Analytic Representations of Affine Subspaces: A) A Point in A and The Director Space Dir (A) L L

Analytic representations of affine subspaces
Let An be an n−dimensional affine space with the K−linear director space

dir (An ) = Ln , and A0p An a p−dimensional affine subspace. Then A0p is
uniquely determined by
A) a point in A0p and the director space dir (A0p ) = L0p ≤ Ln .
Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 1 / 29

B) a system of p + 1 affine independent points in A0p .

Assume F = (O; B = (vi )i=1,n ) is a cartesian frame of An . We shall write
analytic representations for A0p according to the two possibilities:

Assume F = (O; B = (vi )i=1,n ) is a cartesian frame of An . We shall write
analytic representations for A0p according to the two possibilities:

A) Let P0 ∈ A0p and let (uj )j=1,p be a basis in L0p . For a point P ∈ A0p , we
can write
OP = OP0 + P0 P .
Since P0 P ∈ L0p , there are scalars t1 , t2 , . . . , tp ∈ K such that
P0 P = t1 u1 + t2 u2 + · · · + tp up .
Also,
OP = x1 v1 + x2 v2 + · · · + xn vn and
OP0 = x10 v1 + x20 v2 + · · · + xn0 vn .
If
uj = u1j v1 + u2j v2 + · · · + unj vn , j = 1, p,

A) Let P0 ∈ A0p and let (uj )j=1,p be a basis in L0p . For a point P ∈ A0p , we
can write
OP = OP0 + P0 P .
Since P0 P ∈ L0p , there are scalars t1 , t2 , . . . , tp ∈ K such that
P0 P = t1 u1 + t2 u2 + · · · + tp up .
Also,
OP = x1 v1 + x2 v2 + · · · + xn vn and
OP0 = x10 v1 + x20 v2 + · · · + xn0 vn .
If
uj = u1j v1 + u2j v2 + · · · + unj vn , j = 1, p,

we obtain
n
P n
P p
P
xk vk = xk0 vk + t j uj =
k=1 k=1 j=1
n p n

P P P
= xk0 vk + tj ukj vk =
k=1 j=1 k=1
!
n
P p
P
= xk0 + ukj tj vk .
k=1 j=1
Hence, for any k = 1, n, we have

p
X
xk = xk0 + ukj tj .
j=1
These are called the scalar parametric equations of the affine subspace A0p .

we obtain
n
P n
P p
P
xk vk = xk0 vk + t j uj =
k=1 k=1 j=1
n p n

P P P
= xk0 vk + tj ukj vk =
k=1 j=1 k=1
!
n
P p
P
= xk0 + ukj tj vk .
k=1 j=1
Hence, for any k = 1, n, we have

p
X
xk = xk0 + ukj tj .
j=1
These are called the scalar parametric equations of the affine subspace A0p .

Denoting
   
x1 x10
 x2   x20 
X = [OP]B =  , X = [OP ] =  ,
   
..  0 0 B  ..
 .   . 
xn xn0
respectively for each j = 1, p

 
u1j
 u2j 
Uj = [uj ]B =  ,
 
..
 . 
unj
we obtain the matrix parametric equation

p
X
X = X0 + tj Uj .
j=1

B) Let now (P0 , P1 , . . . , Pp ) be a system of p + 1 affine independent
points in A0p . Then (P0 Pj )j=1,p is a basis of L0p , so that we can write
parametric equations in the form
p
X
xk = xk0 + (xkj − xk0 ) , k = 1, n .
j=1
Pp
Denoting αj = tj for j = 1, p and α0 = 1 − j=1 tj , we can rewrite the
last equations in the form
p
X p
X
xk = αj xkj , k = 1, n , where αj = 1 .
j=0 j=0
In matrix form, we have

p
X p
X
X = αj Xj , where αj = 1 .
j=0 j=0

B) Let now (P0 , P1 , . . . , Pp ) be a system of p + 1 affine independent
points in A0p . Then (P0 Pj )j=1,p is a basis of L0p , so that we can write
parametric equations in the form
p
X
xk = xk0 + (xkj − xk0 ) , k = 1, n .
j=1
Pp
Denoting αj = tj for j = 1, p and α0 = 1 − j=1 tj , we can rewrite the
last equations in the form
p
X p
X
xk = αj xkj , k = 1, n , where αj = 1 .
j=0 j=0
In matrix form, we have

p
X p
X
X = αj Xj , where αj = 1 .
j=0 j=0

Special cases
Affine line(p = 1)

