1. Find the image of the point (2,5) by a product of rotations through an angle
of 15 and 75 in counterclockwise direction about the same center
C (7,2).
Solution: From the above theorem, C , C , ( x, y) C , ( x, y) where
( x, y) (2,5) , 15 , 75 and C (7,2).
x' (2 7) cos 90 (5 2) sin 90 7 4
Thus, C ,15 C ,75 (2,5) C ,90 (2,5)
y ' (2 7) sin 90 (5 2) cos 90 2 3
2. Suppose R is a counterclockwise rotation about the origin whose equations
3 1
x' x y
2 2 . Find the equations for the inverse, 1 , of this
are given by R
y' 1
x
3
y
2 2
rotation.
3 1
Solution: Here, cos , sin 30 .
2 2
These, by part (b) of the above theorem, we have
Prepared by Begashaw M. 101
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3 1
x' x cos(30 ) y sin(30 ) x y
2 2
R 1 ( x, y ) ( x' , y ' ) where
y ' x sin(30 ) y cos(30 ) 1 x 3 y
2 2
3. If R R , then what should be the possible values of ? Particularly for
0 2 .
Solution: Using part (c) of the above theorem, we have
R R R ( P) R ( P) R R ( P) R R ( P)
R2 ( P) R0 ( P) i( P) P R2 i
2 2n, n Z n, n Z
Particularly for 0 2 , , when n 1.
4. Suppose R and R are counterclockwise rotations about the origin such that
R R ( x, y) ( y, x) and 35 . Find angle .
Solution: Here, using part (a) of the above theorem, we have
x' x cos( 35 ) y sin( 35 ) y
R R ( x, y ) ( y, x)
y ' x sin( 35 ) y ( 35 ) x
cos( 35 ) 0
35 90 55
sin( 35 ) 1
Half-turns
Definition: A half turn is a rotation by 180 . A half turn about a point P is
denoted by H P . If A is rotated by 180 about point P(a, b) , then A' P AP . In
other words P is the mid point of A and A' .
x' x
a
A' A
Thus using the midpoint formula, P 2 x' x 2a, y ' y 2b .
2 y ' y b
2
Prepared by Begashaw M. 102
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
This will lead us to the formal definition of a half turn. Let P (a, b) . Then,
half turn H P about a point P is a transformation whose equation is given by
x ' x 2a
H P ( x, y) ( x' , y' ) , where
y ' y 2b
Examples:
1. Find the image of a point (2,7) by a half-turn about the point P (5,3) .
Solution: By definition,
x' x 10
H P ( x, y) ( x' , y' ) where . Therefore, H P (2,7) (8,1).
y' y 6
2.If the image of (2,3) by a half-turn is (10,11) , find the center of the half-turn.
Solution: Let the center be P (a, b) . Then, using the definition, we have
H P (2,3) (10,11) (2a 2,2b 3) (10,11)
2a 2 10, 2b 3 11
a 4, b 7 P (4,7)
3. Let H P be a half turn about P (3,2). Find,
a) The image of the line l : y 5x 7
b) The pre-image of the line m : y 2 x 17
Solution:
a) Let ( x, y) be any point on the given line. Then,
( x' , y' ) H P ( x, y) ( x 6, y 4) . Solving for x and y from this equation
gives us x x'6, y y'4 . So, substitute these values in the equation of the
line so that the equation will be in terms of x' , y' and that will be the image
line. l : y 5x 7 l ': y'4 5( x'6) 7 l ': y' 5x'27
Therefore, H P (l ) l ': y 5x 27 . Here, the line l ': y 5x 27 is called the image
of l under H P and the line l : y 5x 7 is called the pre-image of l ' .
Prepared by Begashaw M. 103
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
b) In this case the image of some line m is given we are required to find the pre
image m of this line. To do this problem we can use different methods. For
instance first calculate the inverse H P1 of H P and then find the image of the
given line under H P1 . That will be the pre-image of the given line. Since half
turn is an involution, every half turn is its own inverse.
So, H P1 ( x, y) ( x 6, y 4) ( x' , y' ) .
Using this equation we get the pre-image of the given line to be m : y 2 x 1 .
Proposition 3.12(Characterization Theorem of a half turn):
a) A Half turn H P fixes a line l if and only if P l .
b) A Half turn H P fixes a point A if and only if A P .
c) H P (Q) R if and only if P is the mid point of Q and R .
QR
That is H P (Q) R P .
2
d) Half turn is an involution. That is H 2 P H P H P i .
Proof:
a) Let l : ax by c 0 be any line and let P (h, k ) . Let ( x, y) be arbitrary point
on line l . Then, H P ( x, y) ( x' , y' ) ( x 2h, y 2k ) . Thus, solving for x and
y and substituting in the equation of l , we get,
ax by c 0 ax'by'c 2(ah bk c) 0.
But, the last equation defines equation of a line. Thus, the image of the line l
under H P is also a line l ' H P (l ) .
Hence, H P fixes a line l if and only if l l ' . But from equation of l and l ' ,
they will be the same line if and only if
2(ah bk c) 0 ah bk c 0 (h, k ) l P l.
Thus, H P fixes a line l if and only if P l .
b) Let A ( x, y) and P (h, k ) such that H P ( A) A . We need to show A P .
But,
H P ( A) A ( x 2h, y 2k ) ( x, y) x h, y k ( x, y) (h, k ) A P
Proposition 3.13: The composition of two half turns H P and H Q is a translation
by a vector 2 PQ in the direction from P to Q .That is H Q H P T2 PQ
Proof: Let A ( x, y) be any object point. Then,
Prepared by Begashaw M. 104
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
( H Q H P )( A) H Q ( H P ( A)) H Q ( A 2 P)
(( A 2 P) 2Q) ( A 2 P 2Q)
( A 2(Q P)) A 2 PQ
T2 PQ ( A) T 2 P ,Q T2 PQ
The composition of two half turns H P and H Q is a translation by a vector 2 PQ in
the direction from P to Q .That is H Q H P T2 PQ
Proposition 3.14: Let P , Q and R be any three points. Then, Q is the
midpoint of the P and R if and only if H R H Q TP, R H Q H P .
Proof: () Suppose Q is the midpoint of P and R . In the above theorem, we
have proved that the composition of any two half turns is a translation. Thus,
both H R H Q and H Q H P are translations. Besides, by proposition 3.12, since
Q is the midpoint of P and R , H R H Q ( P) H R ( R) R and
H Q H P ( P) H Q ( P) R . Here, both H R H Q and H Q H P takes P to R .
But, from the previous discussion of translation (in proposition 3.9), there is a
unique translation TP , R that takes P to R . Therefore, H R H Q TP, R H Q H P .
Conversely, suppose H R H Q TP, R H Q H P . We need to show Q 12 ( P R) .
But, for any point X ,
( H R H Q )( X ) H Q H P ( X )
H R ( X 2Q) H Q ( X 2 P)
( X 2Q) 2 R ( X 2 P) 2Q
X 2Q 2 R X 2 P 2Q
4Q 2 P 2 R
Q 12 ( P R)
Here one can observe that not only the composition of two half turns is a
translation but also any translation is the composition of two half turns and this
is left for the reader to justify.
Corollary 3.2: Let Q be the midpoint of the points P and R . If S is the
midpoint of the points P and Q , then H R H Q T4 PS H Q H P
Proposition 3.15: If A C, then the following conditions are equivalent.
a) B is the midpoint of A and C
Prepared by Begashaw M. 105
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
b) H B H A H C H B
c) H B H A H B H C
Proof: (a) (b). Suppose B is the mid point of A and C .Then,
2 AB AC 2BC T2 AB T2 BC H B H A H C H B ( From proposition 3.14)
(b) (c). Suppose H C H B H B H A . Then,
H B H A HC H B H B H A H B HC H B H B
H B H A H B H C i, H B H B i
H B H A H B HC
(c) (a). Suppose H B H A H B H C . Then,
H B H A H B HC H B H A HC H B
T2 AB T2 BC
2 AB 2 BC
AB BC
AC AB BC 2 AB 2 BC
Therefore, B is the midpoint of A and C . From this, we can infer that for any
three collinear points A, B, C , B is the midpoint of A and C if and only if
H B H A H C H B if and only if T2 AB T2 BC .
Example: If H A and H B are half-turns such that H B H A H C H B where
A (1,3), B (1,5) , find the coordinates of C .
Solution: By proposition 3.15, we have that
H B H A H C H B B 12 ( A C ) C 2B A (2,10) (1,3) (3,7)
Proposition 3.16: The composition of three half turns is a half turn.
Proof: Let H P , H Q and H R be three half turns about the points
P (a, b) , Q (c, d ) and R (e, f ) respectively. Then, for any object X ( x, y) ,
H R H Q H P ( x, y) ( x 2[a e c], y 2[b f d ]) . But this is an equation
of a half turn with center at the point (a e c, b f d ) .
Prepared by Begashaw M. 106
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Proposition 3.17: If points P, Q, R and T are non-collinear points, then
H R H Q H P H T whenever PQRT is a parallelogram.
One of the applications of this corollary is that in any parallelogram PQRT ,
from the knowledge of the three vertices of the parallelogram PQRT , any one
of the fourth vertex can be determined from the relation H R H Q H P H T .
Example: Let PQRT be a parallelogram with vertices P(1,2), Q(6,2), T (1,4) .
Find the vertex R .
Solution: From the above corollary, PQRT is a parallelogram if and only if
H R H Q H P H T . Let R (a, b) and X ( x, y) be any point.
Then,
H R HQ H P HT
H R HQ H P (X ) HT (X )
H R H Q ( x 2, y 4) ( x 2, y 8)
H R ( x 10, y ) ( x 2, y 8)
( x 10 2a, y 2b) ( x 2, y 8) (2a 10,2b) (2,8)
(2a,2b) (12,8) (a, b) (6,4)
Thus, the unknown vertex is R (a, b) (6,4) .
Problem Set 3.4
1. If a rotation takes the vector 10 to a , then find the value(s) of a .
5 11
Answer : a 2
Prepared by Begashaw M. 107
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2. *Suppose an X 'Y ' -coordinate axis is obtained by rotating an XY
coordinate axis CCW through an angle of 45 about the origin. Find the
equation of
a) the curve 3x'2 y'2 6 in the XY coordinate axis
b) the curve 2 x 2 2 y 2 2 xy 3 in the X 'Y ' coordinate axis.
Answer : a) x 2 xy y 2 3 b) x' 2 3 y' 2 3
2
3. Suppose is a counterclockwise rotation about the origin with tan .
3
Find the image of ( 13, 13 ) under . Answer : (1,5)
4. Find the image of the line l : y 5x 7 under a half turn H P with center
P(3,2) . Answer : y 2 x 27
5. If a half turn H P ( x, y) (4 x,12 y) fixes the line l : y mx , find the slope m
and give the equation of the line l . Answer : l : y 3x
6. If A (1,2), B (3,6), find C such that H B H A (C ) (4,2).
Answer : C (12,10)
7. If E (2,3), then find D such that H E H D (1,2) (3,5).
8. For any point ( x, y) , if H A H B ( x, y) ( x 2, y 6) , find the vector AB .
Answer : (1,3)
9. Let P (2,3), Q (1,5), R (0,2) . Find a point T so that the product of the
half turns H P H Q H R is equal to a single half turn about T .
Answer : T (3,4)
10. Let P and Q be any two points. Then, show that
a) TPQ H Q H P P Q b) H P H Q H Q H P P Q .
11. Show that rotation preserves parallelism of lines.
12. Show that a translation can be written as a product of two rotations.
13. Prove that if Q TA, B ( P) , then TA, B H P TA1,B H Q .
14. Show that for any two points P, Q , TPQ H Q H P P Q.
Prepared by Begashaw M. 108
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
15. Suppose H P and H Q are half turns about the points P and Q respectively.
If PQ (2,3) , find H P H Q (2,4) . Answer :(6,2)
16. Suppose P, Q, R and T are non-collinear points.
a) Prove that H R H Q H P H T whenever PQRT is a parallelogram.
b) If P(1,2), Q(6,2), T (1,4) , find the vertex R such that PQRT forms a
parallelogram.
3.3.4 Glide Reflection
Definition: A glide reflection g is the composition of a reflection S l over a
line l followed by a translation Tv with non-zero vector v where the line l is
parallel to the direction of the translation or parallel to the translator vector v .
The vector v in this case is called glide vector and the line l is called axis of the
Prepared by Begashaw M. 109
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
glide reflection. Here, the vector v is required to be non-zero otherwise
translation by a zero vector will be identity map and the composition also will
be the usual reflection but not glide reflection. We can easily justify that the
same result is obtained by first reflecting and then translating or vice versa. As
a result, the order of the two transformations
(Translation and reflection) is immaterial. So, g Tv S l S l Tv .
Example : Let S l be a reflection on the line y x 1 and Tv be a translation by
v (1,1) . Here, S l ( x, y) ( y 1, x 1) and Tv ( x, y) ( x 1, y 1) for any point
( x, y) . So, g ( x, y) Tv S l ( x, y) ( y, x 2) and g ( x, y) S l Tv ( x, y) ( y, x 2)
which verifies that g Tv S l S l Tv .
General Equations of Glide-Reflections:
Let g be a glide reflection with axis l : ax by c 0 and glide vector v (d, e)
with the condition a.d b.e 0 .Then, the general equation of g is given by
2a(ax by c)
x ' x d
a2 b2
g ( x, y) Tv Sl ( x, y) Sl ( x, y) v ( x' , y' ) where
y ' y 2b(ax by c) e
a2 b2
Conversely, if g is a glide reflection given by g ( x, y) ( x' , y' ) where
x' ax by c
y ' bx ay d
Then, the axis of g is given by l : 2bx 2(a 1) y ad bc d 0 or 2 x c .
Examples:
1. Let g be a glide reflection with axis l : 3x 4 y 2 0 and glide vector
v (4,3) . Find the equation of g and calculate the image of the point (0,0)
Solution: Here, g ( x, y) Tv Sl ( x, y) Sl ( x, y) v ( x' , y' ) where
Prepared by Begashaw M. 110
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2a(ax by c) 6(3x 4 y 2) 7 x 24 y 88
x' x d x 4
a b
2 2
9 16 25 25 25
y ' y 2b(ax by c) e y 8(3x 4 y 2) 3 24 x 32 y 91
a2 b2 9 16 25 25 25
Therefore the image of the point (0,0) is given by
7(0) 24(0) 88 88 24(0) 32(0) 91 91
g (0,0) ( x' , y ' ) where x' , y'
25 25 25 25 25 25 25 25
88 91
Hence, g (0,0) ( , ).
25 25
2. Suppose g is a glide reflection with axis l : 5x ky 7 0 and glide vector
v (6,10) . Then, find the value of the constant k .
Solution: The condition for g to be a glide reflection with axis
l : ax by c 0 and glide vector v (d, e) is that a.d b.e 0 . In our case, g to
be a glide reflection with axis l : 5x ky 7 0 and glide vector v (6,10) , we
have 5.(6) k (10) 0 30 10k 0 k 3
3. Let g be a glide reflection with axis l : 2 x 7 y 9 0 and glide vector
v (d ,4) . Then, find the value of the constant d .
Solution: Similarly, as in example (2), 2d 28 0 2d 28 d 14.
4. Let g be a glide reflection with axis l : 2 x y 5 0 and glide vector
v (d , e) . If g (0,0) (6,2) , then find the glide vector v .
Solution: Here, the general equations of g are given by
2a(ax by c) 4(2 x y 5) 3 4
x' x d x d x y4d
a b
2 2
5 5 5
y' y 2b ( ax by c ) 2( 2 x y 5) 4 3
e y e x y2e
a2 b2 5 5 5
3 4 4 3
That is for any point ( x, y) , g ( x, y) x y 4 d , x y 2 e
5 5 5 5
Particularly, g (0,0) (6,2) (4 d , 2 e) (6,2) 4 d 6, 2 e 2
d 2, e 4 v (2,4)
Prepared by Begashaw M. 111
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5. Let g be a glide reflection with axis l : x y 1 0 and glide vector v (3,3) .
If g ( p, q) (5,8) , then find the point ( p, q) .
Solution: Using the general equations of glide- reflection, we have
2( p q 1)
x ' p 3 q2 5 q 3
2
( p, q) (3,4)
y' q 2( p q 1)
3 p4 8 p 4
2
Proposition 3.19: The square of a glide reflection (composition of a glide
reflection with itself) is a translation. That is g 2 ( P) ( g g )( P) T ( P) for some
translation T . Any glide reflection fixes exactly one line, its axis. The midpoint
of any point P and its image always lies on the axis of the glide reflection.
Proof: We will prove here the first part of the theorem the others follow from
the above discussions. Let g be a glide reflection. Then, by definition of glide
reflection, g S l Tv Tv S l where S l is a reflection and Tv is a translation
with vector v parallel to the line of reflection l .
Thus,
g 2 ( X ) g g( X )
( S l Tv ) ( S l Tv )( X )
(Tv S l ) ( S l Tv )( X )
Tv ( S l S l ) Tv ( X )
Tv i Tv ( X ), S l S l i
Tv Tv ( X )
Tv ( X )
2
T2 v ( X )
g 2 ( X ) g g ( X ) T2 v ( X ), X
g 2 T2 v
Hence, composition of a glide reflection with itself is a translation by twice its
original glide vector. We know that in any reflection S l , the mid point of any
point P and its image S l (P) lies on the line of reflection or axis of reflection.
Prepared by Begashaw M. 112
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Thus, for any glide reflection g , P and its image g (P) always lies on the axis
of the glide reflection, too.
Examples:
1. Suppose g is a glide reflection such that g 2 (3,5) (7,1) .Then, find the glide
vector of g .
Solution: Let the glide vector be v . Then, by the above proposition, g 2 T2v .
So, for any point P , we have
g 2 ( P) T2v ( P) 2v P g 2 (3,5) (7,1)
2v (3,5) (7,1)
2v (7,1) (3,5) v (2,3)
2. Suppose g is a glide reflection such that g (1,2) (3,4) and g (1,3) (5,7).
Then, find the axis of g .
Solution: For any glide reflection g , the mid point of any point P and its
image always lies on the axis of g . In particular, the mid point of (1,2) and
(3,4) , the midpoint of (1,3) and (5,7) lie on the axis of the glide reflection g .
Hence, the axis of g passes through the points (1,3) and (2,5) . Let the axis of
53 2
g be the line given by l : y mx b . Then, the slope is m .
2 1 3
2
So, l : y xb.
3
2 2 11
Taking one of the above points say (1,3) , y xb b 3b .
3 3 3
2 11
Hence, the axis of the glide reflection is l : y x 2 x 3 y 11 0 .
3 3
Proposition 3.20: The product (composition) of two glide reflections about the
same axis is a translation with translation vector of the sum of the two glide
vectors. That is if g and h are glide reflections with the same axis l and glide
vectors v and w respectively, then g h Tv w .
Prepared by Begashaw M. 113
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Proof: Suppose g and h are glide reflections with the same axis l and glide
vectors v and w respectively. That is g Sl Tv , h Sl Tw . We need to show
g h is a translation. Here, using associativiy of composition and the fact that
g Sl Tv Tv Sl , h Sl Tw Tw Sl , for any point P , we have
( g h)( P) g (h( P)) ( S l Tv ) ( S l Tw ( P))
(Tv S l ) ( S l Tw ( P)
Tv ( S l S l ) Tw ( P)
Tv i Tw ( P)
(Tv Tw )( P) v w P Tv w ( P) Tu ( P), u v w g h Tu
Example: Suppose g and h are glide reflections with the same axis such that
h 2 ( x, y) ( x 6, y 2) and ( g h)(3,7) (9,6) .Then, determine the glide
vectors of g and h .
Solution: Let the glide vector of g be v and that of h be w .Then, using
proposition 3.19, we have
h 2 ( P) T2 w ( P) h 2 ( x, y ) T2 w ( x, y )
2w ( x, y ) ( x 6, y 2)
2w (6,2) w (3,1)
A gain, using proposition 3.20, we have
g h Tv w ( g h)( P) Tv w ( P) v w P, for any point P
( g h)(3,7) (9,6), particularly for P (3,7)
v w (3,7) (9,6)
v (3,1) (3,7) (9,6)
v (9,6) (0,6) (9,12)
Problem Set 3.5
1. Let g be a glide reflection with axis l : x y 3 0 and glide vector v (4,4)
Then, find the general equation of g and calculate the image of the point (0,0)
Prepared by Begashaw M. 114
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Answer : g ( x, y) ( y 1, x 7), g (0,0) (1,7)
2. Suppose g is a glide reflection with fixed (invariant) line y x 1 . If it
maps the point (0,1) to (3,4) , find its equation. Answer : g ( x, y) ( y 2, x 4)
3. Let g be a glide reflection with axis l : x y 7 0 . If g 2 (13,14) (3,4) ,
then find its equation. Answer : g ( x, y) (2 y,12 x)
4. Let g S l Tv be a glide reflection where S l is a reflection on line l passing
through the point (2,7) . If g 2 (0,0) (4,12) , find the equation of the line l .
Answer : y 3x 1
5. For what value of the constant k will an isometry g be a glide reflection with
axis l : 3x ky 17 0 and glide vector v (6,2) ? Answer : k 9
6. Let g be a glide reflection with axis l : 2 x 7 y 9 0 and glide vector
v (14, e) . Then, what must be the value of the constant e ? Answer : e 4
7. Let g be a glide reflection with axis l : 2 x y 15 0 and glide vector v . If
g (3,4) (20,0) , then find the glide vector v . Answer :v (3,6)
8. Let g be a glide reflection with axis l : y x 1 and glide vector v (1,1)
Then, find the image of the point (3,2) . Answer : (2,5)
9. Let g be a glide reflection with axis l : x y 6 0 and glide vector v (2,2)
Then, find the general equation of g and calculate the image of the point (0,0)
Answer : g ( x, y) ( y 4, x 8)
10. Find the equation of a translation T of 2 2 units at 45 into the first
quadrant and a glide reflection g of 2 2 units at 45 with axis l : x y 4 .
Find the image of A BC with A (0,0), B (1,0), C (0,2) under each
isometgry.
Prepared by Begashaw M. 115
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Answer :T ( x, y) ( x 2, y 2), g ( x, y) ( y 6, x 2)
3 0
11. Consider the line l : 2 x 3 y 2 0 and the points P , Q .
1 1
Then, give the equations of a glied reflection g Sl TP,Q .
5 12 31 12 5 38
Answer : g ( x, y) ( x y , x y )
13 13 13 13 13 13
12. Let g be a glide reflection with axis l : ax by c 0 and glide vector
b d
v (d , e) .Then, show that .
a e
3. 5 Orientation and Orthogonal Transformations
3.5.1 Orientation of Vectors
x z
Consider a pair of vectors X and Y regarded as order pair ( X , Y ) .
y w
Denote the angle measured from X to Y in the counter clockwise direction by
where 0, as shown in the diagram below.
Prepared by Begashaw M. 116
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Then, with the help of these diagrams we will give the following definitions.
Definition:
a) The pair of vectors ( X , Y ) is said to be positively oriented if and only if
sin 0 . In this case, we say that the vectors X and Y have positive
orientation. The first diagram above shows how positively oriented vectors are
placed.
b) The pair of vectors ( X , Y ) is said to be negatively oriented if and only if
sin 0 . In this case, we say that the vectors X and Y have negative
orientation. The second diagram shows how negatively oriented vectors are
placed.
x z
Now, if we are given any two vectors X and Y , how can we
y w
determine whether they are positively oriented or negatively oriented simply by
using their coordinates x, y, z, w ?
The method how can we determine whether a pair of vectors is positively
oriented or negatively oriented from their coordinates is given below.