Special cases
Affine line(p = 1)
a) Let A01 An be an affine line with director subspace

dir (A01 ) = L01 = hui, and P0 ∈ A01 . If
u = u1 v1 + u2 v2 + · · · + un vn ,
the scalar parametric equations of the affine line A01 are
xk = xk0 + t · uk , k = 1, n .

Special cases
Affine line(p = 1)
a) Let A01 An be an affine line with director subspace

dir (A01 ) = L01 = hui, and P0 ∈ A01 . If
u = u1 v1 + u2 v2 + · · · + un vn ,
the scalar parametric equations of the affine line A01 are
xk = xk0 + t · uk , k = 1, n .

Eliminating the parameter t, we obtain the equations
x1 − x10 x2 − x20 xn − xn0
= = ··· = .
u1 u2 un

b) Let now P0 , P1 ∈ A01 . Then L01 = hP0 P1 i, and we can write the
previous equations as
x1 − x10 x2 − x20 xn − xn0
= = ··· = .
x11 − x10 x21 − x20 xn1 − xn0

Affine hyperplane(p = n − 1)
a) The equations
xk − xk0 = t1 uk1 + t2 uk2 + · · · + tn−1 uk,n−1 , k = 1, n
express exactly the fact that the columns X − X0 , U1 , U2 , . . . , Un−1 of the

matrix [X − X0 U1 U2 . . . Un−1 ] are linear dependent. This is equivalent to
the equation

x1 − x10 u11 u12 . . . u1,n−1

x2 − x20 u21 u22 . . . u2,n−1
= 0.

.. .. .. .. ..

. . . . .

xn − xn0 un1 un2 . . . un,n−1

Affine hyperplane(p = n − 1)
a) The equations
xk − xk0 = t1 uk1 + t2 uk2 + · · · + tn−1 uk,n−1 , k = 1, n
express exactly the fact that the columns X − X0 , U1 , U2 , . . . , Un−1 of the

matrix [X − X0 U1 U2 . . . Un−1 ] are linear dependent. This is equivalent to
the equation

x1 − x10 u11 u12 . . . u1,n−1

x2 − x20 u21 u22 . . . u2,n−1
= 0.

.. .. .. .. ..

. . . . .

xn − xn0 un1 un2 . . . un,n−1

b) In case (P0 , P1 , P2 , . . . , Pn−1 ) is a set of n affine independent points of
the hyperplane A0n−1 , we obtain the equation

x1 − x10 x11 − x10 x12 − x10 . . . x1,n−1 − x10

x2 − x20 x21 − x20 x22 − x20 . . . x2,n−1 − x20
= 0.

.. .. .. .. ..

. . . . .

xn − xn0 xn1 − xn0 xn2 − xn0 . . . xn,n−1 − xn0
This equation is equivalent to

1 1 1 ... 1

x1 x10 x11 . . . x1,n−1

x2 x20 x21 . . . x2,n−1 = 0.

.. .. .. .. ..
.
. . . .

xn xn0 xn1 . . . xn,n−1

b) In case (P0 , P1 , P2 , . . . , Pn−1 ) is a set of n affine independent points of
the hyperplane A0n−1 , we obtain the equation

x1 − x10 x11 − x10 x12 − x10 . . . x1,n−1 − x10

x2 − x20 x21 − x20 x22 − x20 . . . x2,n−1 − x20
= 0.

.. .. .. .. ..

. . . . .

xn − xn0 xn1 − xn0 xn2 − xn0 . . . xn,n−1 − xn0
This equation is equivalent to

1 1 1 ... 1

x1 x10 x11 . . . x1,n−1

x2 x20 x21 . . . x2,n−1 = 0.

.. .. .. .. ..
.
. . . .

xn xn0 xn1 . . . xn,n−1

Homomorphisms of affine spaces
Let (A, L, ϕ) and (B, M, ψ) be two K−affine spaces.
Definition
A function τ : A −→ B is called affine function or affine homomorphism if
τ (αP + βQ) = ατ (P) + βτ (Q) , (∀)P, Q ∈ A, α, β ∈ K : α + β = 1 .

Proposition
Let τ : A −→ B be a function, O ∈ A and O 0 = τ (O) ∈ B. Then τ is an
affine homomorphism if and only if the unique function t : L −→ M
defined by
t ◦ ϕO = ψO 0 ◦ τ
is a linear homomorphism.

Remark
Let τ : A −→ B be an affine map, O ∈ A, and t : L −→ M the linear
map defined by t(OA) = τ (O)τ (A). Then for any points A, B ∈ A we
have
t(AB) = t(OB − OA) = t(OB) − t(OA) =
= τ (O)τ (B) − τ (O)τ (A) = τ (A)τ (B) .
t is called the associated linear map or director map of τ , denoted
t = dir (τ ).

Corollary
1) An affine application τ : A −→ B is completely determined by a pair of
correspondent points O ∈ A and O 0 = τ (O) ∈ B, and the associated
linear application t : L −→ M.
2) The affine application τ : A −→ B is injective(resp. surjective) if and
only if t = dir (τ ) is injective(resp. surjective).

Corollary
1) An affine application τ : A −→ B is completely determined by a pair of
correspondent points O ∈ A and O 0 = τ (O) ∈ B, and the associated
linear application t : L −→ M.
2) The affine application τ : A −→ B is injective(resp. surjective) if and
only if t = dir (τ ) is injective(resp. surjective).

Proposition
Let τ : A −→ B be an affine application, and t = dir (τ ) its director map.
a) If A0 A is an affine subspace with dir (A0 ) = L0 , then τ (A0 ) B is an
affine subspace with dir (τ (A0 )) = t(L0 ).

Proposition
b) If B 0 B is an affine subspace with dir (B 0 ) = M 0 then τ −1 (B 0 ) A is
an affine subspace with dir (τ −1 (B 0 )) = t −1 (M 0 ).

Proposition
Corollary
Let τ : A −→ B be an affine application. Then

Proposition
Corollary
1) Im(τ ) B.

Proposition
Corollary
1) Im(τ ) B.
2) If A0 , A00 A are subspaces such that A0 k A00 , then τ (A0 ) k τ (A00 ).

Proposition
Corollary
1) Im(τ ) B.
3) If τ is injective, then τ −1 (B) is a point in A for each B ∈ Im(τ ).

Proposition
Corollary
1) Im(τ ) B.
4) Generally, τ −1 (B) A, with dir (τ −1 (B)) = ker (t) for each B ∈ Im(τ ).

Proposition
Corollary
1) Im(τ ) B.
4) Generally, τ −1 (B) A, with dir (τ −1 (B)) = ker (t) for each B ∈ Im(τ ).

Proposition
Let (A1 , L1 ), (A2 , L2 ), (A3 , L3 ) be affine spaces, and τ : A1 −→ A2 ,
σ : A2 −→ A3 affine applications with associated linear applications
t : L1 −→ L2 and s : L2 −→ L3 . Then σ ◦ τ : A1 −→ A3 is an affine
application with associated linear application s ◦ t : L1 −→ L3 .
Proposition
If τ : A −→ B is a bijective linear application, with associated linear map
t : L −→ M, then τ −1 : B −→ A is also an affine application, and its
director map is t −1 : M −→ L.

Proposition
Let (A1 , L1 ), (A2 , L2 ), (A3 , L3 ) be affine spaces, and τ : A1 −→ A2 ,
σ : A2 −→ A3 affine applications with associated linear applications
t : L1 −→ L2 and s : L2 −→ L3 . Then σ ◦ τ : A1 −→ A3 is an affine
application with associated linear application s ◦ t : L1 −→ L3 .
Proposition
If τ : A −→ B is a bijective linear application, with associated linear map
t : L −→ M, then τ −1 : B −→ A is also an affine application, and its
director map is t −1 : M −→ L.