Theorem 3.7 (Orientation Test) :
x z
Let X and Y be any two non-zero and non- parallel vectors. Then,
y w
x z
a) X and Y are positively oriented if and only if det( X , Y ) 0
y w
Prepared by Begashaw M. 117
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x z
b) X and Y are negatively oriented if and only if det( X , Y ) 0
y w
This theorem is known as Orientation Test Theorem.
Proof: Consider the following diagram (figure 3.6)
From these values,
sin sin( )
sin cos cos sin
w x z y
z 2 w2 x2 y2 z 2 w2 x2 y2
xw yz
x 2 y 2 z 2 w2
Now, from the above definition the pair ( X , Y ) is positively oriented if and only
if sin 0 . But,
xw yz
sin 0 0
x y 2 z 2 w2
2
xw yz 0; x 2 y 2 . z 2 w 2 0
x z
0
y w
On the other hand, the pair ( X , Y ) is negatively oriented if and only if sin 0 .
But,
Prepared by Begashaw M. 118
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
xw yz
sin 0 0
x 2 y 2 z 2 w2
xw yz 0; x 2 y 2 . z 2 w 2 0
x z
0
y w
Hence, the proof is complete.
x z
In general, the vectors X and Y are
y w
x z
a) Positively oriented if and only if sin 0 det( X , Y ) 0.
y w
x z
b) Negatively oriented if and only if sin 0 det( X , Y ) 0.
y w
Examples: Determine whether the following pair of vectors are positively or
negatively oriented.
1 0 2 5
a) X and Y b) X and Y
0 3 3 1
Solution:
1 0 1 0
a) For X and Y , det( X , Y ) 3 0 . So, the pair ( X , Y ) is
0 3 0 3
positively oriented. That means X and Y have positive orientation.
2 5 2 5
b) For X and Y , det( X , Y ) 17 0 . So, the pair ( X , Y ) is
3 1 3 1
negatively oriented. That means X and Y have negative orientation.
Remarks:
i. Orientation is not defined for parallel vectors. Because for parallel vectors, if
x z tx
X and Y , then Y tX for some scalar t (definition of parallel
y w ty
x tx
vectors). Thus, det( X , Y ) 0 . But zero is neither positive nor negative.
y ty
In this case, we say that X and Y have zero orientation.
Prepared by Begashaw M. 119
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
ii. Orientation is not defined for three collinear points. For any three collinear
points A, B, C , their orientation is determined from the orientation of the
vectors AB, AC . But if the points A, B, C are collinear, then the vectors
AB, AC will be parallel and hence from the first remark their orientation is not
defined.
iii . A pair of vectors ( X , Y ) is assumed to have the same orientation with a pair
of vectors (Z ,W ) if and only if det( X , Y ) and det(Z ,W ) have the same sign.
If these determinants det( X , Y ) and det(Z ,W ) have opposite sign, then we say
that the pairs ( X , Y ) and (Z ,W ) have opposite orientation.
Examples:
1 2 0 3
1. Let X , Y , Z , W . Determine whether (Y ,W ) and
2 3 2 5
( Z , X ) have the same or opposite orientation.
2 3 0 1
Solution: Here, det(Y ,W ) 1 0, det(Z , X ) 2 0 .
3 5 2 2
So, (Y ,W ) and (Z , X ) have the same orientation. But if we take ( X , Y ) and
1 2 0 3
( Z ,W ) , det( X , Y ) 1 0, det(Z ,W ) 6 0.
2 3 2 5
Thus, ( X , Y ) and (Z ,W ) have the opposite orientation.
2. Let X , Y be any two vectors and t 0 be a scalar. Then,
a) ( X , Y ) and (tX , Y ) have the same orientation.
b) ( X , Y ) and (Y , X ) have opposite orientation.
x z tx z x z
Solution: Let X , Y and t 0 . Then, det(tX , Y ) t .
y
w ty w y w
x z
Since t 0 , the sign of det(tX , Y ) depends on the sign of det( X , Y ) .
y w
Prepared by Begashaw M. 120
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Hence, ( X , Y ) and (tX , Y ) have the same orientation.
x z z x
det( X , Y ) and det(Y , X ) . But, from properties of determinants,
y w w y
z z x z
det(Y , X ) (1) det( X , Y ).
w w y w
Thus, ( X , Y ) and (Y , X ) have opposite orientation.
3.5.2 Orientation of Plane Figures
Here, we will see how to determine the orientation of a triangle and then the
orientation of other plane figures can be defined in the same way.
Definition (Orientation of Triangles): For any triangle ABC , there are two
possibilities for its orientation. If we move from A to B to C again to A in
Prepared by Begashaw M. 121
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
counterclockwise direction (figure 3.7a), then ABC is assumed to have
positive orientation. If A, B, C follow the clockwise direction (figure 3.7b),
then ABC is assumed to have negative orientation.
Let A, B, C be vertices of ABC . Then, the orientation of ABC is determined
from the orientation of the vectors AB, AC . Thus, for
x z u zx u x
A , B , C , AB , AC
y w v w y v y
So, from Orientation Test Theorem, the orientation of the pair of vectors
zx ux
( AB, AC ) depends on the sign of det( AB, AC ) .
w y v y
a) If det( AB, AC ) 0, then ( AB, AC ) will have positive orientation and so is
ABC b) If det( AB, AC ) 0, then ( AB, AC ) will have negative orientation and so
is ABC
Example: Determine the orientation of ABC with vertices
Prepared by Begashaw M. 122
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1 3 4
A , B , C .
3 2 6
2 3 2 3
Solution: Here, AB , AC det( AB, AC ) 9 0.
1 3 1 3
This means the pair of vectors ( AB, AC ) has positive orientation which is the
orientation of ABC .
3.5.3 Orientation Preserving and Orientation Reversing Isometries
Definitions: Let g be any orthogonal transformation. Then, we say that g
preserves orientation if and only if for any positively oriented vectors X and Y
, their images X ' g ( X ), Y ' g (Y ) are again positively oriented vectors. In this
case, g is said to be orientation preserving orthogonal transformation. In
general, if the pair ( X , Y ) and the pair ( g ( X ), g (Y )) have the same orientation,
then g preserves orientation. But, if they have opposite orientation, then g
reverses (changes) orientation. In this case, g is said to be orientation reversing
(changing) orthogonal transformation.
Examples:
1. Determine whether the following isometries preserve or reverse orientation.
x y x x x x 3
a) g b) g c) g
y x y y y y 2
x z
Solution: Let X and Y be positively oriented vectors. Then,
y w
x z
det( X , Y ) 0
y w
a) From the given formula,
x y z w
X ' g ( X ) g , Y ' g (Y ) g
y x w z
y w x z
det( X ' , Y ' ) xw yz 0
x z y w
Prepared by Begashaw M. 123
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Thus, the pair ( X ' , Y ' ) is positively oriented, has the same orientation to the pair
( X , Y ) , which shows that g preserves orientation.
x z
b) Similarly as in part (a), X ' g ( X ) , Y ' g (Y )
y w
x z x z
det( X ' , Y ' ) xw yz yz xw 0
y w y w
Thus, the pair ( X ' , Y ' ) is negatively oriented, has opposite orientation to the pair
( X , Y ) , which implies that g reverses or changes orientation.
c) Similarly, we get that g preserves orientation.
Theorem 3.8 (General Orientation Test for Transformations):
x ax by
i) Let g : R 2 R 2 be any transformation given by g . Then, g
y cx dy
a b
preserves orientation if and only if 0.
c d
x ax by h
ii) In general, if g is given by g .
y cx dy k
a b
Then, g preserves orientation if and only if 0.
c d
x z
Proof: Suppose g preserves orientation and let X and Y be
y w
positively oriented vectors. Then,
x ax by az bw ax by az bw a b x z
X ' g , Y ' det( X ' , Y ' )
y cx dy cz dw cx dy cz dw c d y w
x
Since g preserves orientation, for any positively oriented vectors X and
y
z x z
Y , det( X , Y ) 0 implies X ' and Y ' are also positively oriented.
w y w
a b x z a b x z
Hence, det( X ' , Y ' ) 0 0 because 0.
c d y w c d y w
Prepared by Begashaw M. 124
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
a b
Conversely, suppose 0 . Then, for any positively oriented vectors
c d
x z
X and Y , we need to show X ' and Y ' are also positively oriented.
y w
x z
But, ( X , Y ) is positively oriented implies that det( X , Y ) 0.
y w
a b x z
On the other hand, from 0 and det( X , Y ) 0 , we have that
c d y w
ax by az bw a b x z
det( X ' , Y ' ) 0 det( X ' , Y ' ) 0 .
cx dy cz dw c d y w
Therefore, g preserves orientation.
In general, whether a given transformation g preserves or reverses orientation
is determined from its effect on the orientation of a triangle. This means if g
preserves the orientation of any triangle ABC , then it is orientation preserving
and if g reverses the orientation of ABC , then it is orientation reversing
transformation.
Now, having this fact as basis, let’s see the prove of the second part. Let
ABC be arbitrary triangle. Then its orientation is determined from the
orientation of the vectors AB and AC . Suppose ABC has positive orientation.
That means det( AB, AC ) 0 . On the other hand, let A' B' C ' be the image of
ABC under g . Then, the orientation of A' B' C ' is determined from the
orientation of the pair ( A' B', A' C ') .
Now suppose g is orientation preserving.
x ax by h x z u
But from g , using A , B , C
y cx dy k y w v
ax by h az bw h au bv h
A' g ( A) , B' g ( B) , C ' g (C )
cx dy k cz dw k cu dv k
a ( z x ) b( w y ) a(u x) b(v y )
A' B' , A' C '
c( z x) d ( w y ) c(u x) d (v y )
Prepared by Begashaw M. 125
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Thus,
a( z x) b( w y ) a(u x) b(v y )
det( A' B', A' C ') det
c( z x) d ( w y ) c(u x) d (v y )
a b z x u x
det (Using matrix multiplication property)
c d w y v y
a b z x u x
det det (Using the property,det (AB) det A.detB )
c d w y v y
a b z x u x
det det det( AB , AC ), (det det( AB , AC )
c d w y v y
a b
det( AB , AC )
c d
Hence, g preserves orientation if and only if
a b a b
det( A' B', A' C ') 0 det( AB, AC ) 0 0 because det( AB, AC ) 0
c d c d
from our assumption.
Similarly, g reverses orientation if and only if
a b a b
det( A' B', A' C ') 0 det( AB, AC ) 0 0.
c d c d
Examples: Determine whether the following transformations are orientation
preserving or orientation reversing.
x 3x 2 y 7
a) g : R 2 R 2 given by g
y x 5 y 11
x x 1
b) : R 2 R 2 given by
y y 11
Solution:
x 3x 2 y 7 a b 3 2
a) Here, g a 3, b 2, c 1, d 5 17 0
y x 5 y 11 c d 1 5
Hence, by the above theorem g preserves orientation.
x x 1 a b 1 0
b) Here, a 1, b 0, c 0, d 1 1 0
y 2 y c d 0 1
Prepared by Begashaw M. 126
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Hence, by the above theorem reverses orientation.
3.5.4 Orientation and the Fundamental Types of Isometries
In our discussion of isometries, we have seen as there are four basic types of
isometries: Translation, Rotation, Reflection and Glide- reflections. Now, using
the above general orientation tests, let’s see which of those orthogonal
transformations are orientation preserving and orientation reversing (changing)
isometries.
I. Rotations:
Let RC , be a counter clockwise rotation with center C through an angle of .
x z
Then, for any vectors X and Y , the isometer RC , is given by
y w
x cos y sin z cos w sin
X ' RC , ( X ) and Y ' R (Y )
x sin y cos z sin w cos
Here, using the orientation test for vectors, we have
x cos y sin z cos w sin
det( X ' , Y ' )
x sin y cos z sin w cos
( x cos y sin )( z sin w cos ) [( x sin y cos )( z cos w sin )]
xw cos 2 yz sin 2 xw sin 2 yz cos 2
xw(sin 2 cos 2 yz (sin 2 cos 2 )
xw yz
x z
0 det( X ' , Y ' ) 0
y w
Thus, the pair ( X ' , Y ' ) is positively oriented for any positively oriented vectors
( X , Y ) . This means RC , preserves orientation.
Alternatively, using the above theorem,
x cos y sin h
X ' R ( X )
x sin y cos k
a cos , b sin , c sin , d cos
a b cos sin
c d sin cos
cos 2 sin 2 1 0
Prepared by Begashaw M. 127
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Thus, RC , is orientation preserving isometry.
II. Translations: Let T : R 2 R 2 be a translation with translation vector
a x x h
v . Then, T . Is T orientation preserving or orientation
b y y k
reversing?
x x h a b 1 0
Here, T a 1, b 0, c 0, d 1 1 0
y y k c d 0 1
Hence, T is orientation preserving isometry.
III. Reflections:
Let l : ax by c 0 be any line and S l be a reflection on line l .
2a(ax by c)
x x x' x' x a2 b2
For any vector X in a plane, S l where
y y y' y ' y 2b(ax by c)
a2 b2
After some rearrangement, we get
(b 2 a 2 ) 2ab 2ac
x' 2 x 2 y 2
a b 2
a b 2
a b2
y ' 2ab x (a b ) y 2bc
2 2
a2 b2 a2 b2 a2 b2
From, these equations we obtain the determinant of the coefficients as
b2 a2 2ab
a b 2 (b a )(a b ) 4a b (b a ) 4a b
2 2 2 2 2 2 2 2 2 2 2
a2 b2 2
2ab a2 b2 (a 2 b 2 ) 2 (a 2 b 2 ) 2
a2 b2 a2 b2
(b 4 2a 2 b 2 a 4 ) 4a 2 b 2 [b 4 2a 2 b 2 a 4 )
(a 2 b 2 ) 2 (a 2 b 2 ) 2
[b 4 2a 2 b 2 a 4 ) (b 2 a 2 ) 2
1 1
(a 2 b 2 ) 2 (a 2 b 2 ) 2
Thus, a reflection, S l on any line l is orientation preserving isometry.
IV. Glide-Reflections
Any glide reflection is a composition of a reflection S l over a line l and a
translation Tv with non-zero vector v where the line l is parallel to the
direction of the translator vector v . But from (I ) and (III ) cases above, we saw
that translations and reflections are orientation preserving and orientation
Prepared by Begashaw M. 128
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
reversing isometries respectively. Therefore, their composition will be
orientation reversing isometry (we will prove this later on as a theorem in
Chapter 5) and thus glide reflection is orientation reversing.
To sum up our discussion, let’s summarize our results above as conclusion.
Conclusion: Translations and rotations (including identity) are the only types
of isometries preserving orientation. Reflections and glide-reflections are the
only types of isometries reversing (changing) orientation.
Example: Suppose is an isometry which maps PQR into P' Q' R' where the
vertices of the triangles are
4 1 12 2 1 10
P , Q , R , P' , Q' , R' .
2 2 4 3 1 3
Determine whether is a translation, rotation, glide reflection or reflection
and find its equation.
Solution: To apply the above test, first determine the orientation of PQR and
P' Q' R' .The orientation of PQR is determined from det PQ, PR where
det PQ, PR 3 8
4 6
50 0 .Hence, PQR has positive orientation.
Similarly, the orientation of P' Q' R' is determined from det P' Q', P' R' where
det P' Q', P' R'
3 8
4 6
50 0 which shows that P' Q' R' also has positive
orientation. So, we have got that PQR and P' Q' R' have the same orientation.
As a result, is orientation preserving isometry. Therefore, is either a
translation or a rotation.
Now, to determine whether is a translation or a rotation, find the vectors
PP ', QQ', RR ' .
2 2 2
Here, PP ' , QQ' , RR ' . Thus, PP ' QQ' RR '.
1 1 1
Prepared by Begashaw M. 129
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x x 2
So, is a translation given by .
y y 1
(What happens if PP' QQ' RR' ? We will see the genral case in section 3.9)
3.6 Fixed Points of Isometries
Classification of isometries based on their fixed points:
So far we have seen about fixed points of transformations. Now, among the
fundamental types of isometries, we are going to see which isometrics are with
fixed points and without fixed points.
i) Exacltly one fixed point: Isometries that have exactly one fixed point are
only Rotations: Any rotation has exactly one fixed point and the fixed point is
exactly the center of the rotation.
Example: Show that that the only fixed point of a rotation R about the origin
is the origin itself.
Solution: Let P ( x, y) be arbitrary fixed point of R . Then we need to find
the coordinates of P . From the definition of fixed point, we have that
x x cos y sin
R ( P) R ( x, y) P ( x, y) . Now, we have to solve this
y x sin y cos
system for x and y .
x cos x
From x x cos y sin , solving for y we get y . Substitute this
sin
value of y in the second equation y x sin y cos .
x cos x x cos x
x sin cos
sin sin
x cos x x sin 2 x cos 2 x cos
sin sin
x x
x cot x cot
sin sin
x cot 1
x cot x( )0x0
sin sin
So the only solution for is x 0 and substituting this value of x in
Prepared by Begashaw M. 130
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x cos x
y , we get y 0 .
sin
Therefore, the only fixed point of a rotation R about the origin is P (0,0)
which is the origin itself.
ii) Two or more fixed points:
Any isometry that has two fixed points but not identity is a reflection over a
line and the whole points on the line of reflection are also fixed points.
As a result the line of reflection is a fixed line.
iii) Three non-collinear fixed points:
An isometry that has three non-collinear fixed points is an identity.
iv) No fixed poin:
Isometries that have no fixed point at all- This category includes translation
and glide reflection.
Examples:
1. An isometry has exactly one fixed point (5,1) and maps the point (7,2) into
(3,0) . Then, find the equation of .
Solution: An isometry with exactly one fixed point is a rotation where the
fixed point is the center itself.So, is a rotation with center (5,1) .
Hence, its equation at any point ( x, y) becomes C , ( x, y) ( x' , y' ) where
x' ( x 5) cos ( y 1) sin 5
y ' ( x 5) sin ( y 1) cos 1
x' 2 cos sin 5 3 2 cos sin 2
In particular, C , (7,2) (3,0)
y' 2 sin cos 1 0 2 sin cos 1
Here, adding twice the second on the second gives 5 cos 5 cos 1 .
Similarly, subtracting twice the second from the first gives
5sin 0 sin 0 . Thus, cos 1, sin 0 .
Therefore, the equation of is given by (5,1), ( x, y) (10 x,2 y) .
2. A non identity isometry fixes the points (3,5) and (1,4) . Then, find the
equation of .
Prepared by Begashaw M. 131
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Solution: A non identity two given fixed points is a reflection on aline through
the fixed points. So, is a reflection on a line through the points (3,5) and
9 9
(1,4) . Thus, the line of reflection has slope m . Hence, l : y x b but
4 4
(3,5) l .
9 27 7 9 7
That is y x b 5 b b y x 9x 4 y 7 0 .
4 4 4 4 4
Hence, you can find the equations of using general equations of a
reflection.
3.7 Linear and Non-linear Isometries
Any given isometry is said to be linear or non-linear based on whether the
origin is its fixed point or not. In what follows, we are going to see linear
isometry and decomposition of an isometry using linear isometry.
Definition: Any isometry is said to be linear isometry if and only if it fixes
the origin. That means is linear if and only if (0) 0 .
Examples:
Prepared by Begashaw M. 132
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
a) : R R given by ( x) x is a linear isometry because (0) 0 .
b) : R 2 R 2 given by ( x, y) ( x, y) is a linear isometry because
(0,0) (0,0) .
c) : R 2 R 2 given by ( x, y) ( 12 x 2
3
y, 2
3
x 12 y) is a linear isometry
because (0,0) (0,0) .
d) : R 2 R 2 given by ( x, y) ( x 1, y 2) is not linear isometry. Because it
is an isometry but it does not fix the origin, (0,0) (1,2) (0,0).
Proposition 3.20: The composition of any two linear isometries is linear
isometry.
Proof: Let and be any two linear isometries on the same plane. Since the
composition of any two isometries is an isometry, is also an isometry.
Besides, as and are linear, (0) 0, (0) 0 .
Therefore, (0) ( (0)) (0) 0 .
Hence the composition preserves the origin and it is a linear isometry.
Proposition 3.21: Any given isometry can be expressed as a composition of a
translation and a linear isometry.
Proof: Let be any isometry. Define a vector v by v (0) and consider
Tv .
Claim: is a linear isometry. Since Tv and are isometries and so is their
composition (as the composition of any two isometries is again an isometry) .
Hence, Tv is an isometry. Beside,
(0) Tv (0) Tv ( (0)) Tv (v) v v 0 . This means fixes the origin.
Therefore, is a linear isometry. Thus, Tv Tv .
Hence, for any isometry , Tv where Tv is a translation and is a
linear isometry. But as is an arbitrary isometry, any isometry can be
expressed as a composition of a translation and a linear isometry.
Prepared by Begashaw M. 133
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Example: Let : R 2 R 2 be an isometry given by ( x, y) ( x 2, y 14) .
Express as a composition of a translation and a linear isometry.
Solution: Following the proof of the above proposition, take the vector
v (0,0) (2,14).
Then, for any object ( x, y) , take ( x, y) Tv ( x, y) ( x, y)
Here, clearly is an isometry. Besides, (0,0) (0,0) which means is a
linear isometry and Tv is a translation with translator vector v (2,14) .
Therefore, Tv where Tv ( x, y) ( x 2, y 14), ( x, y) ( x, y).
Proposition 3.22: Every isometry f in R has a unique decomposition of the
form f Tc where c f (0) and ( x) x .
Proof: Let f be an isometry in R such that c f (0) and let Tc be a translation
by c in R such that Tc f .
Claim: is a linear isometry. Since Tc and f are isometries and so is their
composition (as the composition of any two isometries is again an isometry).
Hence, Tc f is an isometry.
A gain, (0) Tc f (0) Tc ( f (0)) Tc (c) c c 0 . This means fixes
the origin and thus is a linear isometry. Besides, in R , d ( x, y) x y .
( x)
Hence, ( x) ( x) (0) x 0 x ( x) x 1, x 0, x R.
x
( x)
If 1 at least for some x1 , x2 , we get ( x1 ) x1 and ( x2 ) x2 which
x
implies ( x2 ) ( x1 ) x2 x1 .
But this is a contradiction because is an isometry.
Therefore, ( x) x, x R.
Hence, from Tc f , composing both sides on the left by Tc gives
Tc Tc (Tc f ) f Tc which means f ( x) Tc ( x) x c
where c f (0) and ( x) x . Complete the uniqueness.
Prepared by Begashaw M. 134
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Example: Let f : R R be an isometry given by f (3) 6 . Find a translation T
and a linear isometry with (2) 2 .
Solution: Any isometry in R is of the form f ( x) x c, x R for a fixed
constant c .
But f (3) 6 3 c 6 c 3 or c 9 f ( x) x 3 or f ( x) x 9 . Then,
following the proof of the above proposition,
f ( x) x 3 ( x) (Tc f )( x) (T3 f )( x) x 3 3 ( x) x or
f ( x) x 9 ( x) (Tc f )( x) (T9 f )( x) x 9 9 x ( x) x
Here, we have two options for , that is ( x) x or ( x) x .
But (2) 2 ( x) x .
Hence, we have got a linear isometry and a translation T9 such that
f T9 where T9 ( x) x 9, ( x) x .
Problem Set 3.6
1. Determine the orientation of ABC whose vertices are
4 1 1
A , B , C
1 1 3
2. Let f : R R be an isometry with f (2) 10 . Find a translation T and a linear
isometry with (2) 2 .
Answer : f T12 where T12 ( x) x 12, ( x) x
3. Let be a half turn about C (1,0) . Find a vector v and a linear isometry
such that Tv . Answer : v (2,0), ( x, y) ( x, y)
Prepared by Begashaw M. 135
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
4. Let f be a CCW rotation about C (4,2) by an angle of . Find a
3
translation Tv and a linear isometry such that f Tv .
1 3 3 1
Answer :Tv ( x, y) ( x 2 3, y 1 2 3 ), ( x, y) ( x y, x y)
2 2 2 2
5. Let : R 2 R 2 be given by ( x, y) ( 12 x 2
3
y 1, 2
3
x 12 y 1) . Show that
is an isometry but not linear isometry. Find a translation Tv and a linear
isometry such that Tv .
1 3 3 1
Answer :Tv ( x, y) (1,1), ( x, y) ( x y, x y)
2 2 2 2
6. If S l fixes the points (1,4) and (5,0) ,then find the equation of line l .
Answer : l : 2 x 3 y 10 0
7. Suppose f is a non-identity isometry such that f ( 5,4) ( 5,4) and
f (0,1) (0,1) . Find the value of f (1,0) and explain how you know f (1,0)
must have that value. Determine the general equation of this isometry.
Answer : It a reflection on a line l : y 5 x 1
8. If an isometry is involution, prove that for any point P , the mid point of P
and (P) is a fixed point of .
9. Suppose f is a non-identity isometry which fixes (1,1) and (2,2) . Find the
equation of f and calculate f (1,0).
10. Prove that the inverse of any linear isometry is also linear isometry.
3.8 Representations of Orthogonal (Isometric)
Transformation as a Product of Reflections
So far, we have discussed that the product of two isometrics is a gain an
isometry. Besides, we have encountered four types of isometries: reflections,
translations, rotations and glide reflections. From now onwards, our concern is
Prepared by Begashaw M. 136
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
to investigate: ”Are these the only isometries or do any other exist?” In other
words, from the fact that the product of two isometries is again an isometry, it
is natural to ask ourselves whether the product of any of two isometries we
have seen so far results among one of those or a new one that we did not
discussed. By investigating, all of the possible combination products of
reflection, rotation and translation, we will find out that the result is any one of
these but not a new form of isometry. Finally, we will generalize that there are
no other isometries or our bag of isometries is complete. Besides, we will
conclude that reflections are the building blocks for plane isometries from the
result that every isometry is the product of three or fewer reflections. Further
more; if it is a rotation, then through what angle and with what center? If it is a
translation, what is the translation vector? If it is a reflection, what is the line of
reflection? and so on will be answered.
3.8.1 Product of Reflections on Two Lines
When we say product of reflections on two lines , we mean S n S m , reflection
on line m followed by a reflection on line n . Here, m and n may be
intersecting or parallel. From our next discussion, we will point out that the
composition of two reflections S n S m is a rotation, a translation or a reflection
itself. So, later on after the discussion, you should be able to describe in detail
in what situation it will be a rotation, translation or reflection.
Case I: When the two Lines are Intersecting
Theorem 3.7: The composition of two reflections on two intersecting lines l
and m is a rotation about the point of intersection through twice the directed
angle between the lines.
Restatement: Let l and m be any two lines intersecting at c (a, b) and the
angle measured from l to m being . Then S m S l c, 2
Prepared by Begashaw M. 137
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Proof: Consider the figure 3.9a: To prove this theorem, it suffices to show that
PQP" 2RQT and QP QP '' . In order to show these two conditions, let’s
consider two different cases:
Case-1: Suppose P m . Then, PR P ' R by definition of reflection and
QR QR by reflexive property. Again, QR is perpendicular to PP ' as m is the
line of reflection of P to P ' . Hence, PQR P' QR by SAS which implies that
PQR P' QR .
Similarly, P' QT P"QT by SAS and hence P' QT P' ' QT . But,
m(PQP" ) mPQP' ) m(P' QP" ) 2m(RQT ) 2 . (Do you see how?).
Besides, as PQP ' and P ' QP '' are isosceles, PQ P ' Q and P ' Q P '' Q , then
by transitivity PQ P '' Q .
Case-2: Suppose either P m or P ' n . For P m , Sm ( P) P and Sn ( P) P ' .
Yet, with similar argument as in the first case, PQR P' QR by SAS,
so, PQR P' QR and thus
m(PQP' ) mPQR) m(P' QR) 2m(PQR) 2 and QP QP ' .
The same reasoning holds when P ' n . Therefore, the theorem follows.
Examples:
1. Let l : 2 x y 3 0 and m : 3x y 7 0. Then find the image of the point
(5,7) by the composite reflection S m S l using single rotation.
Solution: Since the two lines have different slopes, they must be intersecting
lines. So, to apply theorem 3.7, first let’s find the point of intersection and the
angle between them. The point of intersection is obtained as
Prepared by Begashaw M. 138
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2 x y 3 0
x 2, y 1 c (2,1) . Let be the angle between l and m .
3x y 7 0
m1 m2
Then, from coordinate geometry, tan where m1 and m2 are slopes
1 m1 m2
of the non-vertical lines l and m . Thus,
m1 m2 3 2 5
tan 1
1 m1m2 1 2(3) 5
tan 1 tan 1 (1) 45 or 135
To calculate S m S l (5,7) determine which angle will be used. (Be careful here!
On how to determine the angle). Hence, S m S l (5,7) R( 2,1), 90 (5,7) (10,4) .
2. If m : y x 1 and S m S l (1,1) (3, 2), 90 (1,1) , find the equation of l.
Solution: By Theorem 3.7, the equation S m S l (1,1) (3, 2), 90 (1,1) gives us the
two lines are intersecting at (3,2) and the angle between them is 45 . Then,
1 m
if m is the slope of l , we have that tan 45 1 m 1 m m 0 .This
1 m
means that the line l is a horizontal line passing through the point (3,2) .
Therefore, its equation is l : y 2 .
Corollary 3.2: Let l and m be two lines intersecting at c (a, b). Then,
S l S m c,2 where is the angle measured from l to m .
Proof: Consider figure 3.9b. Since the directed angle from l to m is , the
directed angle from m to l (in the same direction) is .
So, by theorem 3.7, S l S m ( x, y) c, 2 ( x, y) , where
2( ) . Then, by using the generalized rotation theorem,
Prepared by Begashaw M. 139
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
S l S m ( x, y ) c , 2 ( x, y )
( x a) cos(2 2 ) ( y b) sin( 2 2 ) a
( x a ) sin( 2 2 ) ( y b ) cos( 2 2 ) b
( x a) cos(2 ) ( y b) sin( 2 ) a
( x a ) sin( 2 ) ( y b ) cos( 2 ) b
c , 2 ( x, y )
Hence, S l S m c,2 whenever S m S l c, 2 . This result shows that if we
interchange the order of the composition the sign of the angle will also be
changed. So, S m S l c, 2 if and only if S l S m c,2 .
Corollary 3.3: Let l and m be any two lines intersecting at c (a, b) and the
angle measured from l to m being .
x' ( x a) cos 2 ( y b) sin 2 a
Then, S m S l ( x, y) ( x' , y' ) where
y' ( x a) sin 2 ( y b) cos 2 b
Proof: (Expand the result of theorem 3.7 using theorem 3.4)
Theorem 3.8 (Converse of theorem 3.7): Given a rotation C , . Let m and n
be any two lines intersecting at C so that the angle between them is . Then,
2
C , S n S m .
Proof: Given point C , lines m and n with (m, n) . Consider figure 3.10.
2
Let M and N be points on m and n so that d (C, M ) d (C, N ). Clearly, C, M
and N are non-collinear points. Now, let S m ( N ) N ' and S n (M ) M ' .
Since center of rotation is fixed under any rotation and bedsides C is on both
lines m and n , S m (C ) C and S n (C ) C which implies S n S m (C ) C.
Prepared by Begashaw M. 140
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Thus, C is fixed both under C , and S n S m . Again, as S n is an isometry,
d (C, M ) d (C, M ' ). Besides, as isometry preserves angles,
m( M C , M 'C ) m( M C , N C ) m( N C , M 'C ) .
2 2
So, from d (C, M ) d (C, M ' ) and m(M C, M 'C) , we get C , (M ) M '.
On the other hand, we have S n S m (M ) S n (M ) M ' which results in
C , (M ) S n S m (M ) M '. Similarly, from d (C, N ) d (C, N ' ) and
m( N C, N 'C ) , we have C , ( N ' ) N and then
C , ( N ' ) S n S m ( N ' ) S n ( N ) N .
But, as C, N ' and M are non-collinear, C, N and M are also non-collinear
besides C , (C ) S n S m (C ) , C , (M ) S n S m (M ) and C , ( N ' ) S n S m ( N ' ) .
This shows that C , and S n S m agrees on three non-collinear points. But, by
the three point theorem (Theorem 3.1), any two isometries are equal if they
agree on three non-collinear points. Therefore, C , S n S m .
Theorem 3.9: The product of two reflections on perpendicular lines m and n
at a point P is a half turn with center P . That means S n S m H P . Conversely,
a half turn about a point is the product of two reflections on two perpendicular
lines intersecting at the center of the half turn.
Proof: Let m and n be perpendicular lines intersecting at P . By theorem 3.7,
the product of two reflections on two intersecting lines is a rotation with center
at the point of intersection by twice the angle between them.
In this case, since the two lines are perpendicular, .
2
Hence, S n S m P, 2 P, H P .From this theorem, we can generalize that
the product of two reflections on intersecting lines is a half turn about their
point of intersection if and only if the two lines are perpendicular.
Prepared by Begashaw M. 141
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Examples:
1. Find the equations of two lines m and n such that (2,4) is on m and
S n S m H P where H P is a half turn with center P (3,0).
Solution: From theorem 3.9, the product of two reflections is equal to a half
turn with center P if and only if the two lines are perpendicular at P . So, we
have to assume that the lines m and n are perpendicular at P (3,0) so as
S n S m to be a half turn about P . Let the equation of m be y ax b .
From the given, the points (3,0) and (2,4) are on m which implies that slope
40
of m is a 4.
23
So, m : y ax b y 4 x b and (3,0) m 0 12 b b 12.
Therefore, m : y 4 x 12 .
1 3 3
Now, m n n : y x c . Besides, (3,0) n 0 c c .
4 4 4
1 3
Therefore, n : y x .
4 4
2. Let m : y 2 x 1. Find the equation of line n such that S n S m is a half turn
about P (1,3) .
Solution: By Theorem 3.9, S n S m H P if and only if the two lines are
1
perpendicular at P . Now since the slope of m is 2 , the slope of n must be .
2
1
Thus, n : y x b .
2
1 1 7
Besides as P (1,3) is on n . y x b b 3 b .
2 2 2
1 7
Hence its equation n is y x .
2 2
Prepared by Begashaw M. 142
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Remark (Condition for product of reflections to be Commutative):
Let l and m be two lines intersecting at point C . We need to investigate the
condition at which S m S l S l S m .
First recall from theorem 3.6, , 1 and
i 2n , n Z . Now, from theorem 3.7 and 3.8, we have
S m S l C , 2 where is the directed angle from l to m .
On the other hand, from corollary 3.2 , S l S m C ,2 .
Consequently,
S m S l S l S m RC , 2 RC , 2
RC , 2 RC , 2 RC , 2 RC , 2 , RC , 2 RC , 2 i
RC , 4 i 4 2k , k Z
If k 0, 0 , which is impossible since the lines are intersecting and the angle
between any two intersecting lines is non zero.
If k 1, 2 which means the lines are perpendicular.
If k 1, 2 which means again the lines are perpendicular.
Since the smaller of the two angles between two intersecting lines is less than
or equal to 2 , there is no need to consider any other values of k . Otherwise if
we consider k 2 , we get which is a contradiction with the fact that if
is the angle between any two distinct intersecting lines, then 0 .
From this analysis, we can conclude that the product of reflections on two
distinct intersecting lines is commutative if and only if the two lines are
perpendicular. Hence, S m S l S l S m l m .
Example: Let l and m be intersecting lines such that S m S l S l S m where
m : x 5 y 17 0 . If S m S l (1,1) (5,5), find the equation of line l .
Solution: Our previous discussion tells us that S m S l S l S m l m .
Besides, S m S l P, 2 where P is the point of intersection say (a, b) and is
the angle between the lines which is 2 in this case.
Thus, S m S l ( x, y) P, ( x, y) ( x 2a, y 2b) for any point ( x, y) (The aim is
to determine the intersection point of the lines). But,
S m S l (1,1) (5,5) (1 2a,1 2b) (5,5) a 2, b 3 P (2,3) .
Hence, the point of intersection is determined to be (2,3).
Prepared by Begashaw M. 143
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
On the other hand, m l m n implies slope of line l is 5 .Thus, l : y 5x b
and (2,3) l 3 10 b b 7 .
So, the equation of l is found to be l : y 5x 7.
Case II: When the two Lines are Parallel
Theorem 3.10: The composition of two reflections on two parallel lines l and
m is a translation by a vector 2v where v is a vector perpendicular to both l
and m . Conversely, if Tv is a translation, then there are parallel lines l and m
such that Tv Sl Sm , where l and m are parallel lines perpendicular to the
translator vector v and the distance between them is half the length of v .
Note that for any given vector v and any line l , there is a unique line m parallel
to l such that Tv Sl Sm . This can be justified as follow: If v is perpendicular
to both l and m , the relation holds true. Now, if l is perpendicular to v , then
take line m to be m : ( x, y) l 12 v M : M p 12 v, p l.
1
This is the translation of line l by v.
2
So, from the previous theorem, we can easily show that Sl Sm T 1 Tv
2( v )
2
Proof: (Left as an exercise)
Remarks: This theorem ensures that the product of reflections on two parallel
lines is a translation and any translation is also a product of reflections on two
parallel lines. If S m S l T2v , then S l S m T2v where l and m are parallel
lines and v is a vector perpendicular to both l and m . If L is on l and M is
on m where the line through L and M is perpendicular to both l and m , then
S m S l T2 LM T 2 LM H M H L and S l S m T2 LM T2 ML T 2 ML H L H M .
The composition reflections S m S l and S l S m are inverse of each others.
Prepared by Begashaw M. 144
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Examples:
1. Given the lines m : y 2 x 1 and n : y 2 x 3 . Find the image of the point
(1,1) by a product of reflection on line m followed by line n using direct
formula, translation, half turn and compare the answers.
Solution: We need to find S n S m (1,1) . This problem can be done using three
methods:
Method I: Using reflection (Direct) Formula: First calculate S m (1,1) using
reflection equation as
4(2 1 1) 3
x' 1
4 1 5
S m (1,1) ( x' , y ' )
y ' 1 2(2 1 1) 9
4 1 5
6 9
4( 3)
3 5 5 21
x' '
Now, S n S m (1,1) S n ( S m (1,1)) S n ( 3 , 9 ) 5 4 1 5
5 5 6 9
2( 3)
9 5 5 3
y ' '
5 4 1 5
21 3
Therefore, S n S m (1,1) ( , ) .
5 5
Method II: Using Translation:
Since the given lines are parallel, the product of reflections on these lines is the
same as a translation by the double of the perpendicular vector pointing from
m to n. That means, S n S m (1,1) T2 MN where M m and N n such that
MN m, n . Here emphasis should be given on how to find such points.
This is accomplished by taking any line that is perpendicular to the given lines
and it is simple to find the intersections of this line with the lines m and n.
Prepared by Begashaw M. 145
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1
Take l : y x . The intersection of this line with the lines m and n is found
2
as follow:
Intersection of l and m :
1 5 2 1
y 2 x 1 x x 1 x , y
2 2 5 5
2 1
M ( x, y ) ( , )
5 5
Intersection of l and n :
1 5 6 3
y 2x 3 x x 3 x , y
2 2 5 5
6 3
N ( x, y ) ( , )
5 5
8 4
Thus, the vector MN is found to be MN N M ( , )
5 5
16 8 21 3
Therefore, S n S m (1,1) T2 MN (1,1) (1 ,1 ) ( , )
5 5 5 5
Method III: Using Half turns: This problem can also be done using product
of half turns as S n S m (1,1) H N H M (1,1) where the centers of the half turns
M and N are as explained above.
So,
9 3 21 3
S n S m (1,1) H N H M (1,1) H 6 3 (H 21 (1,1)) H 6 3 ( , ) ( , ) .
( , ) ( , ) ( , ) 5 5 5 5
5 5 5 5 5 5
The use of different methods in solving the above example will help us to grasp
the relations among reflections, translations and half turns.
2. Suppose the line l : y 2 x 1 is parallel to the line m. If the point (9,2) is
on line m , then find S m S l (2,5) and S l S m (2,5) using a single translation.
Solution: Since, l and m are parallel, to do the problems, we need only two
point L on l and M on m where the line through L and M is perpendicular
to both l and m . But, M (9,2) m (Given).
Here, L (a, b) on l b 2a 1.
Prepared by Begashaw M. 146
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1
Again, LM l implies slope of a line a long LM is which gives
2
b 2 1
2b a 13 . Combining these two equations, we get
a 9 2
b 2a 1
a 3, b 5 L (a, b) (3,5)
2b a 13
So, S m S l (2,5) T 2 L,M (2,5) where T is a unique translation that takes L to M .
Let this translator vector be v (c, d ) . But,
T ( L) M T (3,5) (9,2) 3 c 9, 5 d 2 c 6, d 3 . Hence, the
translator vector of T is v (6,3) so that T ( x, y) ( x 6, y 3) for any point
( x, y) . So, S m S l (2,5) T 2 L,M (2,5) TL,M (TL,M (2,5) TL,M (8,2) (14,1) .
Besides, using the relation S m S l T2 LM S l S m T2 LM in the remark, we
get, S l S m (2,5) T LM (T LM (2,5) T LM (4,8) (10,11) . Once the points L and
M are determined, this problem can also be done using half turns about L and
M using proposition 3.13 and Theorem 3.10 as follow.
S m S l (2,5) H M H L (2,5) H M (4,5) (14,1) and
S l S m (2,5) H L H M (2,5) H L (16,1) (10,11) .
This means that from proposition 3.13 and Theorem 3.10 we can relate
reflections, translations and half turns as S l S m T2 ML H L H M where l // m
and the points L and M are on the lines l and m respectively such that the
vector LM is perpendicular to the lines.
3. Let l and m be two parallel lines. If S m S l (2,1) (8,7) , calculate
S m S l (3,4) and S l S m (3,4) .
Solution: Since the lines l and m are parallel, by Theorem, 3.10, we have
above theorem, we have S l S m T2v where v is a perpendicular vector directed
from m to l . So, for any point P ,
P' S l S m ( P) T2v ( P) 2v P P' .
Prepared by Begashaw M. 147
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Particularly, for P (2,1) ,
S l S m (2,1) T2v (2,1) 2v (2,1) (8,7) 2v (10,6) v (5,3) .
Hence, S m S l (3,4) T2v (3,4) 2(5,3) (3,4) (7,10) . On the other hand,
S l S m T2v S m S l T2v .Thereofe,
S l S m (3,4) T2v (3,4) 2(5,3) (3,4) (13,2) .
4. Given l : y 2 x 1and a vector v (12,6) . Find a line m parallel to l such
that S m S l Tv .
Solution: By the converse of the above theorem, the line m is obtained by
translating the given line by 12 v . For any point ( x, y) on l ,
T1 v ( x, y) ( x 6, y 3) ( x' , y' ) x x'6, y y'3 .
2
So, l : y'3 2( x'6) 1 l : y' 2 x'16 .
5. Let l : y x and m : y x 4 . Then, show that S m Sl is a translation and find
its equation.
Solution: Since the two lines are parallel, by the above thorem S m Sl is a
translation. Besides, Sl ( x, y) ( y, x), S m ( x, y) ( y 4, x 4) .
Hence, S m Sl ( x, y) S m ( y, x) ( x 4, y 4) .
Therefore, S m Sl Tv where v (4,4) .
6. Let l : y 3x 20 and m : y 3x 10 . Then, find a translation T with
translator vector v such that S m S l Tv and calculate S m S l (0,0) .
Solution: By using similar procedures, as in example 5, we get
Tv ( x, y) ( x 18, y 6), S m S l (0,0) (18,6) .
Note: If the lines m and n are parallel and the points M and N are as
indicated in figure below, we always have the relation
S n S m T2 MN H N H M .
Prepared by Begashaw M. 148
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Here, we must give attention how to write the order of the subscripts in the
translation and half turns with the order of the reflections. For instance, if we
interchange the order of the product of the reflections, the order of the half
turns and the direction of the vector in the translation will also be changed. It
becomes S m S n T2 NM H M H N and the resulting image is also different.
Example: Given the line l : y 2 x 3 and a vector v (6,3) . Find a line m
such that Tv Sl Sm .
Solution: Apply the above result.
Remark (Characterizing products of rotations): Prior to this, we have seen
that the product of rotations about the same center is a gain a rotation about that
center. Besides, the product of rotations is commutative as far as the individual
rotations are rotations about the same center. Here, one may ask that what will
happen if the rotations are performed about different centers with different
angles of rotations as indicated in figure 3.11. In figure 3.11a, we can see that
RB, R A, ( P) P' ' .
Prepared by Begashaw M. 149
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Now we want to find a single isometry that will have the same effect as the
product RB, R A, . To investigate this, consider the two rotations R A, and RB ,
where A and B are different centers. Once we have two different points A and
B , we can determine a unique line AB . So, by the previous theorem there are
lines m and n through A and B respectively (Refer figure 3.11b) such that the
rotations R A, and RB , are expressed as RA, S l S m and RB, S n S l .
Thus, from the product RB, R A, is also expressed as
RB, RA, (S n S l ) (S l S m ) S n S m . This means if m // n , then the product
S n S m will be a translation and so is RB , R A, . If the lines m and n
intersect at some point, then S n S m will be a rotation about their intersection
point and so is RB, R A, .
Therefore, this observation is summarized by the following theorem.
Proposition 3.18 (Characterization of products of rotations):
Let R A, and RB , be two rotations about different centers A and B . Then, the
product RB, R A, is a rotation if and only if is not a multiple of 2 and it
is a translation whenever is a multiple of 2 . Since the proof of this
theorem needs some sketch pad construction, it is omitted.
Any way, the important result that the theorem tells us the product of two
rotations is either a rotation or a translation and in general the product is not
commutative.
Proof: (Follows from the general formula of rotation)
Prepared by Begashaw M. 150
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 3.7
1
1. Let l : y 2 x 1 , m : y 2 x 3 and n : y x . Find the image of (1,1) by
2
the composite reflections S m S l and S n S l . State the procedure you followed
21 3 9 3
Answer : S m S l (1,1) ( , ), S n S l (1,1) ( , )
in each case. 5 5 5 5
2. Let l be a line through the points (1,9), (4,1) and m be a line through the
points (6,2), (5,5) . Find the center and angle of a rotation C , such that
27 19
S m S l C , . , ),
Answer :C (
5 5 2
3. Let l be y x and v be vector v (2,2) . Find line m such that Tv Sl Sm .
4. Let T ( x, y) ( x, y 2). Find two lines l, m such that T S l S m .
5. Let l : y 2 and m : y 5 . Show that Sl Sm T2v where v (0,3) .
6. Let l and m be any two parallel lines. If S m S l (1,2) (7,8) , calculate
S l S m (3,4) . Answer : (3,14)
7. Let l be the line y 5 . Find line m such that S m S l .
( 3, 5),
2
Answer : m : y x 2
8. Given the lines l : x y 3 and m : y x 1. Express S l S m using a single
rotation C , . Answer : S l S m C , where C (1,2),
9. If m n such that S m S n (1,2) (5,6) , then find S m S n (3,4) .
Answer : S m S n (3,4) (3,12)
10. Let l and m be any two perpendicular lines intersecting at the point (2,3) ,
calculate S l S m (2,7) and S m S l (2,7) . Are the results equal? Why?
Answer : (6,1)
11. If A (2,4), find equations of lines m and n such that (0,3) is on m and
H A Sn Sm . Answer : n : 2 x y 8 0, m : x 2 y 3 0
12. Let l be the line x 3 and P R 2 be the point (5,4). Show that S l H P is a
glide reflection by finding the axis of the glide reflection and the glide vector.