Theorem
The set of all bijective affine applications(called also affine
transformations, or affine automorphisms) of an affine space A onto itself
forms a group with respect to function composition, called the affine group
of the affine space A, and denoted GA(A).

Definition
By affine geometry of the space A one understands all the properties of
figures(=subsets) in A, which remain invariant(=unchanged) with
respect to the affine group of A. These are called affine properties.
Example
1) The property of a set of points of being affine dependent,
respectively affine independent, is an affine property.

Definition
Example
2) The ratio in which a point divides a bipoint is invariant with respect
to affine transformations, hence it is an affine invariant.

Definition
Example
3) The cross-ratio of four collinear points is an affine invariant.

Definition
Example
4) The property of two subspaces of being parallel is an affine property.

Definition
Example
4) The property of two subspaces of being parallel is an affine property.

Definition
An affine transformation τ : A −→ A such that dir (τ ) = idL is called a
translation of A.
Remark
The affine map τ : A −→ A is a translation of A if and only if
Oτ (O) = Pτ (P) , (∀)O, P ∈ A .
The common vector Oτ (O) is called the translation vector of τ .

Definition
An affine transformation τ : A −→ A such that dir (τ ) = idL is called a
translation of A.
Remark
The affine map τ : A −→ A is a translation of A if and only if
Oτ (O) = Pτ (P) , (∀)O, P ∈ A .
The common vector Oτ (O) is called the translation vector of τ .

Proposition
The set of all translations of an affine space A is a subgroup GT (A) of
the affine group GA(A), and
(GT (A), ◦) ∼
= (L, +) .

Definition
Let O ∈ A be a fixed point of an affine space A. An affine
transformation τ : A −→ A such that τ (O) = O is called a center
affinity of A with center O.
Remark
The affine transformation τ : A −→ A is a center affinity of A with
center O if and only if the associated linear application t : L −→ L
satisfies the relation
t(OP) = τ (O)τ (P) , (∀)P ∈ A .

Definition
Let O ∈ A be a fixed point of an affine space A. An affine
transformation τ : A −→ A such that τ (O) = O is called a center
affinity of A with center O.
Remark
The affine transformation τ : A −→ A is a center affinity of A with
center O if and only if the associated linear application t : L −→ L
satisfies the relation
t(OP) = τ (O)τ (P) , (∀)P ∈ A .

Proposition
Let O ∈ A be a fixed point of the affine space A. The set of all center
affinities of A with center O is a subgroup GCAO (A) of the affine group
GA(A), and
(GCAO (A), ◦) ∼= (GL(L), ◦) .

Proposition
Let A be an arbitrary affine space, and O ∈ A a fixed point in A. Then
any affine transformation τ : A −→ A can be decomposed as τ = τ1 ◦ τO ,
where τO ∈ GCAO (A) and τ1 ∈ GT (A).

Homomorphisms of finite dimensional affine spaces
Let (An , Ln , ϕ) and (Bm , Mm , ψ) be two finite dimensional K−affine
spaces, τ : An −→ Bm an affine application, and t : Ln −→ Mm the
associated linear application of τ .
Definition
The rank of the affine application τ is the rank of the linear map
t(= dimK (Im(t))). The defect of τ is the defect of the linear map
t(= dimK (Ker (t))).

Homomorphisms of finite dimensional affine spaces
Let (An , Ln , ϕ) and (Bm , Mm , ψ) be two finite dimensional K−affine
spaces, τ : An −→ Bm an affine application, and t : Ln −→ Mm the
associated linear application of τ .
Definition
The rank of the affine application τ is the rank of the linear map
t(= dimK (Im(t))). The defect of τ is the defect of the linear map
t(= dimK (Ker (t))).

Proposition
Let τ : An −→ Bm be an affine map. Then
a) rank(τ ) = dim(Im(τ )).

Proposition
b) def (τ ) = dim(τ −1 (B)), (∀)B ∈ Im(τ ).

Proposition
b) def (τ ) = dim(τ −1 (B)), (∀)B ∈ Im(τ ).
c) rank(τ ) + def (τ ) = n.