13. Given a line m : y x 3 . Find the equation of line l such that
a) C , S m S l where C (7,10), / 2 b) S m S l H P where P (2,5)
c) S m S l Tv where v (3,3) d ) S m (l ) l with P (2,7) l
Answer : a) l : y 10 b) l : x y 7 0 c) l : x y 6 0 d ) l :x y 9 0
Prepared by Begashaw M. 151
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3.8.2 Product of Reflections on Three Lines
Case-I: When the three lines are concurrent
So far, we discussed that the product of reflections on two intersecting lines is a
rotation. But what will be the result if the product of reflection is on three
concurrent lines ( on three lines intersecting at a single point). This is what we
are going to address in the next theorem.
Theorem 3.11: Let m, n, p be lines intersecting at a point C. Then, there is a
line q through C so that S n S m S p S q .
Proof: Since m and n are intersecting lines at a point C , S n S m is a rotation
C , where the angle between m and n is .
2
Now, suppose q is a line through C such that the angle between p and q is
2
(this is supposition or assumption of line q is possible by angle construction
postulate). Then, S q S p C , .
So, C , S n S m and C , S q S p imply that
Sn Sm Sq S p
Sn Sm S p Sq S p S p
Sn Sm S p Sq i
Sn Sm S p Sq
From this theorem, we can conclude that the composition of reflections on
three concurrent lines is again a reflection on a line through the same point.
Case-II: When the three lines are parallel
So far, we discussed that the product of reflections on two parallel lines is a
translation. But what will be the result if the product of reflection is on three
parallel lines.
Prepared by Begashaw M. 152
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Theorem 3.12: Let l , m, and n be three parallel lines in a plane. Then there is a
unique line p parallel to the given lines l , m, and n such that Sm Sl Sn S p .
Equivalently S m S l S p S n .
Proof: Since l and m are parallel lines, from Theorem 3.10, Sm Sl Tv , where
v is perpendicular to both l and m . Similarly, as n is parallel to l and m , v
is also perpendicular to n .
Again, by the converse of theorem 3.10, there exists line p parallel to the
given line n such that Tv S p Sn . Thus combining the two results, we get
Sm Sl (Q) S p Sn (Q), for any point Q
( Sm Sl ) Sn (Q) ( S p S n ) S n (Q)
( Sm Sl ) Sn (Q) S p ( Sn Sn )(Q), because composition is associative
( Sm Sl ) Sn (Q) S p (Q), because S n Sn id , for any reflection Sn
S m Sl S n S p
From this theorem by considering one additional line q like that of p , we can
state a useful corollary.
Corollary 3.4: Let l , m, and n be three parallel lines in a plane. Then there exist
unique lines p and q parallel to the given lines l , m, and n such that
S m Sl S n S p S q S n
Proof: Consider the diagram below and look for the lines p and q from the
equation S m S l S n S p S q S n (This equation actually has unique solutions
for p and q ).
Prepared by Begashaw M. 153
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Let P and Q be unique points on line k such that the products of the half turns
with centers L, M , N , P, Q are equal as follow:
H M H L H N H P H Q H N . But, we know that a unique line can be drawn
through a point on a given line perpendicular to the line. So, let’s drop line p
through P and line q through Q both perpendicular to line k .
Now, l // m S m S l H M H L ,
n // p S n S p H N H P and q // n S q S n H Q H N .
Combining these and our previous relations H M H L H N H P H Q H N , we
get S m S l S n S p S q S n . From the first equality, S m S l S n S p we have
S n S m S l S p and from the second equality
S m S l S q S n , we get S m S l S n S q .
Remarks: The uniqueness of p and q is shown from the fact that reflections
on two lines are equal if and only if the two lines are equal.
That means S n S m n m.
So, if we assume there is another line h for which S n S p S n S h . Then,
S n S p S n S h S n S n S p S n S n S h S p S h p h.
1
Examples: Given the lines l : y 2 x 5, m : y 2 x , n : y 2 x 5 . Find the
2
lines p and q parallel to the given lines such that S m S l S n S p S q S n .
Solution: To do this problem, we use the analysis followed in the proof of the
above corollary. Since the given lines are parallel, take any line k
perpendicular to the given lines but the line k should be simple for analysis.
1
Let k be the line k : y x .
2
(Note the problem can also be done using any other line without affecting the
solution for p and q ). Now find the points L, M , N on the lines l , m, and n
respectively which are the intersections of line k with these lines.
Prepared by Begashaw M. 154
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Intersection of l and k :
1 5
y 2 x 5 x x 5 x 2, y 1 L ( x, y) (2,1)
2 2
Intersection of m and k :
1 1 5 1 1 1 1 1
y 2x x x x , y M ( x, y) ( , )
2 2 2 2 5 10 5 10
Intersection of n and k :
1 5
y 2 x 5 x x 5 x 2, y 1 N ( x, y) (2,1)
2 2
Now determine two points P and Q on line k which are the intersection of the
lines p and q with k respectively (as labeled in figure 3.12) such that
H M H L H N H P HQ H N .
Let P (a, b) and Q (c, d ). But, for any point ( x, y) ,
18 9
H M H L ( x, y ) ( x ,y )
5 5
H N H P ( x, y ) ( x 2a 4, y 2b 2)
H Q H N ( x, y ) ( x 2c 4, y 2d 2)
Thus,
18 9
H M H L ( x, y ) H N H P ( x, y ) ( x
, y ) ( x 2a 4, y 2b 2)
5 5
18 9
x x 2a 4, y y 2b 2
5 5
1 1
a , b
5 10
1 1
P ( a, b) ( , )
5 10
On the other hand, using the second equality H M H L H Q H N ,
18 9
H M H L ( x, y ) H Q H N ( x, y ) ( x , y ) ( x 2c 4, y 2d 2)
5 5
18 9
x x 2c 4, y y 2d 2
5 5
19 19
c , d
5 10
19 19
Q ( c, d ) ( , )
5 10
Since the lines p and q are parallel to the given lines, then they are also
perpendicular to line k .
Prepared by Begashaw M. 155
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Thus, we are required to determine lines p and q through points P and Q
which are perpendicular to the line k at these points of intersections.
1 1
p // l p : y 2 x b . But p passes through P ( , ) , then
5 10
1 2 1 1
y 2x b b b . Hence, line p is found to be p : y 2 x
10 5 2 2
Similarly, q is a line through Q and perpendicular to line k and parallel to the
19 38 19
lines. Thus, it has the form q : y 2 x c cc .
10 5 2
19
Hence, line q is found to be p : y 2 x . In this solution the choice of line
2
k is arbitrary, one can choose any other line perpendicular to the given lines
and gets the same result but the choice of line k should be in such a way that
one can calculate the points of intersection easily.
Case-III: When the three lines are neither parallel nor concurrent
Theorem 3.13: Let l , m, n be neither concurrent nor parallel lines (two may be
parallel). Then, S n S m S l is a glide reflection.
Proof: To prove that S n S m S l is a glide reflection, it suffices to find a line
p and a vector v with v // p such that S n S m S l Tv S p . Since all the three
lines are not parallel assume that m and n intersect at P. (Refer figure 3.13a).
Besides, l , m, n are not all concurrent, l does not pass through P . So, let k be a
line perpendicular to l through P . Here, m, n, k becomes concurrent at P .
Prepared by Begashaw M. 156
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Thus, by theorem 3.11, there is a line t through P such that S k S n S m S t
which in turn implies S n S m S t S k . Again let Q be the intersection of k and
l . As k l , we have S k S l H Q (because composition of reflections on two
perpendicular lines is a half turn about their intersection). Now, construct line
p perpendicular to t through Q and line q through Q parallel to t . Since t // q ,
there is a vector v perpendicular to lines t and q such
that S t S q Tv . Hence, combining these results above
S n S m S t S k , S k S l H Q and S t S q Tv , we get,
S n S m Sl St S k Sl
St H Q
St S q S p
Tv S p
But, from our construction, v t and p t implies v // p which shows that
Tv S p is a glide reflection. Therefore, S n S m S l Tv S p implies that
S n S m S l is a glide reflection. In this theorem, the lines l , m, n are neither
concurrent nor parallel (of course two may be parallel).
If l // m and n is perpendicular to l and m , then S m S l is a translation and S n is
a reflection, so the product S n S m S l is called glide reflection and line n is
known as the axis of the glide reflection.
Corollary 3.5: Let l be any line and P be a point not on l . Then, g H P S l
is a glide reflection.
Proof: Since P l , consider line n through P perpendicular to l and line m
perpendicular to line n at P . See figure 3.13
Prepared by Begashaw M. 157
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Now, m n and n l gives S n S m H P (Because product of two reflections
on perpendicular lines is a half turn about their intersection). So,
g H P S l S n S m S l . But l , m, n are neither concurrent nor all parallel from
theorem 3.11 S n S m S l is a glide reflection and so is H P S l as they are
equal. From figure 3.13, one can also easily verify that
g S n S l S m H Q S m S l S m S n S l H P using the fact that product of
two reflections on perpendicular lines commute and is a half turn through their
intersection. In general, for l // m , g H Q S m S l H P is a glide reflection if
and only if Q l and P m . This shows that glide reflection is the product of
three reflections on three neither parallel nor concurrent lines,
Or the product of a half turn and a reflection in either order where the center of
the half turn is not on the line like as P or Q above, Or the product of a
translation and a reflection in any order where the translation vector is parallel
to the line of reflection.
Examples:
1. Let m and n be parallel lines through M and N where M (1,8) and n is the
line y x 1 0 . If g S n H M H N S m is a glide reflection, find the
equation of H N , the axis of g and line m.
Solution: From corollary 3.5, g S n H M H N S m is a glide reflection
whenever M m , N n and the axis of g passes through M and N
perpendicular to the lines m and n . Now, m // n implies the slope of m is 1 so
then m : y x b . But, M m 8 1 b b 7 m : y x 7.
On the other hand, the axis of g is perpendicular to m implies the slope of the
axis of g is 1 and then l : y x c . But, the axis also passes through
M (1,8) which implies that l : y x c 8 1 c c 9 l : y x 9.
Prepared by Begashaw M. 158
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Similarly, this axis l : y x 9 which is perpendicular to m at M is also
perpendicular to n at N . That means N is the intersection point of
n : y x 1 0 and l : y x 9 .
So, y x 1 x 9 x 4 N ( x, y) (4,5).
Hence, H N ( x, y) (8 x,10 y) .
2. Let l be the line y 0 and A R 2 be the point (0,3). Show that H A S l is a
glide reflection by finding the axis of the glide reflection and the glide vector.
Solution: Let ( x, y) be any point so that H A ( x, y) ( x, y 6), S l ( x, y) ( x, y)
Then, H A S l ( x, y) H A ( x, y) ( x, y 6) …………………………… (i )
Here, take the line m : x 0 and the vector v (0,6).
Thus, S m Tv ( x, y) S m ( x, y 6) ( x, y 6) Tv S m ( x, y) …………………. (ii )
Combining (i ) and (ii ) , we get H A S l Tv S m . Besides, m // v and then
Tv S m is a glide reflection by definition.
Therefore, H A S l Tv S m implies H A S l itself is a glide reflection with axis
m : x 0 and a glide vector v (0,6).
Theorem 3.14: If f is an isometry with exactly one fixed point C , then
f S n S m where m and n are lines intersecting at the point C .
Proof: Suppose f is an isometry that fixes only the point C. Let P be a point
different from C , such that f ( P) P' and let m the perpendicular bisector of
the segment PP ' .
Since f is an isometry f (C ) f ( P) C ' P' CP' CP CP' .
Prepared by Begashaw M. 159
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
This means C is at equal distance from the two end point of the segment PP ' .
So, C is on line m (because if a point is at equal distance from the end points of
a line segment, then it is on the perpendicular bisector of the segment.). Thus,
S m (C ) C and S m ( P' ) P. Then, S m f (C ) S m (C ) C and
S m f ( P) S m ( P' ) P .
This shows that S m f fixes two different points C and P .
So, S m f i or S m f S n where n is the line
through the points C and P . But, S m f i because if S m f i , then f will
have more than one fixed point which contradicts from the hypothesis that f
has exactly one fixed point. Thus, S m f S n f S m S n where m and n
are lines intersecting at the point C . This theorem guarantees that whenever an
isometry has exactly one fixed point, it is a composition of two reflections by
lines intersecting at the fixed point of the isometry. In what follows we are
going to consider one of the fundamental theorems which is useful in the
classification of isometrics.
Prepared by Begashaw M. 160
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3.8.3 The Fundamental Theorems of Isometries
Theorem 3.14 (The First Fundamental Theorem of Isometries):
Every isometry can be expressed as a composition of three or fewer reflections.
Proof: The identity map is one of the isometrics which is the composition of
two reflections, any reflection with itself. That is i S l S l . Now, let f be any
isometry different from the identity. Then, there is some point P which is not
fixed by f . Let f ( P) Q, P Q . Let m be the perpendicular bisector of PQ .
Then, by definition of a reflection S m , S m (Q) P.
Then, Q f ( P) S m f ( P) P . So, P is the fixed point of S m f (P).
If S m f (P) has other fixed points besides P , then it must be either a reflection
S n which in turn implies S m f ( P) S n S m S m f S m S n f S m S n
which is a product of two reflections or the identity in which case
S m f i f S m which is a reflection.
On the other hand, if S m f (P) has only P as a fixed point, then by theorem
3.12, there are lines l and n intersecting at P such that
S m f S l S n S m S m f S m S l S n f S m S l S n which is the
composition of three reflections. Besides, f S m S l S n will be reflection or
glide reflection based on the relation of the lines. Thus, in any of the above
cases, any isometry can be written as a composition of one, two, or three
reflections. As we see from Theorem 3.14, reflection is the building blocks of
all isometries. That means every isometry can be expressed as a product of one,
two or three reflections.
Theorem 3.15: Any isometry of the plane is a reflection, rotation, translation
or a glide reflection.
Proof: Apply the above Theorems
Prepared by Begashaw M. 161
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Theorem 3.16 (The second Fundamental Theorem of Isometries):
If A, B, C are three non-collinear points, and A' , B' , C ' are also non-collinear
such that A' B' AC , A' C' AC , B' C' BC , then there exists a unique isometry
that maps A to A' , B to B' and C to C ' . Besides, this isometry can be
expressed as a product of at most three reflections.
Proof : This theorem has two parts to be proved: Existence and Uniqueness.
Existence: Construct line k as the perpendicular bisector of A and A' . Then
S k ( A) A' and S k ( B) P . If P B' , consider a line l which is the
perpendicular bisector of P and B' . Then, S l S k ( B) S l ( P) B'.
Claim I: S l S k ( A) A'. Here, S k ( B) P and S k ( A) A' which implies that
A' P AB . Besides from the hypothesis, B' A' BA. Combining these two
relations we get B' A' PA'. This implies that B' and P are equidistant from A'
which means A' is on the perpendicular bisector of P and B' . But this line is l
and hence A' is on l . Thus, S l S k ( A) S l ( A) A' and S l S k ( B) S l ( P) B'.
Now consider, S k (C ) R and S l S k (C ) S l ( R) Q. If Q C ' , the proof is
complete. However, suppose Q C '. Consider line m through A' and B'
Claim II: S m (Q) C '. From the hypothesis, A' C ' AC and from S k (C ) R ,
S k ( A) A' , we get A' R AC . Similarly, S l ( A' ) A' and S l ( R) Q gives
A' Q A' R.
Prepared by Begashaw M. 162
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
By combining these three equalities, we obtain A' Q A' C ' .
This implies that A' is equidistant from the end points of the segment
determined by Q and C ' . So A' is on the perpendicular bisector of the segment
C' Q . Similarly, from S l S k ( B) S l (S k ( B)) S l ( P) B' and
S l S k (C ) S l (S k (C )) S l ( R) Q , we have B' Q BC .
Again from the hypothesis B' C ' BC . Thus, B' Q B' C ' which implies that
B' is equidistant from the end points of the segment determined by Q and C ' .
This in turn results that B' is on the perpendicular bisector of the segment C' Q .
Hence, both A' and B' are on a line which is the perpendicular bisector of the
segment C' Q and this line is m so then S m (Q) C ' .
As a result,
S m S l S k ( A) S m S l ( A' ) S m ( A' ) A'
S m S l S k ( B) S m S l ( P) S m ( B' ) B'
S m S l S k (C ) S m S l ( R) S m (Q) C '
Therefore, there exists f S m S l S k which maps the three non- collinear
points A, B, C into three non-collinear points A' , B' , C ' .
The second part of the proof is to show that this isometry is unique. Since,
A, B, C are three non-collinear points, the uniqueness follows from Theorem
3.1. Thus, the proof is complete. Here, give attention on how to find such lines
m, l , k which satisfies the given conditions.
a) Line k is the perpendicular bisector of A and A'
b) Line l is the perpendicular bisector of B' and S k (B)
c) Line m is the perpendicular bisector of A' and B'
Example: Given ABC DEF where the vertices are
A (0,0), B (5,0), C (0,10) and D (4,2), E (1,2), F (12,4) .
Find equations of lines such that product of reflections on these lines takes
ABC to DEF .
Solution: Follow the procedure in the proof of the above theorem.
Prepared by Begashaw M. 163
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 3.8
1
1. Given the three lines l : x , m : x 2, n : x 5 . Find a line p parallel to these
2
7
lines such that Sn Sm Sl S p . Answer : p : x
2
2. Let m be the x-axis, n the y-axis and p the line y x . Find the line q such
that S n S m S p S q . Answer : q : y x
3. Let l : y 3x 20 and m : y 3x 10 . Then, find a translation T with
translator vector v such that S m S l Tv .
Answer : Tv ( x, y) ( x 18, y 6)
4. Given the lines M : y 0, n : y 2 x, x 0 . Find line q such that
a) S q S p S n S m
b) S q S p S m S n
Prepared by Begashaw M. 164
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3.9 Equations of Orthogonal Transformations in Coordinates
In the previous sections, we have studied about types of isometries and their
properties.
Now let’s summarize and generalize their equations using coordinate geometry.
Equation of Rotations: Equation of rotation about any center C (h, k ) with
angle of rotation is given by C , ( x, y) ( x' , y' ) where
x' ( x h) cos ( y k ) sin h
y ' ( x h) sin ( y k ) cos k
x' x cos y sin c
Expanding these equations gives
x' x sin y cos d
Now, by letting a cos , b sin , a 2 b 2 1 , these equations can be reduced
x' ax by c
to the form , a2 b2 1
y ' bx ay d
Here, if 2n, n Z , then
a cos , b sin a cos(2n) 1, b sin( 2n) 0 and the equations become
x' x c
which are the general equations of a translation.
y' y d
Otherwise, for 2n, a cos , b sin a cos( 2n) 1, b sin( 2n) 0 ,
the equations are equations of a rotation.
x' ax by c
Consequently; the equations , a 2 b 2 1 are the general
y ' bx ay d
equations of a translations and or rotations.
Prepared by Begashaw M. 165
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3.10 Equations of Even and Odd Isometries
Definition: Even isometry is an isometry that can be expressed as a product of
even number of reflections. On the other hand, an isometry that can be
expressed as a product of odd number of reflections is said to be odd isometry.
Translation and rotations are even isometries while reflection and glide
reflection are odd isometries.
3.10.1 Equations of Even Isometries
We discussed above that only translations and rotations are isometries that can
be expressed as a product of even number of reflections. Thus, for any even
x' ax by c
isometry, its equations are of the form , a 2 b 2 1.
y ' bx ay d
3.10.2 Equations of Odd Isometries
Any odd isometry is a product of odd numbers of reflections. In other words,
any odd isometry is a product of even isometry followed by a reflection on any
line l . Now, let be any odd isometry. Then, for any line l , S l becomes
even isometry where as S l S l is an odd isometry. For any point ( x, y ) ,
taking the line l to be the x axis (in particular, since it works for all), we get
S l ( x, y) ( x' , y' ) ( x, y) . Hence, equation of the odd isometry becomes,
( x' , y ' ) ( x, y ) S l ( S l )( x, y )
S l (ax by c, bx ay d )
(ax by c,[bx ay d ])
Thus, the general equations of any odd isometry becomes ( x, y) ( x' , y' )
x' ax by c
where , a2 b2 1
y ' [bx ay d ]
Since no isometry is both even and odd, the above two equations (equations of
even and odd isometries) constitute the equations for any types of isometries.
Prepared by Begashaw M. 166
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Theorem 3.17 (The generalized equation Theorem for Isometries):
Let ( x, y ) be any point. Then, the general equations for any isometry is given
x' ax by c
by ( x, y) ( x' , y' ) where , a2 b2 1
y ' [bx ay d ]
The plus sign is used when we assume that is an even isometry and the minus
sign is applied when we assume that is an odd isometry. Conversely, any
equations of this form are equations of isometries.
Examples:
1. Let be an isometry whose equations are given by ( x, y) ( x' , y' ) where
x' 53 x 54 y 4
y' 5 x 5 y 2
4 3
Determine whether is an even or odd isometry.
Solution:
x' ax by c
From equations of isometries, we know that , a2 b2 1
y ' [bx ay d ]
Equate the general equations and the given values of x' and y ' . This gives,
x' ax by c 53 x 54 y 4
y ' [bx ay d ] 5 x 5 y 2
4 3
So, x' ax by c 53 x 54 y 4 a 53 , b 54 , c 4
Using these values, we obtain d from the equations of y ' as follow.
y' [bx ay d ] [ 54 x 53 y d ] 54 x 53 y 2
To solve these equations we have two options either to use the plus sign or the
minus sign. If we use the plus sign the equation becomes
54 x 53 y d 54 x 53 y 2 54 54 , 53 53 , d 2 (by equating
coefficients) which is absurd (impossible!). If we use the minus sign the
equation becomes 4
5 x 53 y d 54 x 53 y 2 54 54 , 53 53 , d 2 (by equating
coefficients)
which is the only logical option.
Prepared by Begashaw M. 167
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Now, using the values a 53 , b 54 , c 4, d 2 ,
x' 53 x 54 y 4 ax by c
we get
y ' 5 x 5 y 2 [bx ay d ]
4 3
But from our previous discussions, these are equations of odd isometry and
hence is odd.
2. Let be an isometry whose equations are given by ( x, y) ( x' , y' ) where
x' 53 x 54 y 54
y' 5 x 5 y 5
4 3 2
Show that is an even isometry and conclude that is a rotation and find the
center of the rotation.
Solution:
An isometry is even if and only if its equations are given by
x' ax by c
, a2 b2 1
y ' bx ay d
In our case, x' ax by c 53 x 54 y 54 a 53 , b 54 , c 4
5
and
y' bx ay d 54 x 53 y 52 d 52 .
x' 53 x 54 y 54 ax by c
Hence, the equations of become
y ' 5 x 5 y 5 bx ay d
4 3 2
But, these are the forms of an even isometry and thus is even. Since the only
even isometries are rotation and translation, must be arotation because as
we see from its general equation, cannot be a translation. Now, if is a
rotation withcenter C (h, k ) , it must satisfy the equations
x' ax by c x cos y sin c
y' bx ay d x sin y cos d
From which we can get the equations to determine the intersection point
h(1 cos ) k sin c
C (h, k ) which is given as
h sin k (1 cos ) d
Thus, h and k can be solved as follow.
h(1 53 ) k ( 54 ) 54 2h k 1 k 1 2h
h( 54 ) k (1 53 ) 52 2h 4k 1
2h 4(1 2h) 1
h 12 , k 1 2h k 0
C (h, k ) ( 12 ,0)
Prepared by Begashaw M. 168
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Alternatively, one can also solve the values of h and k from the equation that
we derived earlier using a cos 53 , b sin 54 , r 54 , t 52 in the
equations we will get the same results.
r sin 4 4
2 1 1
h t 5 52 5
2 2(1 cos ) 2 2(1 53 ) 5 10 2
t sin 2 4 4
1 1
k r 5 5
0
2 2(1 cos ) 2 5 2(1 53 ) 5 5
(h, k ) ( 12 ,0)
Hence, we conclude that is a rotation which is the product of two
reflections on lines intersecting at the point ( 12 ,0).
3. Let be an isometry with (0,0) (2,1), (1,1) (1,0), (2,3) (5,1) .Then
find the equations of .
Solution: Let A (0,0), B (1,1), C (2,3), A' (2,1), B' (1,0), C' (5,1).
Clearly, A, B, C are non-collinear and so are A' , B' , C ' . But we know that there
is a unique isometry that takes three non-collinear points into three non
collinear points. Now, let this isometry be given by ( x, y) ( x' , y' ) where
x' ax by c
such that the constants a, b, c, d are to be determined from the
y ' bx ay d
given points and their images.
So, ( A) (0,0) (2,1) a.0 b.0 c 2,b.0 a.0 d 1 c 2, d 1.
Using these values c 2, d 1, the equation of becomes ( x, y) ( x' , y' )
x' ax by 2
where
y ' bx ay 1
a b 2 1
Again, ( B) (1,1) (1,0) b 1, a 0
b a 1 0
x' y 2
Hence, ( x, y) ( x' , y' ) where .
y' x 1
If we use the minus sign, in the general equation we get the same result.
x' 3x 4 y 4
4. Let ( x, y) ( x' , y' ) where .
y' 4 x 3 y 2
Verify that is not an isometry.
Solution: If the equations of represent equations of an isometry, it must
satisfy the following equations:
x' ax by c
( x, y) ( x' , y' ) where , a2 b2 1
y ' [bx ay d ]
Prepared by Begashaw M. 169
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
But, x' ax by c 3x 4 y 4 a 3, b 4, c 4 a 2 b 2 25 1 .
Hence, the equations of does not represent equations of an isometry and thus
is not an isometry.
Theorem 3.18: The product of any two even or any two odd isometries is even
and the product of an even and odd isometries is odd isometry.
Proof:
Let and be any two even isometries given by the following equations.
x' ax by c
( x, y) ( x' , y' ) where , a2 b2 1
y ' bx ay d
x' mx ny p
( x, y) ( x' , y' ) where , m2 n2 1
y ' nx my q
We need to show that the product is also a direct similarity.
For any point ( x, y) , ( x, y) ( ( x, y)) ( x' ' , y' ' ) where
x' ' mx'ny ' p
m(ax by c) n(bx ay d ) p
(ma nb) x (mb na) y mc nd p
tx ly r , with t ma nb, l mb na, r mc nd p
y ' ' nx' my' q
n(ax by c) m(bx ay d ) q
(na mb) x (ma nb) y nc md q
lx ty s, with s nc md q
Now, these equations to be equations of isometries, we are left to show
that t 2 l 2 1. But,
t 2 l 2 (ma nb) 2 (mb na) 2
(ma) 2 2abmn (nb) 2 (mb) 2 2abmn (na ) 2
(ma) 2 (na) 2 (mb) 2 (nb) 2
m 2 (a 2 b 2 ) n 2 (a 2 b 2 )
(m 2 n 2 )(a 2 b 2 ) 1, sin ce a 2 b 2 1, m 2 n 2 1
t 2 l 2 1.
x' tx ly r 2
Hence, ( x, y) ( ( x, y)) ( x' , y' ) where , t l2 1
y ' lx ty s
But these are equations of an isometry obtained by applying the plus sign in the
general equation of isometries (Theorem 3.17).
Hence, is an even isometry whenever and are even isometries. The
other part follows similarly.
Prepared by Begashaw M. 170
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
3.11 Test for Type of Isometries
From the observation on the type of isometries and their effects on orientation,
let’s develop a simple test for the type of isometry from the image of three
non-collinear points. Given three non-collinear points P, Q, R and their images
P' ( P), Q' (Q), R' ( R) . We need to determine whether is a
translation, rotation, glide reflection or reflection. Since an isometry maps any
three non-collinear points in to non-collinear points, P' , Q' , R' are non-collinear
points. Now, take PQR and P' Q' R' and determine their orientation. We know
that the orientation of PQR is determined from det PQ, PR and that of
P' Q' R' from det P' Q', P' R' .
Test-I: If PQR and P' Q' R' have the same orientation, then is orientation
preserving isometry. Thus, is either a translation or a rotation (Because
translation and rotation are the only types of orientation preserving isometries).
Furthermore, to determine whether is a translation or a rotation, find the
vectors PP ', QQ', RR ' . If PP' QQ' RR' , then is a translation, but if they
are different, is a rotation.
Test-II: If PQR and P' Q' R' have different orientation, then is orientation
reversing (changing) isometry. Thus, is either a glide-reflection or a
reflection (Because glide-reflection and reflection are the only types of
orientation reversing (changing) isometries). Furthermore, to determine
whether is a reflection or glid-reflection, find the equation of the reflecting
line using any of the points and its respective image.
If all the points are reflected on the same line, then is a reflection,
otherwise it is glide-rflection.
Prepared by Begashaw M. 171
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Example: Suppose is an isometry which maps PQR into P' Q' R' where the
vertices of the triangles are
4 1 12 1 3 5
P , Q , R , P' , Q' , R' .
2 2 4 3 0 11
Determine whether is a translation, rotation, glide reflection or reflection
and find its equation. Based on the type of , find the translator vector, center
and angle of rotation or line of reflection.
Solution: To apply the above test, first determine the orientation of PQR and
P' Q' R' .The orientation of PQR is determined from det PQ, PR where
det PQ, PR 3 8
4 6
50 0 .Hence, PQR has positive orientation.
Similarly, the orientation of P' Q' R' is determined from det P' Q', P' R' where
det P' Q', P' R'
4 6
3 8
50 0 which shows that P' Q' R' also has positive
orientation. So, we have got that PQR and P' Q' R' have the same orientation.
As a result, is orientation preserving isometry. Therefore, is either a
translation or a rotation. Now, to determine whether is a translation or a
rotation, find the vectors PP ', QQ', RR ' .
5 2 7
Here, PP ' , QQ' , RR ' . But PP ' QQ' RR '.
1 2 15
So, is not a translation. As a result, the only option for to be is a rotation.
Now, determine the equation of . We know that rotation is an even isometry.
So, we can determine the equation of from the general equations of even
x x' x' ax by c
isometries where , a2 b2 1.
y y' y ' bx ay d
4a 2b c 1 a 2b c 3
So, using ( P) P' , (Q) Q' , we get ;
2a 4b d 3 2a b d 0
Prepared by Begashaw M. 172
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Collecting equations with like terms gives
4a 2b c 1
3a 4b 4......................................(i )
a 2b c 3
2a 4b d 3
4a 3b 3..........................................(ii )
2a b d 0
Multiplying equation (i ) by 3 and equation (ii ) by 4 yields
9a 12b 12
25a 0 a 0, b 1
16a 12b 12
Substituting these values of a and b gives c 1, d 1 .
x x' x' y 1
Therefore, the equation of becomes where .
y y' y' x 1
Since is a rotation, our last task is to find the center and angle of the rotation.
But, in the general equation of isometry, we know that a cos , b sin
where is the angle of rotation. Thus, cos 0 , sin 1 2 . Besides, the
center of rotation C (h, k ) is given by
c d 1 1 d r 1 1
h
1, k
0.
2 2 tan 2 2 2 2 2 tan 2 2 2
As a result, is a rotation with center C (1,0) and angle of rotation / 2 .
Prepared by Begashaw M. 173
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 3.9
1. Suppose RC , is a counterclockwise rotation with center C (h, k ) whose
x' 10 y
equations are given by . Find the angle and center of this rotation.
y ' x 30
Answer : (h, k ) (20,10)
2. If 2 x' 3x y 14 4 3 and 2 y' x 3 y 8 6 3 are equations of a
CCW rotation, then find the center and angle of the rotation.
Answer :C (4,6), 30
x 2ax 4 y 7
3. Suppose 4 5
is an odd isometry.Then, find the value (s) of
y 5 x 5 y 5
3
the constant a . Answer : a 3 / 10
x 3 px 5
y
4. If 5 13
is an isometry, find the value (s) of the constant p .
y 13 x 13 y
12
Answer : p 4 / 13
3x 4 y 4x 3y
5. If x' and y ' are equations for a reflection S m , then find
5 5 5 5
equation of line m . Answer : m : x 2 y
6. Suppose is an isometry that maps the three non-collinear points
A (1,1), B (2,1), C (1,5) into the points A' (1,1), B' (4,1), C' (1,3) .
Determine whether is a translation, rotation, glide reflection or reflection
and find its equation.
5 12 31 12 5 38
7. Suppose g ( x, y) ( x y , x y ) is the equations of a
13 13 13 13 13 13
glide reflection g . Answer : l : 2 x 3 y 2 0
8. Suppose ( x, y) ( x' , y' ) where x' ax by c and y' bx ay d where
a 2 b2 1. Show that is an odd isometry.
Prepared by Begashaw M. 174
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Review Problems On Chapter-3
1. Given the lines l : x y 6 0, m : x y 4 0 . Find the center P of a half
turn H P such that Sm Sl H P . Answer : P (5,1)
2. Given a line m : 2 x y 3 0 . Find two possible equations for a line l such
that S m (l ) l with P (1,5) l . Answer : l :2x y 3 0 Or l : x 2 y 11 0
3. Let l : 2 x 3 y 1 0 and m : 4 x 6 y 5 0. Then find the image of the point
(1,3) by the composite reflection S m S l
4. Let l : y 2 3x 1 and m : y 5 x 2 . Show that for any point ( x, y) ,
3 3 3
x 3 4 3
x' y
S m S l ( x, y) ( x' , y' ) where 2 2 13
y' 3 x 1 y 4 3
2 2 13
5. Let l : y ax, a 0 and m : y bx, b 0. If the angle measured from l to m
x' x cos 2 y sin 2
is , then S m S l ( x, y) ( x' , y' ) where y ' x sin 2 y cos 2
6. Let l : ax y 0 and m : cx y 0 . If ac 1 , calculate S m S l (10,12) .
Answer : (10,12)
7. True or False? If f S n S m and f S l S t , then n l and m t .
Answer: False
8. Let l : y 3x 20 and m : y 3x 10 . Then, find a translation T with
translator vector v such that S m S l Tv . Answer : v (9,3)
9. Given two lines l and m intersecting at (1,2) . Suppose S m S l (3,5) (2,4) .
Then find the acute angle measured from l to m . Answer :
4
10. Let m and n be lines intersecting at C (2,1). If S m S n RC , , then find the
3
equation of line n when m : y x 1.
Prepared by Begashaw M. 175
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
11. Let g be a glide reflection with axis l : y x 1 and glide vector v (1,1)
Then, find the image of the point (3,2) . Answer : (2,5)
12. Let g be a glide reflection with axis l : x y 6 0 and glide vector
v (2,2) . Then, find the general equation of g and calculate the image of the
point (0,0) . Answer : g ( x, y) ( y 4, x 8)
13. Let g be a glide reflection with axis l : x y 3 0 and glide vector
v (4,4) . Then, find the general equation of g and calculate the image of the
point (0,0) . Answer : g ( x, y) ( y 1, x 7), g (0,0) (1,7)
14*. Let m and n be parallel lines 13cm a part. A point A is 4cm and 17cm
from line m and n respectively. Suppose A is reflected across line m , and then
its image A' is reflected across line n to create a second image A' ' .
Answer : d 26cm
a) Draw the diagram showing the position of A, A' , A' ' and find the distance
between A and A' ' . b) If the order of reflection is changed, repeat the
problem of part (a).
How do the answers to part (a) and (b) differ?
15. Suppose m and n are two distinct lines. Then, show that
S n S m ( P) P P m n .
16. Let l and m be any two lines. Prove that S l S m S m S l l m
17. Let m and n be lines intersecting at P . Show that if m and n are fixed
lines under an isometry , then ( P) P .
18. Given two lines l and m intersecting at a point (1,2) and l : y x 1 .
Suppose Rm Rl (3,5) (2,4) . Find the acute angle measured from l to m .
19. If k // l and k m , then show that S m (S l S k ) (S l S k ) S m .
20*. Let , , be interior angles at the vertices A, B, C respectively of ABC
(oriented counter clockwise). Show that R A, 2 RB, 2 RC , 2 i.
Prepared by Begashaw M. 176
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
21. Show that non-identity rotations with different centers do not commute.
22. What must be true of two rotations R A, and RB , if their product R A, RB,
is a translation?
23. Show that the composition of any two rotations is either a rotation or a
translation.
24. Let R1 and R2 be rotations. Then show that R1 R2 R11 R21 is a translation.
25. Prove that
a) The set of all isometries forms a transformation group.
b) The set of all rotation with fixed center forms abelian group of
transformations.
26. Suppose l intersects the lines m and n so that alternate interior angles are
congruent. Then, using the concept of isometry, prove that m and n are
parallel.
27. Show that the set of all rotations, all reflections, or all half turns does not
form a transformation group. State the reasons clearly in each case.
28. Suppose the lines l and m passes through the origin in R 2 make angles
and respectively with the positive x axis. Show that Rm Rl R2( ) . Where
Rm Rl is a reflection on l followed by a reflection on m and R2 ( ) is a rotation
by angle 2( ) about the origin.
Prepared by Begashaw M. 177
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
CHAPTER-4
SIMILARITY TRANSFORMATIONS
4.1 Introduction
Consider a right angle triangle ABC with right angle at C and hypotenuse AB
as in the figure 4.1.
If the altitude CD is constructed, we get three similar triangles
ADC, ACB, CDB . Then, using similarity of these triangles we see that there is
proportion of the segments of the hypotenuse AD, DB and the altitude CD
AD DC AC
given by .
CD DB CB
If we assign the lengths of these segments as a, b, h in the figure, we get
a h
h 2 ab .
h b
In this explanation, even though the lengths, areas in general sizes of the figure
are different we see that they do have the same shape and the ratio of their sides
remains constant as well.
Such geometric figures which have the same shape and their sides proportional
are said to be similar figures and a transformation which maps one figure in to
a similar figure is known as similarity transformation.
Prepared by Begashaw M. 178
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Definition: A transformation of the plane is said to be a similarity
transformation if there exists a positive number k such that for all points A and
B of the plane and their images A' and B' , we have A' B' kAB. That is
( A) ( B) A' B' kAB . The constant k is called the coefficient or factor or
ratio of the similarity.
Examples: Determine whether the following transformations are similarity
transformation or not.
a) : R 2 R 2 given by ( x, y) (2 x 5,2 y 6)
b) : R 2 R 2 given by ( x, y) ( x 1, y 3)
c) : R 2 R 2 given by ( x, y) (2 x 3,3 y 5)
Solution:
a) Clearly, is a transformation. Now, let A (a, b) and B (c, d ) be any two
points in the plane R 2 . Then,
( A) A' (2a 5,2b 6), ( B) B' (2c 5,2d 6)
A' B' B' A' (2c 2a,2d 2b)
A' B' 2 AB A' B' kAB, for k 2
Hence, is a similarity transformation with ratio k 2.
b) Similar calculation here also yields A' B' AB A' B' kAB, for k 1.
So, is a similarity transformation with ratio k 1.
c) In this case, A' B' (3c 3a,2d 2b), AB (c a, d b) A' B' kAB, k R .
It means there is no positive constant k for which A' B' kAB. So, is not a
similarity transformation.
Prepared by Begashaw M. 179
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
4.2 Properties of Similarity Transformations
Proposition 4.1: The product of any two similarities is again a similarity.
Proof: Let and be any two similarities with ratios k and t respectively.
For any two points, A and B , let ( A) A' , ( B) B' .
Then, ( A) ( ( A)) ( A' ) A' ' and ( B) ( ( B)) ( B' ) B' '
Since is a similarity with ratio t , it follows that A" B" tA' B'.
Again, as is a similarity with ratio k , it follows that A' B' kAB.
So, A" B" tA' B' t (kAB) tkAB. Thus, is a similarity with ratio tk .
Proposition 4.2: The inverse of a similarity is again a similarity.
Proof: Let be a similarity with ratio k .Then, for any two points, A and B ,
A' B' kAB where ( A) A' , ( B) B' . But, is a transformation (Definition
of similarity) implies ( A) A' , ( B) B' 1 ( A' ) A, 1 ( B' ) B. Thus,
1 1
A' B' kAB A' B' k 1 ( A' ) 1 ( B' ) 1 ( A' ) 1 ( B' ) A' B' AB A' B'.
k k
1
But for any positive constant k , is also defined and positive. Hence, 1 is a
k
1
similarity transformation with ratio whenever is a similarity with ratio k .
k
Therefore, for any similarity with ratio k , A' B' kAB if and only if
1 1
AB A' B' if and only if 1 is a similarity transformation with ratio .
k k
Proposition 4.3: A similarity maps triangles in to similar triangles.
Proof: Let be a similarity with ratio k and let A' B' C ' be the image of
ABC under . We need to show that A' B' C' ~ ABC , where
( A) A' , ( B) B' , (C) C'.
Prepared by Begashaw M. 180
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Since is a similarity with ratio k , it follows that
A' B'
A' B' kAB k
AB
A' C ' A' B' A' C ' B' C '
A' C ' kAC k k.
AC AB AC BC
B' C '
B' C ' kBC k
BC
This shows that the three sides of the two triangles are proportional. Hence, by
Side-Side-Side similarity theorem, we get A' B' C' ~ ABC .
Here, A' B' C' ~ ABC in turn enables us to infer that the three interior angles of
ABC are congruent to the three interior angles of A' B' C ' (Because
corresponding angles of similar triangles are congruent).
So, the three interior angles of ABC are preserved and this will enable us to
state the following immediate corollary.
Proposition 4.4: Any similarity preserves angles.
Proof: Immediate from proposition 4.3.
Proposition 4.5: Any similarity preserves co-linearity, between ness and mid
point.
Proof: Let be a similarity with ratio k and let A, B, C be three collinear
points where B is between A and C . Then, AB BC AC since B is between
A and C . Now, whenever ( A) A' , ( B) B' , (C ) C' , we have that
A' B' kAB, B' C' kBC, A' C' kAC. As a result,
A' B' B' C' kAB kBC k ( AB BC ) kAC A' C' A' B' B' C' A' C'
But, this is true if and only if the image points A' , B' , C ' are collinear and B' is
between A' and C '. Besides, if B is the mid point of A and C , then
AB BC A' B' kAB kBC B' C' which implies B' is the mid point of A' and
C '. This result tells us that the images of any three or more collinear points are
again collinear. If we consider, the proposition in detail, it further implies that
any similarity is a collineation.
Prepared by Begashaw M. 181
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Because if A' , B' are the images of the points A, B on a line l ,then for any
other point P on l , P' is on a line determined by the image points A' , B' (As
similarity preserves co linearity and between ness). Thus, one can use those
ideas together in order to conclude that the image of a line under a similarity is
again a line.
Theorem 4.1: A similarity with at least two fixed points is an isometry.
Proof: Let be a similarity with ratio k such that ( A) A, ( B) B . By
definition of similarity, for any two distinct points A and B , A' B' kAB . Then,
it follows that A' B' kAB AB kAB k 1 .This shows that for any other
points, P and Q , ( P) P' , (Q) Q' P' Q' PQ . (Any given similarity has
exactly one ratio). Hence, is an isometry.
Corollary 4.1: A similarity with ratio of k 1 is an isometry.
4.3 Common Types of Similarity Transformations
4.3.1 Isometries
Since for any isometry f , f ( A) f ( B) AB for every pair of points A and B , it
shows that every isometry is a similarity transformation.
a) Translations
b) Reflections
c) Rotations
d) Glide reflections
All these isometries transformations are similarity transformations.
Prepared by Begashaw M. 182
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
4.3.2 Homothety (Homothetic Transformations)
Definition: A transformation is said to be homothety transformation if and
only if for a scalar k 0,1 and a fixed point C it maps every point X to X ' such
that X ' C k XC .
Here, the fixed point C is called center of the homothety and the scalar k 0,1
is called ratio or factor of the homothety.
Such a transformation is denoted by H C ,k . Thus,
H C ,k ( X )C k XC C H C ,k ( X ) kC kX
H C ,k ( X ) kX (1 k )C
Therefore, for any point X , H C ,k ( X ) kX (1 k )C .
If H C ,k : R 2 R 2 and C (a, b) R 2 , then H C ,k is given by H C ,k ( x, y) ( x' , y' )
x' kx (1 k )a
where
y ' ky (1 k )b
Hence, all such types of transformations are named as homothety and they are
similarity transformations. Such type of similarity transformations are usually
called homothetic similarities.
Examples:
1. Find the equation of a homothety with center (2,4) and factor k 2 and
calculate the image of the point (3,5).
Solution:
x' kx (1 k )a
For any point ( x, y) , H C ,k ( x, y) ( x' , y' ) where
y ' ky (1 k )b
x' 2 x (1 2)2 2 x 2
H C , 2 ( x, y) ( x' , y' ) where
y ' 2 y (1 2)4 2 y 4
H C , 2 ( x, y) (2 x 2,2 y 4) . In particular, H C , 2 (3,5) (4,14)
Prepared by Begashaw M. 183
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
2. A homothety takes the point (1,1) to (1,5) and the point (0,2) to (4,8). Find
the equation of this homothety.
Solution: Let H C ,k be a homothety with center C (a, b) and factor k . Then,
H C ,k ( X ) kX (1 k )C H C ,k ( x, y ) k ( x, y ) (1 k )(a, b)
H C ,k (1,1) k (1,1) (1 k )(a, b) (1,5) (k a ka, k b kb) (1,5)
H C ,k (0,2) k (0,2) (1 k )(a, b) (4,8) (a ak ,2k b bk ) (4,8)
k a ka 1 k b kb 5
,
a ka 4 2k b kb 8
k 3, a 2, b 1 C (a, b) (2,1)
Hence, H C ,3 ( x, y) (3x 4,3 y 2).
Proposition 4.6: Every homothety is a similarity transformation.
Proof: Let H C ,k be a homothety. Then, for any two points A and B ,
H C ,k ( A) A' kA (1 k )C , H C ,k ( B) B' kB (1 k )C
A' B' B' A' kB kA kAB
A' B' kAB
So, A' B' kAB for any two points A and B where H C ,k ( A) A' , H C ,k ( B) B' .
Hence, any homothety is a similarity.
4.4 Representation of Similarity Transformations
Theorem 4.2: Every similarity is the product of an isometry and a homothety.
Proof: Let be a similarity with ratio k and consider a homothety H C , with 1
k
1
center C and factor . From proposition 4.1 and 4.2, H C , 1 is a similarity
k k
1
with ratio .k 1 . But, a similarity with ratio of 1 is an isometry. Thus, let this
k
isometry be H C , 1 . So,
k
H C , 1 H C ,k H C , 1 H C ,k
k k
i H C ,k , H C , 1 H C ,k i
k
H C ,k
This means that any similarity with ratio k is the product of a homothety with
ratio k and any isometry.
Prepared by Begashaw M. 184
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Definition: A transformation that maps each vector in to a vector parallel to
itself is known as dilation. That means is a dilation if and only if for all A
and B , A' B' k AB where A' ( A), B' ( B).
Example: Show that : R 2 R 2 given by ( x, y) (3x 4,3 y 2) is dilation.
Solution: Let A (a, b), B (c, d ) be arbitrary points. Then,
A' ( A) (3a 4,3b 2), B' ( B) (3c 4,3d 2)
A' B' (3c 3a,3d 3b) 3(c a, d b) 3 AB A' B' // AB
Hence, the given transformation is dilation.
Proposition 4.7: Every dilation is a similarity.
Proof: Let be dilation with factor k 0 .
Then, A' B' k AB , for all points A, B . But from vector analysis, for any two
vectors v and u , v ku v k u for any scalar k .
Thus, A' B' k AB A' B' k AB A' B' k AB A' B' k AB.
But, A' B' k AB is a condition for similarity if k 0. Hence, is a similarity.
Be careful! The converse of this theorem need not be true.
Let : R 2 R 2 be a transformation given by ( x, y) (3 y, x 1). Take any
two points. A (a, b) and B (c, d ) . Then,
( A) A' (3 b, a 1), ( B) B' (3 d , c 1)
A' B' A' B' (b d , c a) (d b, c a) AB AB A' B' k AB, with k 1.
Thus, is a similarity. On the other hand,
A' B' (b d , c a), AB (d b, c a) A' B' k AB, k R .
This means A' B' and AB are not parallel. Thus, is not a dilation.
Hence, from this observation is a similarity but not a dilation which shows
that the converse of the theorem is not true.
Prepared by Begashaw M. 185
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Theorem 4.4 (The Classification Theorem ):
This theorem is useful in classifying whether a given dilation is a translation or
a homothety. That is why it is named as classification theorem.
A' B'
Let : W W be a dilation. Then, k (constant) for all A, B in W .
AB
Furthermore, if k 1, then is a homothety with ratio of k itself and if k 1,
then is a translation.
Proof: Assume that W is not a line. Let be a dilation such that
( A) C, ( B) D as shown in figure 4.3.
Now, suppose the points A, B, C, D are not collinear. Take any point X not
on A, B with ( X ) Y . As is a dilation, AB CD, AX CY and BX DY .
CD CY DY
Thus, by Desagrues Theorem, k . Since X is arbitrary,
AB AX BX
A' B'
k for all points A, B in W . Furthermore, if k 1 (refer figure 4.3), then
AB
CD k AB, CY k AX , DY k BX still by Desagrues Theorem, the lines
A, C , B, D and X , Y are concurrent at a point O with
OC CY OY
k . As a result,
OA AX OX
OY
k OY k OX
OX
O ( X ) k OX
( X ) O kX kO
( X ) kX (1 k )O
Prepared by Begashaw M. 186
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
But, ( X ) kX (1 k )O is an equation of a homothety with center O and
factor k . On the other hand if k 1 (refer figure 4.3), then
CD AB, CY AX , DY BX . But,
CY AX XY AC
X ( X ) AC
(X ) X C A
( X ) X AC
Thus, ( X ) X b for any point X where b AC which shows that is a
translation.
Corollary 4.2: Every dilation is either a translation or a homothety.
Proof : ( Use the classification theorem)
Example: A dilation takes the point (0,20) to (1,12) and the point (10,5) to
(3,15). Determine whether this dilation is a translation or a homothety and
write its equation.
A' B'
Solution: Let be a dilation. Then, by the classification theorem, k , for
AB
all A, B . In particular, it is true for A (0,20) and B (10,5) .
A' B' B' A' (3,15) (1,12) (2,3) (2,3) 1
Thus, k 15 k 15 1.
AB B A (10,5) (0,20) (10,15) (2,3) 5
Hence, is a homothety with factor k 15 by the classification theorem.
So, ( x, y) 15 ( x, y) (1 15 )C 15 ( x, y) 54 C, C (a, b). This is true for all ( x, y).
In particular, it is true for (0,20) , so
(0,20) (1,12) 15 (0,20) 54 C (1,12)
54 C (1,12) (0,4) (1,16)
C 54 (1,16) ( 54 ,20)
Therefore, the equation of is given by ( x, y) ( 15 x 1, 15 y 16).
Prepared by Begashaw M. 187
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Corollary 4.3: Every dilation is a product of a homothety and an isometry.
Proof: The proof of this theorem follows from proposition 4.7 ( every dilation
is a similarity)and from theorem 4.2, every similarity is a product of a
homothety and an isometry. Hence, every dilation is a product of a homothety
and an isometry. This theorem explains that for any dilation , there exist a
homothety H C ,k , and an isometry such that H C ,k .
Theorem 4.5: If A' B' // AB, then there is a unique dilation for which
( A) A' , ( B) B'.
Proof: Left as an exercise.
From this theorem, we are assured that the dilation that takes A, B to A' , B'
respectively exists and is unique whenever A' B' // AB . But, the main point is not
only to assure its existence but also how to find its equation. Now, the main
question is for a given similarity , how can we find the homothety H C ,k , and
the isometry such that H C ,k . To find H C ,k and explicitly, from
H C ,k , let’s proceed case by case as follow:
Case I: When the similarity is dilation:
Suppose is a dilation where ( A) A' , ( B) B' such that A' B' // AB (The
line through A' , B' is parallel to the line through A, B ). Thus, by theorem 4.5,
there is a unique dilation that takes A to A' and B to B' .To find such dilation
follow the following procedures:
Step-1: Find a translation Tv that takes A to A' and calculate the image of B
under Tv . say Tv ( B) P
Step-2: Write the formula for a homothety with center A' and factor k , say
H A',k .
Step-3: Solve the equation H A',k ( P) B' for k .
Step-4: Write the equations of H A',k and Tv using any object point ( x, y) .
Hence, these are the required homothety and isometry for which H A',k .
Prepared by Begashaw M. 188
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Example: Let be a dilation where (1,2) (2,5) and (2,4) (5,11). Find a
homothety H C ,k and an isometry such that H C ,k .
Solution: From the given, A (1,2), B (2,4), A' (2,5), B' (5,11)
Clearly, A' B' // AB, because the line m through A' , B' and the line n through
A, B have the same slope 2. So, to find the decompositions of as a
homothety H C ,k and an isometry such that H C ,k , proceed as follow
using the above steps.
Step-1: Let’s find a translation Tv that takes A to A' and calculate the image of
B under Tv . If Tv is a translation that takes A to A' , then v A' A (1,3) .
Hence, for any point ( x, y) , Tv ( x, y) ( x 1, y 3) and P Tv ( B) Tv (2,4) (3,7) .
Step-2: A homothety with center A' and factor k is given by
H A',k ( x, y) (kx 2k 2, ky 5k 5) .
Step-3: Equating H A',k ( P) B' , yields H A',k (3,7) (k 2,2k 5) (5,11) k 3.
Step-4: For any object point ( x, y) ,
H A',3 ( x, y ) (3x 4,3 y 10)
( x, y) Tv ( x, y) ( x 1, y 3)
Hence, these are the required homothety and isometry for which H C ,k .
Case II: When the similarity is not dilation:
Note that from the example above, we could not generalize that the isometry
part is always a translation rather we must investigate the case when A' B'
and AB are not parallel so as to make generalization. This follows with similar
procedures as above with a little modification and it is left to the readers to
develop the procedures.
Prepared by Begashaw M. 189
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
4. 5 Equations of Similarity Transformations in Coordinates
So far we discussed about the classes of similarities. Now, we are going to see
the equation of similarities in details using coordinates. First let’s consider
special cases which help us to drive the general equations of similarities.
I) Equations of isometric similarities
We have already discussed that the equations of any isometry is given by
x' ( x h) cos ( y k ) sin h
( x, y) ( x' , y' ) where
y' [( x h) sin ( y k ) cos k ]
x' x cos y sin r
Expanding the equations gives where
y ' [ x sin y cos t ]
r h(1 cos ) k sin
t h sin k (1 cos )
x x' x' ax by c
This can be generalized as where , a 2 b 2 1.
y y' y' (bx ay d )
Since any isometry is a similarity, this is one form of equations of similarities.
II) Equations of Homothetic Similarities
The equation of a any homothety with factor r and center C (h, k ) is
x' rx (1 r )h
H C ,r ( x, y) ( x' , y' ) where
y ' ry (1 r )k
III) General Equations of Similarities
The general equations of similarities are derived from the above results using
the fact that any similarity is a homothety followed by an isometry. That is for
any similarity
, H C ,r where H C ,r is a homothety with ratio r , center C and is an
isometry. Thus, using the results in I and II above, we get
x' xr cos yr sin c
( x, y) H C ,r ( x, y) ( x' , y' ) where
y' [ xr sin yr cos d ]
Here c and d are real numbers.
Now, by letting a r cos , b r sin , a 2 b 2 r 2 0 , for r 0 , we state the
general equations of similarities as follows.
Prepared by Begashaw M. 190
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Theorem 4.6[The Generalized Equation Theorem for Similarities]:
Any similarity in a plane has equations of the form ( x, y) ( x' , y' ) where
x' ax by c
, a2 b2 0
y ' [bx ay d ]
Conversely, any equations of this form are equations of a similarity.
4.6 Direct and Opposite Similarities
i) Direct Similarities
Definition: A similarity is said to be direct similarity if and only if is the
product of a homothety about any point followed by an even isometry. Thus,
from the equation of even isometry, the equation of a direct similarity
x' ax by c
becomes , a2 b2 0
y ' bx ay d
ii) Opposite Similarities
Definition: A similarity is said to be opposite similarity if and only if is
the product of a homothety about any point followed by an odd isometry.
Thus, from the equation of odd isometry, the equation of an odd similarity
x' ax by c
becomes , a2 b2 0
y ' bx ay d
Therefore, in the general equation of similarity we discussed above, the plus
sign applies to direct similarities where as the minus sign is taken for opposite
similarities.
Note that a similarity is either direct or opposite but can not be both.
Theorem 4.7: The product of any two direct or any two opposite similarities is
direct similarity while the product of a direct and opposite similarities is
opposite similarity.
Prepared by Begashaw M. 191
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Proof: Let and be any two direct similarities given by the following
x' ax by c
equations. ( x, y) ( x' , y' ) where , a2 b2 0
y ' bx ay d
x' mx ny p
( x, y) ( x' , y' ) where , m2 n2 0
y ' nx my q
We need to show that the product is also a direct similarity.
For any point ( x, y) , ( x, y) ( ( x, y)) ( x' ' , y' ' ) where
x' ' mx'ny ' p
m(ax by c) n(bx ay d ) p
(ma nb) x (mb na) y mc nd p
tx ly r , with t ma nb, l mb na, r mc nd p
y ' ' nx' my' q
n(ax by c) m(bx ay d ) q
(na mb) x (ma nb) y nc md q
lx ty s, with s nc md q
Now, these equations to be a similarity, we are left to show that t 2 l 2 0 .
But,
t 2 l 2 (ma nb) 2 (mb na) 2
(ma) 2 2abmn (nb) 2 (mb) 2 2abmn (na) 2
(ma) 2 (na) 2 (mb) 2 (nb) 2
m 2 (a 2 b 2 ) n 2 (a 2 b 2 )
(m 2 n 2 )(a 2 b 2 ) 0, sin ce a 2 b 2 0, m 2 n 2 0
t 2 l 2 0.
x' tx ly r 2 2
Hence, ( x, y) ( ( x, y)) ( x' , y' ) where , t l 0
y ' lx ty s
But by, these are equations of a similarity obtained by applying the plus sign in
the general equation of similarities. Hence, is a direct similarity whenever
and are direct. Similarly, show that the product of any two opposite
similarities is a direct similarity. Now, let’s proof the last part.
The product of a direct and opposite similarities is opposite similarity.
Prepared by Begashaw M. 192
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Let be a direct similarity given by ( x, y) ( x' , y' ) where
x' ax by c
, a2 b2 0
y ' bx ay d
Let be opposite similarity given by ( x, y) ( x' , y' ) where
x' mx ny p
, m2 n2 0
y ' nx my q
We need to show that the product is opposite similarity.
For any point ( x, y) ,we have ( x, y) ( ( x, y)) ( x' ' , y' ' ) where
x' ' mx'ny ' p
m(ax by c) n(bx ay d ) p
(ma nb) x (mb na) y mc nd p
tx ly r , with t ma nb, l mb na, r mc nd p
y ' ' nx'my'q
n(ax by c) m(bx ay d ) q
(na mb) x (ma nb) y nc md q
lx ty s
Where the constants l , t , s are as given in the first part. Besides, we showed
above that t 2 l 2 0 .
x' tx ly r
Thus, we got ( x, y) ( ( x, y)) ( x' , y' ) where , t2 l2 0
y ' lx ty s
So, these are equations of a similarity obtained by applying the minus sign in
the general equations of similarities. Hence, is an opposite similarity
whenever and are direct and opposite similarities.
Example: Let be a transformation given by ( x, y) ( x' , y' ) where
x' 3x 4 y 2
.
y' 4 x 3 y 2
Show that is a similarity and find the ratio k of this similarity. Finally,
determine whether it is direct or opposite similarity.
Solution: To show that is a similarity, it suffices to show A' B' k AB for
any two arbitrary points A, B and for some positive constant k .
Prepared by Begashaw M. 193
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x' 3x 4 y 2
Let A ( x, y), B ( z, w) . Then, A' ( A) ( x' , y' ) where and
y' 4 x 3 y 2
z ' 3z 4w 2
B' ( B) ( z' , w' ) where .
w' 4 z 3w 2
Thus,
A' B' ( z ' x' ) 2 ( w' y ' ) 2
[3( z x) 4( w y )]2 [4( z x) 3( w y )]2
9( z x) 2 24( z x)( w y ) 16( w y ) 2 16( z x) 2 24( z x)( w y ) 9( w y ) 2
25( z x) 2 25( w y ) 2
5 ( z x) 2 ( w y ) 2
5 AB , AB ( z x) 2 ( w y ) 2
This implies that is a similarity with ratio k 5 .
Now, we are left to determine whether it is direct or opposite similarity . From
the general similarity equation, we have ( x, y) ( x' , y' ) where
x' ax by c
, a2 b2 0 .
y ' [bx ay d ]
x' ax by c 3x 4 y 2
Equating the equation of with this gives us
y ' [bx ay d ] 4 x 3 y 2
Here, from the firs equation we get,
ax by c 3x 4 y 2 a 3, b 4, c 2. From the
second we have [bx ay d ] 4 x 3 y 2 [4 x 3 y d ] 4x 3 y 2 . If
we apply the plus sign and equate corresponding coefficients, we get
[4 x 3 y d ] 4 x 3 y 2 4 4, 3 3, d 2 which is not possible.
Again by applying the minus sign, [4 x 3 y d ] 4 x 3 y 2 d 2 .
Thus, with the values a 3, b 4, c 2, d 2 , ( x, y) ( x' , y' ) where
x' ax by c
, a2 b2 0 .
y ' [bx ay d ]
This shows that is an opposite similarity.
Prepared by Begashaw M. 194
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Review Problems on Chapter-4
1. If x' 3x 5 y 2 and y' tx 3 y are the equations of a similarity, then find
the value of t . Answer :t 5
2. Let H be a homothety with factor k and center c . For what value of k , H
will be a half turn about c ? Answer : k 1
3. If (0,0) (1,0), (1,0) (2,2), (2,2) (1,6) for a similarity , then find
(1,6)
4. Let be a dilation with center C (2,3) and factor k 3. If d (C, P' ) 15cm ,
where ( P) P' , find d (C, P). Answer :5cm
5. A dilation maps the point (1,4) to (6,9) and the point (0,1) to (1,12) . Is it a
homothety or a translation? Write the formula for this dilation.
Answer : A homothety given by H ( x, y) (7 x 1,7 y 19)
6. *Consider the line L : y 2 x 17 such that f : L R is given by
f ( x, y) 5 x, ( x, y) L . Show that f is a transformation and conclude that it
is a similarity. Finally, find a point P if f ( P) 7 5 . Answer : P (7,3)
x 13 x 83
7. Given a homothety H P ,k 1 . Then give the ratio k and the center
10
3
y y 3
P of this homothety. Answer : Ratio k 1/ 3, Center C (4,5)
x px 4 y q
8. Let be a similarity with ratio k 5 .
y 4x 3y r
2 8
If , then find the value of the constants p, q and r .
1 2
Answer : p 3, q 6, r 9
9. Show that a mapping f ( X ) mX B is
a) A translation if m 1
B
b) A homothety with factor m and center C if m 1 .
1 m
Prepared by Begashaw M. 195
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
10. Given a line : ( x, y) (1,3) r (2,5) and a homothety
f ( x, y) 5( x, y) (6,1) . Find the image f (l ) ,the factor k and center C of the
homothety.
(6,1) 1 3 1
Answer : By problem 7, the factor is k 5 and center C (6,1) ( , )
1 5 4 2 4
11. Let the mapping f ( x, y) (h g )( x, y) be defined by
f ( x, y) ( x, y) (6,2) . Then find
a) The formula for g ( x, y) if h is a homothety with factor 2 and center (3,2)
b) The formula for h( x, y) if g is a half turn at (1,1) .
1
Answer : a) h( x, y) ( x 9, y 2)
2
12. Give an example of similarity with one fixed point but not isometry.
13. Let H C ,k be a homothety and l be any line. Show that H C ,k (l ) l C l .
14. Let ( x, y) (2 x 1,2 y 3) be a similarity. Find a homothety H C ,k and an
isometry such that H C ,k .
15. Consider a dilatation C ,r with r 3, C (1,1). Find a vector v and a linear
map such that C ,r Tv . Answer : v (2,2), ( x, y) (3x,3 y)
x' 5 x 7 y 2
16. If the transformation with equations is an opposite
y ' tx 5 y 4
similarity, then find the value of the constant t . Answer : t 7
17. If a homothety H C ,k ( x, y) (3x 6,3 y 2t ) fixes the line l : 4 x 3 y 9 0 ,
find the value of the constant t .(Hint: Use problem 13). Answer : t 7
18. If a half turn H P ( x, y) (6 x, k y) fixes the line l : 4 x 3 y 9 0 , find the
value of the constant k . Answer : k 14
19. Let H o,k be homothety about the origin and let be any similarity with
1
factor . Show that U H o,k is an isometry.
k
Prepared by Begashaw M. 196
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
20. If H 1 and H Q ,k are any two homotheties, then show that H Q,k H 1 TPR
P, P,
k k
for some point R .
21. Given A (2,1), B (2,4), A' (2,4) and B' (14,13) . Find a dilation
such that ( A) A' and ( B) B' .
x' 3x 4 y 2
22. Let be a transformation given by ( x, y) ( x' , y' ) where .
y' 4 x 3 y 2
Show that is a similarity and find the ratio k of this similarity. Finally,
determine whether it is direct or opposite similarity.
23. A dilation maps the point (1,4) to (6,9) and the point (0,1) to (1,12) .
a) Is it a homothety or a translation?
b) Write the formula for the dilation.
c) Find a homothety H C ,k and an isometry such that H C ,k .
24. Show that
a) The ratio of every similarity is unique.
b) The image of a vector under a homothety is a parallel vector.
c) Any similarity is a collineation.
d) The product of two homotheties with common center is commutative
e) If a dilatation fixes two points, then it is an identity transformation.
25. Let A' B' C ' be the medial triangle of ABC . Prove that there exists a
dilation which maps ABC to A' B' C ' . Find the center and scaling factor.
Prepared by Begashaw M. 197
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
CHAPTER-5
AFFINE TRANSFORMATIONS
5.1 Introduction
In affine space, any three points A ( x, y), B ( z, w), C (u, v) are said to be
A 1 x y 1
collinear if and only if B 1 z w 1 0
C 1 u v 1
Consider two mappings in a plane f ( x, y) ( x 3 , y 1), g ( x, y) ( y 1, x 2).
Let A (0,0), B (1,2), C (2,4) . Are these points collinear?
A 1 0 10
1 2
B 1 1 1
2 0 . Hence A, B, C are collinear.
2 4
C 1 2 4 1
On the other hand, consider the two given mappings above.
Under f , A' f ( A) (0,0), B' f ( B) (1,3), C' f (C) (8,3) and
A' 1 0 10
1 3
B' 1 1 1
3 21 0 . Thus, A' , B' , C ' are not collinear.
8 3
C' 1 8 3 1
Again, using g instead of f , we have
A' g ( A) (1,2), B' g ( B) (0,4), C' g (C) (3,2) and
A' 1 1 12
4 1 0 1 0 4
B' 1 0 1
4 2 6 6 12 0 .
2 1 3 1 3 2
C' 1 3 2 1
Thus, A' , B' , C ' are collinear. Such transformations like g which maps
collinear points into collinear points are said to be affine transformations, but
mappings like f are not considered as affine transformation.
Prepared by Begashaw M. 198
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Definition: Let g : W W be a transformation. Then, g is said to be affine
transformation if and only if it preserves co-linearity. (It maps any collinear
points into collinear points).
Example: Let g : R 2 R 2 be given by g ( x, y) ( x y, x y). Show that g is
affine transformation.
Solution: Clearly g is a transformation.
Let A ( x, y), B ( z, w), C (u, v) be any collinear points.
A 1 x y 1
Then, B 1 z w 1 0
C 1 u v 1
Now, consider their images under g ,
A' g ( A) ( x y, x y), B' g ( B) ( z w, z w), C' g (C) (u v, u v) .
A' 1 x y x y 1 x x y 1 x y 1
Thus, B' 1 z w z w 1 z z w 1 z w 1 0 .
C' 1 uv uv 1 u uv 1 u v 1
(Here we used properties of determinants)
Thus, A' , B' , C ' are collinear. This means the three collinear points A, B, C are
mapped into collinear points under g . So g is affine transformation.
Coordinate definition of affine transformation: A mapping g : R 2 R 2 is
said to be affine transformation if and only if it is of the form
x ax by e
g , where ad bc 0. This definition is equivalent to the
y cx dy f
definition we stated first. In this book, we mainly follow this definition.
Consider the two mappings in a plane that we saw above
f ( x, y) ( x 3 , y 1), g ( x, y) ( y 1, x 2). Here, g is of the form given in this
definition and thus it is affine transformation but f is not of the form given in
the definition because the power of x is 3 and thus it is not affine
transformation.
Prepared by Begashaw M. 199
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.2 Basic Properties of Affine Transformations
1.Any affine transformation maps lines into lines.
Proof: Let l be any line and g be any affine transformation. Suppose A, B, C
are collinear points on l .
Since g preserves co-linearity, A' g ( A), B' g ( B), C' g (C) are also collinear.
Hence, A' , B' , C ' are all on the same line say l ' . Now for any point
P on l collinear with A, B , P' g ( P) is collinear with A' , B' and found on l ' .
Otherwise, if for at least one point Q on l , Q' g (Q) is not on l ' , Q ' will be
non-collinear with A' , B' which means g maps three collinear points A, B, Q
into non-collinear points A' , B' , Q' . But this is a contradiction with the
definition of affine transformation (as g is supposed to be affine
transformation)
2. An affine transformation maps parallel lines into parallel lines.
Proof: Let m and n be any two distinct parallel lines and g be an affine
transformation such that m' g (m), n' g (n) . We need to show that m' // n' .
Suppose they are not parallel. Then there exist at least one common point P' .
But, P' m'n' P m, Q n, g ( P) P' , g (Q) P' g ( P) g (Q) P Q.
Because g is one to one. This means, P' m'n' P m n, g ( P) P'. But
this is not possible as the lines m and n are parallel and distinct. Consequently,
m' // n' whenever m // n .
3. The image of a plane under affine transformation is again a plane.
Proof: Let be any plane. If m and n are any two intersecting or parallel
lines on , then from above properties their images are intersecting or
parallel under a given affine transformation. But any two intersecting or
parallel lines determine a plane.
4. An affine transformation preserves „betweeness‟: If P is any point between
A and B , then g (P) is a point between g (A) and g (B) .
Prepared by Begashaw M. 200
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5. An affine transformation preserves mid-point: If M is the mid-point of A
and B , then g (M ) is the mid-point of g (A) and g (B) .
6. An affine transformation preserves length ratio of line segments lying along
a line. If A, B, C are collinear points, then it is always true that
g ( A) g ( B) A B
.
g ( B) g (C ) B C
7. The image of a triangle under affine transformation is another triangle.
8. The image of a parallelogram under affine transformation is another
parallelogram.
9. An affine transformation preserves cenroid of a triangle.
10. An affine transformation preserves bay centric coordinates or center of
mass.
11. The image of an ellipse under affine transformations is another ellipse.
Geometric Properties that are not preserved under AT
As there are preserved properties of geometric figures under affine
transformations, there are also properties which are not preserved under affine
transformations.
1. Lengths of line segments (Distances) are not preserved under affine
transformations.
x 2x
Example: Let g : R 2 R 2 be affine transformation given by g .
y 2y
x z
Take any two points A , B . Then, AB ( z x) 2 (w y) 2 .
y w
But,
2x 2z
A' , B' A' B' 4( z x) 2 4( w y ) 2
2y 2w
2 ( z x) 2 ( w y ) 2
2 AB A' B' AB
2. Circles are not preserved under affine transformations. Under affine
transformations, generally circles are mapped into ellipses.
x 1 x
Example: Let g : R 2 R 2 be affine transformation given by g 13 .
y 4 y
Find the image of the circle C : x 2 y 2 144 .
Prepared by Begashaw M. 201
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x x' x 1 x
Solution: Let be any point on the given circle. Then, g 13 .
y y' y 4
y
Solving this equation for x and y in terms of the images x' and y ' yields,
x 3 x' , y 4 y ' .
Now put these values in the equation of the circle to find the required image.
C : x 2 y 2 144
C ': (3x' ) 2 (4 y ' ) 2 144
C ': 9 x' 2 16 y ' 2 144
x' 2 y ' 2
C ': 1
16 9
Thus, the image is an ellipse.
3. The norm of a vector is not preserved under affine transformations.
4. Angle measures are not preserved under affine transformations.
5. Area and volume are not preserved under affine transformations.
6. Affine transformations do not preserve orientation of plane figures.
7. Right, Isosceles and equilateral triangles are not preserved under affine
transformations.
8. Rectangles are not preserved under affine transformations.
9. Generally, affine transformations do not preserve shape and size.
Proposition 5.1: Let g be an affine transformation that fixes two distinct
points A, B . Then g fixes the whole line through A, B point wise.
Proof: Given g ( A) A, g ( B) B . We need to show that g ( P) P for any point
on a line through A, B . Let P be any point on a line through A, B . Then,
A, B, P are collinear so that AP r AB . On the other hand as g preserves co
linearity, A' , B' , P' are also collinear. So, A' P' r A' B' . But A' A, B' B implies
AP' r AB . Thus combining AP r AB and AP' r AB , gives
AP' r AB AP P' A P A P' P
Hence, g fixes the whole line through A, B point wise.
Prepared by Begashaw M. 202
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.3 Types of Affine Transformations
5.3.1 Line (Skew)-Reflections
Definition: Let l and k be two intersecting lines in a plane. Then, line
refection is a reflection on line k in the direction of line l that maps every point
P to the point P such that the line through P and P is parallel to the line l
and the midpoint of P and P lies on the line of reflection k .
Notation: Let S be line reflection on k in the direction of l .
P, if P k
Then, S ( P) P P'
P' , PP ' // l , k , if P' k
2
Here, the line k is called axis of reflection and the line l is called direction of
reflection.
Example: Find the image of the point (2,1) by a reflection on a line
k : 2 x y 1 0 in the direction of the line l;3x y 6 0 .
Solution: Let the image of the point (2,1) be ( x' , y' ) . Then, the line through
(2,1) and ( x' , y' ) is parallel to the direction line l . As any two parallel lines
have the same slope, we get
y'1
3 3x' y'5 0...................................................................................(i)
x'2
On the other hand the mid point of (2,1) and ( x' , y' ) lies on the axis of
reflection k , hence it satisfies the equation of line k .
( x '22 , y '21 ) k 2( x '22 ) ( y '21 ) 1 0 2 x' y'7 0...................................(ii )
Prepared by Begashaw M. 203
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Combining (i ) and (ii ) , we get the following system of linear equations
2 x' y '7 0
5 x' 2 x' 52 , y' 31
P' ( 52 , 31
5)
3 x ' y '5 0 5
Therefore, the image of the point (2,1) by a reflection on the line
k : 2 x y 1 0 in the direction of the line l;3x y 6 0 is P' ( 52 , 31
5)
.
Remarks:
1. In reflection problems, the line through a point P and its image P' is
perpendicular to the line of reflection, but in skew or line reflection, this is not
necessarily true. For instance, in the above example the line through the point
P (2,1) and its image P' ( 52 , 31
5)
is not perpendicular to the line
k : 2 x y 1 0 .This is because their slopes are m 2 and m' 3 such that
m.m' 6 1 , but the product of the slope of
two non- vertical and perpendicular lines must be -1.
2. If the direction line of reflection is perpendicular to the line of reflection,
then the line through P and its image P' is also perpendicular to the line of
reflection or axis of reflection. In this case, it becomes the normal or the usual
reflection that we know it before. It is simply called orthogonal reflection. That
means, every reflection is line reflection but every line reflection is not
orthogonal reflection.
Prepared by Begashaw M. 204
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.3.2 Compressions
Definition: Let l and k be two intersecting lines in a plane. Then, a
compression on line k in the direction of line l with factor t is a transformation
that maps every point P to the point P such that the line through Q and P is
parallel to the line l where Q is a point on line k .
Notation: Let C be a compression on k in the direction of l . Then,
P, if P k
C ( P)
P' , QP' t QP, Q k , if P k
Here, the line k is called axis of compression, the line l is called direction of
compression, the scalar t is called factor of the compression and the point Q is
called the critical point.
Example: Find the image of the point (1,4) by a compression on the line
k : 2 x y 3 0 in the direction of the line l : 3x y 6 0 with factor t 23
Solution: Let Q (a, b) be the critical point on line k such that the line QP is
parallel to l . Since parallel lines have the same slope, we have
b4
3 3a b 7 0 . On the other hand Q is on line k means it satisfies
a 1
the equation k : 2 x y 3 0 . Thus, 2a b 3 0 . Collecting these two
3a b 7 0
equations yields 5a 10 a 2, b 1 Q (a, b) (2,1)
2a b 3 0
But from the definition of compression, we have
C ( P) P' , QP' t QP P'Q 23 ( P Q) ( x' , y ' ) (2,1) 23 [(1,4) (2,1)]
( x' , y ' ) (2,1) ( 23 ,2) ( 43 ,3)
Note that to solve compression problems, the first task is to identify the axis of
compression and the direction line, then to find the critical point Q and finally
to find the image P' using QP' t QP where t is factor of the compression.
Prepared by Begashaw M. 205
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.3.3 Shears
Definitions:
a) Horizontal Shears:
Shears in the x direction with factor k that sends each point ( x, y) parallel to
the x axis by an amount of ky to the point S x ( x, y) ( x ky, y) are known as
horizontal shears.
Under a horizontal shear points on the x axis are invariant or unmoved
because on the x axis y 0 so that x ky x 0 x .
But when we move away from the x axis, the magnitude of y increases, so
that points farther from the x axis moves a greater distance than which are
closer to or on the x axis.
b) Vertical Shears:
Shears in the y direction with factor t that sends each point ( x, y) parallel to
the y axis by an amount of tx to the point S y ( x, y) ( x, y ty ) are known as
vertical shears.
Under such types of shears points on the y axis are invariant or fixed and
points farther from the y axis moves a greater distance than closer points to
the y axis.
c) Total Shears (Simply called shears):
A total shear or simply a shear with factors k and t is a transformation that
maps any point ( x, y) to the point S ( x, y) ( x ky, y tx) .
Examples: Find the equation of a shear that maps
a) (2,3) to (8,3) b) (3,1) to (3,0) c) (2,6) to (8,2)
Solution: To find the equations, first identify the type of the shear from the
images.
a) Here, the x-coordinate is changed and the y-coordinate is fixed. Hence, the
shear is horizontal.
Prepared by Begashaw M. 206
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
So, S x ( x, y ) ( x ky, y ) S x (2,3) (2 3k ,3)
(2 3k ,3) (8,3)
k 2 S x ( x, y ) ( x 2 y , y )
b) Here, the y-coordinate is changed and the x-coordinate is fixed.
Hence, the shear is vertical.
So,
S y ( x, y ) ( x, y tx ) S y (3,1) (3,1 3t )
1 1
(3,1 3t ) (3,0) t S y ( x , y ) ( x, y x )
3 3
c) Here, both coordinates are changed and thus it is a total shear.
So,
S ( x, y ) ( x ky, y tx ) S (2,6) (2 6k ,2t 6)
(2 6k ,2t 6) (8,2) k 1, t 4
S ( x, y ) ( x y , y 4 x )
Prepared by Begashaw M. 207
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.3.4 Sililarities
All isometries are affine transformations.Generally, all similarities
(Homothety and or Dilations) are affine transformations. All isometries are
similarities. But there are similarities that are not isometries. For instance,
homotheties are similarities but not isometries. All similarities are affine
transformations but there are affine transformations which are not similarities.
For instance, shears, line reflections and compressions are affine
transformations but not similarities.
x x'
Theorem 5.1(Characterization): Let g be affine transformation
y y'
x' ax by h
where , ad bc 0. Then,
y ' cx dy k
I. g is a similarity if and only if a 2 c 2 b 2 d 2 , ab cd 0 .
II. g is an isometry if and only if a 2 c 2 b 2 d 2 1, ab cd 0.
x z u zx
Proof: Let A , B , C be vertices of ABC .Then, AB
y w v w y
For the first part, g is a similarity if and only if A' B' r AB where,
AB ( z x) 2 ( w y ) 2 ,
A' B' (a 2 c 2 )( z x) 2 (b 2 d 2 )( w y ) 2 (2ab 2cd )( z x)( w y )
From these relations,
A' B' r AB
(a 2 c 2 )( z x) 2 (b 2 d 2 )( w y ) 2 (2ab 2cd )( z x)( w y )
r ( z x) 2 ( w y ) 2
a 2 c 2 r 2 , b 2 d 2 r 2 , 2ab 2cd 0
a 2 c 2 b 2 d 2 , ab cd 0
Thus, g is a similarity if and only if a 2 c 2 b 2 d 2 , ab cd 0 .
Finally, g is an isometry if and only if A' B' AB .
A' B' AB (a 2 c 2 )( z x) 2 (b 2 d 2 )( w y ) 2 (2ab 2cd )( z x)( w y )
( z x) 2 ( w y ) 2
a 2 c 2 1, b 2 d 2 1, 2ab 2cd 0
a 2 c 2 b 2 d 2 1, ab cd 0
Prepared by Begashaw M. 208
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Problem Set 5.1
x 3x 4 y 5
1. If : R 2 R 2 is an affine transformation given by , find
y px 2 y 1
all values of p .
x' x y
2. Show that the mapping given by S ( x, y) ( x' , y' ) where is a shear.
y' y
3. Determine whether the set of shears forms a group of transformation with
composition or not.
4. If the image of (1,3) under a compression C on a line k : y 2 x 3 in the
direction of the line l : 2 x y 2 0 is the point P' ( 34 , 72 ) , find the image of the
line m : x y 1 0 .
5. Prove that any affine transformation maps two intersecting lines into
intersecting lines. (Affine transformation preserves point of concurrency).
Preserves ratio of the areas of triangles.
6. Show that the inverse of a horizontal (Vertical) shear is again a horizontal
(vertical) shear with opposite factor.
7. Let u and w be perpendicular vectors in R 2 . Define a mapping T : R 2 R 2
by T (v) v (u.v)w for all v in R 2 . Show that a) T is a linear transformation.
b) T is a shear parallel to a line L spanned by w .
c) If u (2,1) and w (3,6) , find T (v) where v (3,4) .
8. Let S be a shear parallel to a line L in R 2 . Show that
a) S (v) v for all v on L b) S (v) v // L for all v in R 2 .
9. Find the image of the point (2,3) and the line l : 4 x 2 y 6 0
a) By a reflection on the line l1 : 6 x 3 y 12 0 in the direction of the line
l 2 : 3x y 2 0
b) By a compression on the line l1 : 6 x 3 y 12 0 in the direction of the
1
line l 2 : 3x y 2 0 with factor . c) By a shear with factor 2.
3
10. Let g be a reflection on the line l1 : 15x 2 y 10 0 in the direction of the
line l 2 : 3x 7 y 10 0 . Give at least three fixed points and three fixed line of g .
Prepared by Begashaw M. 209
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.4 Affine Transformations and Linear Mappings
Definition: Linear transformations are any transformations from a plane onto a
x' ax by
plane given by ( x, y) ( x' , y' ) where , a, b, c, d R, ad bc 0
y ' cx dy
x
Notation: From now onwards we use the notation instead of ( x, y) for
y
simplicity.
Using this notation, the above linear transformation becomes
x x' ax by
y y ' cx dy
Examples:
a) Reflection on the x axis and y axis are linear.
x x
Let S x be a reflection on x axis. Then, S x .
y y
Hence, S x is linear.
b) Rotations about the origin are linear.
Let be a rotation by an angle about the origin.
x
Then, for any point ,
y
x x cos y sin
.
y x sin y cos
x 2x 3y
c) : R 2 R 2 given by is linear.
y 3x 2 y
x 2x
d) : R 2 R 2 given by is not linear.
y 3x 1
Prepared by Begashaw M. 210
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.5 Matrix Representation of Affine Transformations
So far we have seen that any linear transformation f is given by the formula
x ax by
f , ad bc 0 . This formula can be equivalently expressed as
y cx dy
x a b x ax by a b
f . Now, denote the first matrix by M and
y c d
y cx dy c d
x
the second by X .Thus any linear transformation f is given by
y
f ( X ) MX . This is known as matrix Representation (formula) of linear
transformation and M is called the standard matrix of f .
From now onwards, we use the notation f ( X ) MX , but to use this formula
we have to see the method how to find the matrix M .
a b x ax by
The formula f ( X ) MX is true for a unique M and
c d
y cx dy
x 1 0
for any point X . In particular, this formula works for X , Y .
y 0 1
1 a b 1 a 0 a b 0 b
So, f , f .
0 c d 0 c 1 c d 1 d
1
This means the first column of M is obtained by calculating f and the
0
0 1 0 a b
second by calculating f so that M f f .
1 0 1 c d
Examples:
x 2x y
a) Let f : R 2 R 2 be given by f . Find the standard matrix M
y 3x 7 y
of f and write its matrix representation.
a b
Solution: Let M be the standard matrix of f . Then, from the above
c d
a 1 2 b 0 1 2 1
relation, f , f M .
c 0 3 d 1 7 3 7
b) Let be a rotation by an angle about the origin. Find the matrix M of
Prepared by Begashaw M. 211
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x
Solution: For any point X , a rotation by an angle about the origin is
y
x x cos y sin a b
given by so that for M ,
y x sin y cos c d
a 1 cos b 0 sin cos sin
f , f M
c
0 sin d 1 cos sin cos
In conclusion, for any linear transformation f , there is a unique matrix M
1 0
whose first and second columns are the images of and respectively.
0 1
Conversely, to each 2x2 invertible matrix M there corresponds a unique linear
x
transformation f whose formula is obtained by multiplying any point by
y
x x
the matrix M as f M .
y y
2 5
For instance, for M , the formula of the corresponding linear
4 9
x 2 5 x 2x 5 y
transformation f is f .
y 4 9 y 4 x 9 y
Rules of Thumb!
i. To obtain the standard matrix of a linear transformation whose formula is
given, simply take the coefficients of x and y . Then, form the matrix by
putting the coefficients of x in the first column and the coefficients of y as
second column. For instance, let f : R 2 R 2 be a linear transformation given
x 5x 2 y 5 2
by f . Here, the coefficients of x are and that of y are .
y 3x 4 y 3 4
So, putting the coefficients of x in the first column and the coefficients of y as
5 2
second column yields a 2x2 invertible matrix M .
3 4
Prepared by Begashaw M. 212
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
ii. To find the formula of a linear transformation, whose standard matrix M is
x
given, simply multiply M by the column vector .
y
For instance, let f : R 2 R 2 be a linear transformation whose standard matrix
0 3
is M . Then, the formula of f will be
2 1
x x 0 3 x 3 y
f M .
y y 2 1 y 2 x y
At this point, we are equipped with enough concepts that are useful to find the
matrix representation of any affine transformations.
Theorem 5.2 (Representation Theorem): Let g be any affine transformation.
Then, g can be represented uniquely as a product of linear transformation f
and a translation with translation vector b as g ( X ) f ( X ) b .
Proof: Consider the general coordinate definitions of affine transformation that
we discussed at the beginning of this Chapter,
x ax by e
g , where ad bc 0.
y cx dy f
x ax by e ax by e
Here, g ( X ) g .
y cx dy f cx dy f
0 e
Now, observe that g . So, by taking
0 f
x ax by 0 e
f ( X ) f , b g , we get that g ( X ) f ( X ) b .
y cx dy 0 f
0 0
Besides, f which shows that f is linear transformation.
0 0
0
The uniqueness follows from the uniqueness of b which is b g and from
0
x
the uniqueness of the coefficients in the formula of the image of f . (Note
y
0
that if either the image of or the coefficients are not unique, the map g will
0
not be bijective. That is why we used these facts).
Prepared by Begashaw M. 213
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x x y 3
Example: Represent the transformation g as
y 2x 5 y 2
g( X ) f ( X ) b .
0 3 x x y
Solution: Using b g , f , we get g ( X ) f ( X ) b .
0 2 y 2x 5 y
Theorem 5.3 (Matrix Representation Theorem (MRT) of Affine
Transformations): Let g : R 2 R 2 be any affine transformation. Then, there
exist a unique matrix M and a unique vector b such that g ( X ) MX b . This
representation is said to be matrix representation of g .
Proof: From theorem 5.2, for any affine transformation g , there is a unique
linear transformation f and a unique vector b such that g ( X ) f ( X ) b . Again
from the above discussions, to each linear transformation f , there is a unique
matrix M such that f ( X ) MX .
Consequently, we get that g ( X ) f ( X ) b MX b . Our next task will be
determining the matrix associated with a given affine transformation g .
How to determine M from g ( X ) MX b ?
Here we need to investigate how to find the standard matrix M associated with
a given affine transformation.
x x
From g f b , since any linear transformation preserves the origin
y y
0 0 x x
g f b b and f g b .
0 0 y y
But from the previous observations we have seen that the standard matrix M of
1
any linear transformation is completely determined from the images of and
0
0 a 1 b 0 a b
as f , f so that M .
1 c 0 d 1 c d
Prepared by Begashaw M. 214
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x x
Thus, from f g b ,
y y
a 1 1 b
0 0 a b
f g b ,
f g b so that M .
c 0 0 1 d 1 c d
x x y 3
Example: Let g : R 2 R 2 be affine given by g . Find the
y 2x 3 y 7
standard matrix of g and give its matrix representation.
a b x
Solution: Let the matrix representation be g ( X ) MX b b .
c d y
0 0 3
Now, find g from the formula of g ,that is b g .
0 0 7
Besides, from the above relations,
a 1 4 3 1 b
0 0 2 3 1
g b ,
f g b
c 0 5 7 2 d
1 1 4 7 3
a b 1 1
Consequently, the standard matrix is found to be M .
c d 2 3
Therefore, the matrix representation is given by
x x 1 1 x 3
g M b .
y y 2 3 y 7
How can we determine affine transformations from their images?
In our discussion of isometries, we have seen that any isometry is uniquely
determined from its effect on three non-collinear points. Likewise, any affine
transformation is uniquely determined from the images of three non-collinear
points. That means there is a unique affine transformation that sends three non-
collinear points into three non-collinear points. In other words, for any three
non- collinear points A, B, C , and any two affine transformations f and g if
g ( A) f ( A), g ( B) f ( B), g (C) f (C ) , then g f .
Therefore, to determine the formula of any affine transformation g , it suffices
to know the images of any three non- collinear points A, B, C .
This basic and important concept is illustrated by the following example.
Prepared by Begashaw M. 215
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Example: Let g : R 2 R 2 be affine transformation with
2 6 0 10 4 23
g , g and g . Then, determine the formula of
0 0 5 7 10 10
0
g and calculate g .
0
2 0 4
Solution: Since the points A , B , C are non-collinear, there is
0 5 10
a b
a unique affine transformation g given by g ( X ) MX b , where M
c d
e
and b so that g ( A) MA b, g ( B) MB b, g (C ) MC b . Thus,
f
a b 2 2a 5b 16
i. M ( B A) g ( B) g ( A)
c d 5 2c 5d 7
a b 4 4a 5b 13
ii. M (C B) g (C ) g ( B)
c d 5 4c 5d 17
Equating components and collecting like terms from (i ) and (ii ) , we get
2a 5b 16
6a 3 a 12 , b 3
4 a 5 b 13
2c 5d 7
6c 24 c 4, d 15
4c 5d 17
a b 12 3
Hence, M 1
.
c d 4 5
Now using one of the given points A, B or C , we obtain the vector b as
follow:
2 12 3 2 6 1 5
g ( A) MA b b g ( A) MA g
1
0 4 5 0 0 8 8
x
Therefore, for any X , g ( X ) MX b gives
y
x 1 3 x 5 1 x 3 y 5 0 5
g 2 1 2 , and g .
y 4 5 y 8 4 x 5 y 8 0 8
1
Prepared by Begashaw M. 216
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.6 Orientation and Affine Transformations (Revisited)
Definitions: Let g be any affine transformation. Then, ee say that g preserves
orientation if and only if for any positively oriented vectors X and Y , their
images X ' g ( X ), Y ' g (Y ) are again positively oriented vectors. In this case,
g is said to be orientation preserving affine transformation. In general, if the
pair ( X , Y ) and the pair ( g ( X ), g (Y )) have the same orientation, then g preserves
orientation. But, if they have opposite orientation, then g reverses (changes)
orientation. In this case, g is said to be orientation reversing (changing) affine
transformation.
Theorem 5.4 (Orientation Characterization Theorem):
Let g : W W be any affine transformation given by g ( X ) MX b . Then,
a) g preserves orientation if and only if det M 0.
b) g reverses (changes) orientation if and only if det M 0.
Proof: Whether a given transformation g preserves or reverses orientation is
determined from its effect on the orientation of a triangle. This means if g
preserves the orientation of any triangle ABC , then it is orientateon preserving
and if g reverses the orientation of ABC , then it is orientation reversing affine
transformation.
Now, having this fact as basis, let’s prove our theorem. Let ABC be arbitrary
triangle. Then its orientation is determined from the orientation of the vectors
AB and AC . Suppose ABC has positive orientation. That means det( AB, AC ) 0
On the other hand, let A' B' C ' be the image of ABC under g . Then, the
orientation of A' B' C ' is determined from the orientation of the pair ( A' B', A' C ') .
But from g ( X ) MX b ,
A' g ( A) MA b, B' g ( B) MB b, C ' g (C ) MC b
.
A' B' M ( B A), A' C ' M (C A)
Prepared by Begashaw M. 217
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Thus,
det( A' B', A' C ') det[M ( B A), M (C A)] det[M ( B A, C A)]
det[M ( AB, AC )] det M . det( AB, AC )
Hence, g preserves orientation if and only if
det( A' B', A' C ') 0 det M . det( AB, AC ) 0 det M 0 because det( AB, AC ) 0
from our assumption. Similarly, g reverses orientation if and only if
det( A' B', A' C ') 0 det M . det( AB, AC ) 0 det M 0 .
Note that since any affine transformation g is bijective, it is invertible and so is
its standard matrix M and hence det M can not be zero.
So, any affine transformation g is either orientation preserving or orientation
reversing but cannot be both.
Examples: Determine whether the following affine transformations are
orientation preserving or orientation reversing.
x 3x 2 y 7
a) g : R 2 R 2 given by g
y x 5 y 11
x x 1
b) : R 2 R 2 given by
y y 11
Solution:
x 3x 2 y 7 3x 2 y 7 3 2 x 7
a) Here, g
y x 5 y 11 x 5 y 11 1 5 y 11
3 2
M det M 17 0
1 5
Hence, by the above theorem g preserves orientation.
x x 1 x 1 1 0 x 1
b) Similarly,
y 2 y y 2 0 1 y 2
1 0
M det M 1 0
0 1
Hence, by the above theorem reverses orientation.
Prepared by Begashaw M. 218
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Theorem 5.5: The inverse of orientation-preserving affine transformation is
again orientation-preserving. The inverse of orientation-reversing affine
transformation is again orientation-reversing.
Proof: Let g ( X ) MX b be orientation-preserving affine transformation.
Then, we need to show that g 1 is also orientation-preserving. Since g is
orientation-preserving, det M 0 . Besides, from g ( X ) MX b , by Inverse
Characterization Theorem, g 1 ( X ) M 1 X M 1 b which gives us the standard
matrix of g 1 is M 1 .
1
Thus, det(M 1 ) 0, det M 0 .
det M
Hence, g 1 ( X ) M 1 X M 1 b is also orientation-preserving whenever
g ( X ) MX b is orientation -preserving. Similarly, if g ( X ) MX b is
1
orientation –reversing, then det M 0 . So, det(M 1 ) 0, det M 0 .
det M
Theorem 5.6: The composition of two orientation-preserving affine
transformations is orientation-preserving. The composition of two orientation-
reversing affine transformations is orientation-preserving. The composition of
an orientation-preserving and an orientation-reversing affine transformation is
orientation-reversing.
Proof: Let g ( X ) MX b and h( X ) NX c be any affine transformations. In
section 6.6, we have seen that the matrix of the composition g h is the product
of the matrices of g and h .Thus, the standard matrix of g h is MN .
Now, suppose g and h are orientation-preserving.
That means, det M 0, det N 0 .So, det(MN ) det M . det N 0 . On the other
hand, if g and h are orientation-reversing, det M 0, det N 0 gives
det(MN ) det M . det N 0 .
Thus, in either cases, g h is orientation-preserving.
Finally, if g is orientation preserving and h is orientation reversing, we will
have det M 0, det N 0 det MN det M . det N 0.
Hence, if g and h have opposite effect on orientation, then their composition
g h will be orientation reversing.
Prepared by Begashaw M. 219
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.7 Area and Affine Transformations
Let be a parallelogram whose adjacent sides are given by the vectors
x z
X and Y as shown in the figure 5.3 below. For simplicity of
y w
x z
calculation, assume one of the vertices to be the origin and X , Y
y w
x z
to be positively oriented vectors. That means det( X , Y ) 0.
y w
Now let’s calculate the area of the parallelogram denoted by a( ) .
That is a( ) b.h where b is the base of the parallelogram given by
b X and h is the altitude from the vertex Y to the base X . Here, how to
find h will need a little algebraic calculation as follow. From Pythagoras
Theorem,
2
X .Y ( X .Y ) 2
h d Y h Y d Y PrX Y Y ( Y
2 2 2 2 2 2 2 2 2 2
2
)X 2
X X
X Y ( X .Y ) 2 ( x 2 y 2 )( z 2 w 2 ) ( xz yw) 2
, X .Y xz yw
X
2
x2 y2
( xz ) 2 ( xw) 2 ( yz ) 2 ( yw) 2 [( xz ) 2 2 xzyw ( yw) 2 ]
x2 y2
( xw) 2 2( xw)( yz ) ( yz ) 2 ( xw yz ) 2
x y
2 2
x2 y2
xw yz
h
x2 y2
Prepared by Begashaw M. 220
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Hence, the area of the parallelogram is computed by
xw yz x z
a( ) b.h ( x 2 y 2 )( ) xw yz
x2 y2 y w
Notice: The first bar is to indicate absolute value and the second is to indicate
determinant in the above area formula. From this area of a parallelogram, if we
are interested on the area of the triangle with adjacent sides the vectors
x z 1 x z
X and Y , we simply multiply by half. a(OXY ) .
y w 2 y w
In general, for arbitrary parallelogram with vertices
A , B , C , D , the area of is given by
x z u r
y w v t
zx ux
a( ) . For arbitrary triangle ABC with vertices
w y v y
x z u 1 zx ux
A , B , C , we have a(ABC )
y w v 2 w y v y
(The derivations of these formulas are similar as we did above). The plus sign
of the determinant is obtained when the triangle has positive orientation where
as the minus sign is taken when the triangle has negative orientation. Notation
from now onwards, we use the following notations.
x z u
For vertices A , B , C , the pair ( AB, AC ) is a matrix formed by
y w v
putting the coordinates of AB as first column and the coordinates of AC as the
z x u x
second column as follow: ( AB, AC ) .
w y v y
Let det( AB, AC ) denotes determinant of ( AB, AC ) so that we can write
z x u x
det( AB, AC ) . Hence,
w y v y
1 1 zx ux
a(ABC ) det( AB, AC ) .
2 2 w y v y
Now let g be any affine transformation. We need to investigate the effect of g
on the area of ABC . Let A' B' C ' be the image of ABC under g where
A' g ( A), B' g ( B), C' g (C) .
Prepared by Begashaw M. 221
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Since g is affine transformation, there is a unique matrix M and a vector b
such that g ( X ) MX b for all points X in the plane. Thus,
A' g ( A) MA b, B' g ( B) MB b, C ' g (C ) MC b
A' B' MB b ( MA b) MB MA,
A' C ' MC b ( MA b) MC MA
Hence, from the above area formula,
a(A' B' C ' ) 12 det( A' B', A' C ') 12 det(MB MA, MC MA)
12 det[M ( B A, C A)] 12 det M det(B A, C A) , det( XY ) det X det Y
12 det(B A, C A) det M , xy x y
12 det( AB , AC ) det M a(ABC ) det M , 1
2 det( AB , AC ) a(ABC )
This relation in general is given by the following theorem.
Theorem 5.7 (Area Relation Theorem, ART):
Let g be any affine transformation given by g ( X ) MX b . Then,
a(A' B' C ' ) a(ABC ) det M where A' B' C ' is the image of ABC .
Examples:
1. Given ABC with vertices A (4,1), B (2,0), C (2,2). Let g be affine
x x 4y 3
transformation given by g . If A' B' C ' is the image of ABC
y 3x 7 y 2
under g , compute a(A' B' C ' ) .
Solution: First find the area of ABC so that to apply the area relation theorem.
In ABC , AB (6,1), AC (2,3). Thus,
6 2 1
a(ABC ) 12 det( AB, AC ) 12 det 2 18 2 8 sq.units
1 3
Prepared by Begashaw M. 222
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
1 4
Now from the formula of g , its standard matrix is given by M such
3 7
that det M 5 . Therefore, from area relation theorem,
a(A' B' C ' ) a(ABC ) det M 8 5 40 sq. units.
2. Suppose a certain affine transformation g takes ABC into A' B' C ' such that
a(A' B' C ' ) 13 a(ABC ) .If g takes DEF into D' E' F ' where
a(D' E ' F ' ) 7cm 2 , calculate a(DEF ) .
Solution: For an affine transformation g , if A' B' C ' is the image of ABC
under g , then a(A' B' C ' ) a(ABC ) det M . But we are given that,
a(A' B' C ' ) 13 a(ABC ) a(ABC ) det M 13 a(ABC ) det M 13 det M 13
Thus,
a(D' E ' F ' ) a(DEF ) det M 7cm 2 a(DEF ) 13 a(DEF ) 21cm 2
Definition: Equi-affine transformation is an affine transformation that
preserves area.
An affine transformation preserves area of ABC if a(A' B' C ' ) a(ABC ) .
Corollary:
An affine transformation g ( X ) MX b is equi affine if and only if det M 1
Proof: Suppose g is equi affine transformation. Thus, for any ABC ,
a(A' B' C ' ) a(ABC ) . But from area relation theorem,
a(A' B' C ' ) a(ABC ) det M a(ABC ) det M a(ABC ) det M 1 det M 1.
Conversely, suppose det M 1 . Then,
a(A' B' C ' ) a(ABC ) det M a(ABC ) 1 a(ABC ) .
Hence, g preserves area.
Prepared by Begashaw M. 223
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x x'
Theorem 5.8: Let g be any affine transformation where
y y'
x' ax by h
, ad bc 0. Then, g is equi affine if and only if ad bc 1 .
y ' cx dy k
x z u
Proof: Let A , B , C be vertices of ABC .
y w v
zx u x
Then, AB , AC so that
w y v x
z x u x
a(ABC ) 12 det( AB, AC ) 12 det .
w y v y
On the other hand, let A' B' C ' be the image of ABC under g , then
ax by h az bw h au bv h
A' g ( A) , B' g ( B) , C ' g (C )
cx dy k cz dw k cu dv k
a ( z x ) b( w y ) a(u x) b(v y )
A' B' , A' C '
c( z x) d ( w y ) c(u x) d (v y )
Hence, using the general area formula for the image A' B' C ' , we have
a(A' B' C ' ) 12 det( A' B', A' C ')
a( z x) b( w y ) a (u x) b(v y )
12 det
c( z x) d ( w y ) c(u x) d (v y )
a b z x u x
12 det
c d w y v y
a b z x u x
12 det det
c d w y v y
z x u x a b
12 det det
w y v y c d
a(ABC ) ad bc a(A' B' C ' ) a(ABC ) ad bc)
Hence, g is equi affine if and only if
a(A' B' C ' ) a(ABC ) a(ABC ) ad bc a(ABC ) ad bc 1
Prepared by Begashaw M. 224
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
5.8 Inverse of Affine Transformations
Let g ( X ) MX b and h( X ) NX c be any two affine transformations. Then,
their composition is given by
i. g h( X ) g (h( X )) g ( NX c) MNX M c b
g h( X ) CX d , C MN , d M c b
ii. h g ( X ) h( g ( X )) h( MX b) NMX N b c
h g ( X ) DX e, D NM , e N b c
In either case, the matrix of the composition is the product of the matrices of
the individual affine transformations.
Definition: Any two affine transformations g and f are said to be inverse of
each other if and only if f g ( X ) g f ( X ) i( X ) X for all X in the
domain. For instance in R , f ( x) x 1 and g ( x) x 1 are inverses of each
other because f g ( x) g f ( x) i( x) x, x R. Now our aim is to find the
inverse of an affine transformation whose formula is given. Consider any two
affine transformations g ( X ) MX b and f ( X ) M 1 X M 1 b . Then,
i. f g ( X ) f ( MX b) M 1 ( MX b) M 1 b X M 1 b M b X
ii. g f ( X ) g ( M 1 X M 1 b) M ( M 1 X M 1 b) b X b b X
Thus, from these two cases, f g ( X ) g f ( X ) i( X ) X for all X in the
given space. Hence, the given affine transformations g and f are inverse of
each other. This observation is generalized by the following theorem.
Theorem 5.9 (Inverse Characterization Theorem, ICT):
Let g ( X ) MX b be affine transformation. Then, the inverse of g is given by
g 1 ( X ) M 1 X M 1 b.
Notice: To find the inverse of any given affine transformation g : R 2 R 2
given by g ( X ) MX b , it is advisable to remember the following steps:
Step-1: Find carefully the standard matrix M and the vector b of g using:
0
i) b g
0
a b a 1 b 0
ii) M , where g b, g b
c d c 0 d 1
Prepared by Begashaw M. 225
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
d b
Step-2: Calculate M 1 using M 1 1
ad bc
c a
Step-3: Compute g 1 using inverse characterization theorem.
g 1 ( X ) M 1 X M 1 b
Example: Find the inverse of an affine transformation g given by
x 9x 4 y 2
g .
y 2x y 5
Solution: To find the inverse of g , the first task is to identify its standard
matrix M so that to apply theorem 5.9.
a b 0 2
Let g ( X ) MX b where M . Hence, b g and
c d 0 5
a 1 11 2 9 b 0 2 2 4
g b , g b
c 0 3 5 2 d 1 6 5 1
9 4
Thus, M . Now calculate M 1 .
2 1
a b
For any 2x2 invertible matrix M , its inverse is given by
c d
d b
M 1 1
ad bc
.
c a
9 4 1 4 1 4
So, for M , M 1 11 .
2 1 2 9 2 9
Therefore,
x 1 4 x 1 4 2
g 1 ( X ) M 1 X M 1 b g 1
y 2 9
y 2 9 5
x 4 y 22
2 x 9 y 49
x 4 y 22
2 x 9 y 49
Prepared by Begashaw M. 226
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
Review Problems on Chapter-5
x x 2y
1. Suppose T is an affine transformation given by T .Then, find
y x 2y
a) the image of the y -axis under T
b) the image of the line x y 1 under T
c) the image of the circle x 2 y 2 1 under T
Answer : a) y x b) x 3 y 4 0 c) 4( x y) 2 ( y x) 2 16
2. Find the image of the circle x 2 y 2 16 under affine transformation
x 2x
g . Answer : x 2 y 2 64
y 2y
x2 y2
3. Find the image of the ellipse 1 under affine transformation
4 9
x x
g 6y
y 9
Answer : x 2 y 2 1/ 9
3 1
4. Suppose is a linear transformation with standard matrix M .
2 1
Find the image of the line l : y 2 x 1 under . Answer : 4 x 5 y 1 0
2 2
5*. Suppose g is an affine transformation with standard matrix M .
1 0
0 4
If g , find the image of l : y 3x under g . Answer : x 8 y 20 0
0 2
6. Let S l and S m be a reflections on lines a long the vectors u (1,1) and
v (1,2) . Then, find the equations of S l S m and give its standard matrix.
4 3 3 4 4 / 5 3/ 5
Answer : S l S m ( x, y) ( x y, x y), M
5 5 5 5 3 / 5 4 / 5
Prepared by Begashaw M. 227
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x 3x 2 y
7. Let T : R 2 R 2 be defined by T .
y 4x 3y
a) Find the matrix representation of T .
b) Determine whether T preserves or changes orientation.
c) Find the formula for T 1 .
d) If A' B' C ' D' is the image of the parallelogram ABCD under T where a
A(1,1), B(4,1), C (4,3), D(1,3) , calculate the area of A' B' C ' D' .
8. Let g be a reflection on the line l1 : 2 x y 0 in the direction of the line
l 2 : 12 x 3 y 2 0 . If l : tx 3 y 5 0 is the fixed line of g , then find t
9. Let S be a shear which takes the point (2,6) to (5,6) . Then find S 1 (3,2)
10. Let T : R 2 R 2 be a one to one linear transformation. Then,
2x 6 0 , x
a) If T then find the point
3y 9 0
y
b) If 2 3 16
T , then find the image T
3 5 24
11. Let g : R 2 R 2 be affine transformation. If g is a reflection on the line
x y 1 0 in the direction of the line y 2 x . Then,
a) Find the formula for the linear mapping f (x) associated with g (x)
b) Find the matrix representation of g (x)
c) Give at least three fixed points of g
d) Is the line l : 6 x 3 y 12 0 a fixed line of g ?
12. If the image of the (1,3) under a compression C on a line k : y 2 x 3 in the
direction of the line l : 2 x y 2 0 is the point P' ( 34 , 72 ) , find the image of the
line m : x y 1 0 under this compression.
13. Let : R 2 R 2 be linear transformation such that and
2 1
3 0
1 0 y 5x 2 under .
. Find the formula for and the image of
1
1
Prepared by Begashaw M. 228
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
14. Let be a shear such that (1,4) (7,11) . Then, find the equations of .
Answer : ( x, y) ( x 2 y, y 7 x)
15. Suppose f is an orientation-reversing similarity. Prove that f f is either
dilation or a translation.
x ax by
16. Suppose g is an equi-affine transformation. Then, prove that
y bx ay
a 2 b 2 1 .
17. Given the affine mapping g ( x) 3 0
6
x .
3 1 2
a) Is g orientation preserving or orientation changing?
b) Is g isometric mapping?
x 5 x 3
18. Given the affine mapping g and triangle ABC with vertices
y 6 y 4
A(0,1), B(2,1), C(2,5). Find the area of the image triangle A' B' C ' under g .
19. Let g ( x) MX b be affine tansformation. If g ( x) preserves area , then
prove that det(M ) 1.
20. Prove that any homothety preserves orientation.
21. Determine the effects of shears on area and orientation.
22. Let S l be a reflection where l is a line through the origin that makes an
angle of with the positive x axis.
cos 2 sin 2
a) Show that the standard matrix of S l is given by M .
sin 2 cos 2
b) Using part (a), conclude that S l preserves area but changes orientation.
x x
23. Every affine transformation given by f M b is a similarity if and
y y
x x x
only if there exists k 0 such that M k for all vectors in R 2 .
y y y
Prepared by Begashaw M. 229
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
24. In the cell of the following table, put “ ” if the geometric concept will
be preserved or if it will be changed under the given type of
transformations in general.
25. Produce an example of similarity but not an isometry
a) That preserves area but not orientation.
b) That preserves orientation but not area.
c) That preserves both area and orientation.
x' kx
26. Let k be a mapping for k 0 given by k ( x, y) ( x' , y' ) where
y' y
a) Show that k is affine transformation. Can k be a similarity?
b) Determine whether k is equi-affine or not. Is it orientation preserving?
x ax 2a
27. Let T : R 2 R 2 be defined by T , where a and b are non zero
y by 3b
constants. Show that T is affine transformation and find the formula for T 1 .
1
Let be a plane with area A . Give the area of T () ? Is T equi-affine?
16
Is T orientation preserving or reversing?
28. Determine the orientation of ABC whose vertices are given as follow
4 1 1
A , B , C
1 1 3
Prepared by Begashaw M. 230
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
x 3x 4 y 2
29. Given f
y 4 x 3 y 1
a) Show that f is an affine transformation
b) Show that f is a similarity
c) Find a dilation , a rotation R and a translation
Tv such that f Tv R
d) Is f orientation preserving?
30. Classify the following affine transformations as orientation preserving and
orientation changing.
a. Dilation
b. Line reflection
c. Shears
d. Compression
x2 y2
31. Show that the ellipse 2 2 1 is the image of the circle x 2 y 2 1 under
a b
some affine transformation.
32. If 2 x' 3x y 2 and 2 y' x 3 y 1 are equations of a rotation, then
find the center and angle of the rotation.
33. Let be an isometry with (0,0) (2,1), (1,1) (1,0), (2,3) (5,1) .Then
find the equations of . Determine whether even or odd isometry?
Prepared by Begashaw M. 231
Text Book of Transformation Geometry by Begashaw M. For your comments, use -0938836262
REFERENCES
[1] G.E. Martin. (1982). Transformation Geometry. New York: Springer-
Verlag. [2] H. Anton. (1987). Elementary Linear Algebra. New York: John
Wiley- Sons.
[3] S. Singh. (1982). Modern Algebra. New Delhi: Vikas publishing, 3rd ed.
[4] E. Moise. Elementary Geometry from an advanced standpoint
Prepared by Begashaw M. 232
You might also like
- Duc Tran - Mathematical Induction 101 - With 101 Practice ProblemsNo ratings yetDuc Tran - Mathematical Induction 101 - With 101 Practice Problems250 pages
- Working With OOTB MACH Simulation ExamplesNo ratings yetWorking With OOTB MACH Simulation Examples44 pages
- 52a Trigonometry 2D and Bearings - H - Question PaperNo ratings yet52a Trigonometry 2D and Bearings - H - Question Paper22 pages
- Similar Figures 2 TEST 2023 +ans Key PDFNo ratings yetSimilar Figures 2 TEST 2023 +ans Key PDF10 pages
- Applied Mathematics: Syllabus & Programme StructureNo ratings yetApplied Mathematics: Syllabus & Programme Structure36 pages
- List of Mathematical Symbols - Wikipedia PDFNo ratings yetList of Mathematical Symbols - Wikipedia PDF22 pages
- MATH1179 Vectors and Matrices Book SolutionNo ratings yetMATH1179 Vectors and Matrices Book Solution4 pages
- Chapter 3 Vectors in 2-Space and 3-SpaceNo ratings yetChapter 3 Vectors in 2-Space and 3-Space68 pages
- Partial Derivatives: Chapter 2: Function of Several VariablesNo ratings yetPartial Derivatives: Chapter 2: Function of Several Variables26 pages
- HIGH STANDARDS IN MATHS G10 DK-2-unlockNo ratings yetHIGH STANDARDS IN MATHS G10 DK-2-unlock96 pages
- Transformations of Exponential & Logarithmic Functions100% (1)Transformations of Exponential & Logarithmic Functions3 pages
- Linear Equations in 2 Variables Worksheet 3No ratings yetLinear Equations in 2 Variables Worksheet 31 page
- Applied Math III (Math 2042), Worksheet IINo ratings yetApplied Math III (Math 2042), Worksheet II6 pages
- 2.5: Numerical Integration - Midpoint, Trapezoid, Simpson's RuleNo ratings yet2.5: Numerical Integration - Midpoint, Trapezoid, Simpson's Rule8 pages
- Vectors - Definition, Properties and Algebra: Topic B1.1No ratings yetVectors - Definition, Properties and Algebra: Topic B1.124 pages
- Matrix & Determinant (Solution of Some Questions)No ratings yetMatrix & Determinant (Solution of Some Questions)23 pages
- Math Practice Questions - Coordinate GeometryNo ratings yetMath Practice Questions - Coordinate Geometry2 pages
- C4 Vectors Exam Questions: (C4 June 2014 (R) Q6) 1100% (1)C4 Vectors Exam Questions: (C4 June 2014 (R) Q6) 126 pages
- Transformations in Geometry: By: Mathryn Obligado and Reshel Ladio BSED - Mathematics 2No ratings yetTransformations in Geometry: By: Mathryn Obligado and Reshel Ladio BSED - Mathematics 259 pages
- Name: Teacher: Date: Score:: TransformationsNo ratings yetName: Teacher: Date: Score:: Transformations2 pages
- Creating 3D NX CAM Tool Assemblies V3.1 PDF100% (2)Creating 3D NX CAM Tool Assemblies V3.1 PDF37 pages
- Calc - Sec 6-2 - IC Volume by Known Cross-Section Worksheet With Answers #2No ratings yetCalc - Sec 6-2 - IC Volume by Known Cross-Section Worksheet With Answers #24 pages
- Chapter S1 (Celestial Timekeeping and Navigation)50% (2)Chapter S1 (Celestial Timekeeping and Navigation)28 pages
- Large Deflection Analysis of A Thin Plate Computer Simulations and ExperimentsNo ratings yetLarge Deflection Analysis of A Thin Plate Computer Simulations and Experiments24 pages
- Building Assemblies With Mates: Creating An Assembly100% (1)Building Assemblies With Mates: Creating An Assembly20 pages
- Handbook ISO GPS Tolerancing Cetim CVDL EN V1.02d PDF100% (4)Handbook ISO GPS Tolerancing Cetim CVDL EN V1.02d PDF76 pages
- 28fff394a46039b4c6de932a38a5f707c8614346f566972741cc7ef682c74af7No ratings yet28fff394a46039b4c6de932a38a5f707c8614346f566972741cc7ef682c74af76 pages
- The Configuration Space Method For Kinematic Design of Mechanisms100% (2)The Configuration Space Method For Kinematic Design of Mechanisms213 pages
- Tsuneo Yoshikawa - Foundations of Robotics - Analysis and Control-The MIT Press (1990)No ratings yetTsuneo Yoshikawa - Foundations of Robotics - Analysis and Control-The MIT Press (1990)294 pages
- Line of Position On The Sun: Sin (T) Sin (Lat .) Sin (T) Sin (Lat .) 1No ratings yetLine of Position On The Sun: Sin (T) Sin (Lat .) Sin (T) Sin (Lat .) 12 pages