Proposition
b) def (τ ) = dim(τ −1 (B)), (∀)B ∈ Im(τ ).
c) rank(τ ) + def (τ ) = n.

Proposition
Let F = (O, B = (vi )i=1,n ) and F 0 = (O 0 , B 0 = (wj )j=1,m ) cartesian
frames of the affine spaces An , respectively Bm , and τ : An −→ Bm an
affine map. Then there are scalars aij , bj ∈ K, with i = 1, n, j = 1, m such
that for any point P(xi )i=1,n ∈ An , the image τ (P) = P 0 (yj )j=1,m is given
by
Xn
yj = aij xi + bj , (∀)j = 1, m .
i=1
In matrix form, if X = [OP]B and Y = [O 0 P 0 ]B0 , then
Y = AX + B ,
or
1 1 01×n 1
= · .
Y B A X

Proposition
Let F = (O, B = (vi )i=1,n ) and F 0 = (O 0 , B 0 = (wj )j=1,m ) cartesian
frames of the affine spaces An , respectively Bm , and τ : An −→ Bm an
affine map. Then there are scalars aij , bj ∈ K, with i = 1, n, j = 1, m such
that for any point P(xi )i=1,n ∈ An , the image τ (P) = P 0 (yj )j=1,m is given
by
Xn
yj = aij xi + bj , (∀)j = 1, m .
i=1
In matrix form, if X = [OP]B and Y = [O 0 P 0 ]B0 , then
Y = AX + B ,
or
1 1 01×n 1
= · .
Y B A X

Definition
The matrix
1 01×n
B A
is called the matrix associated to the affine map τ with respect to the
cartesian frames F and F 0 , and is denoted [τ ]F
F0.

Proposition
Let τ : An −→ An be an affine application, F = (O, B = (vi )i=1,n ) a
cartesian frame of An , and
Y = AX + B , X ∈ Kn
the matrix equation of τ . Then

a) τ is an affine transformation of An ⇐⇒ det(A) 6= 0.

Proposition
Y = AX + B , X ∈ Kn

b) τ is a translation of An ⇐⇒ A = In .

Proposition
Y = AX + B , X ∈ Kn

c) τ is a center affinity with center O ⇐⇒ B = 0n×1 .

Proposition
Y = AX + B , X ∈ Kn

c) τ is a center affinity with center O ⇐⇒ B = 0n×1 .

Theorem
Let F = (O, B = (vi )i=1,n ) and F 0 = (O 0 , B 0 = (vi0 )i=1,n ) be two cartesian
frames of an n−dimensional affine space An . Then there is a unique affine
transformation τ : An −→ An such that
τ (O) = O 0 and t(vi ) = vi0 , (∀)i = 1, n .
If a point P ∈ An has coordinates (xi ) with respect to F, and its image

τ (P) has coordinates (xi0 ) with respect then the equations of τ with
respect to F and F 0 are
xi0 = xi , (∀)i = 1, n .

Theorem
Let F = (O, B = (vi )i=1,n ) and F 0 = (O 0 , B 0 = (vi0 )i=1,n ) be two cartesian
frames of an n−dimensional affine space An . Then there is a unique affine
transformation τ : An −→ An such that
τ (O) = O 0 and t(vi ) = vi0 , (∀)i = 1, n .
If a point P ∈ An has coordinates (xi ) with respect to F, and its image

τ (P) has coordinates (xi0 ) with respect then the equations of τ with
respect to F and F 0 are
xi0 = xi , (∀)i = 1, n .

Analytic Representations of Affine Subspaces: A) A Point in A and The Director Space Dir (A) L L

Uploaded by

Copyright:

Available Formats

Analytic Representations of Affine Subspaces: A) A Point in A and The Director Space Dir (A) L L

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Analytic Representations of Affine Subspaces: A) A Point in A and The Director Space Dir (A) L L

Uploaded by

Copyright:

Available Formats

Analytic representations of affine subspaces

Let An be an n−dimensional affine space with the K−linear director space

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 1 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 1 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 1 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 1 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 2 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 2 / 29

Hence, for any k = 1, n, we have

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 3 / 29

Hence, for any k = 1, n, we have

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 3 / 29

respectively for each j = 1, p

we obtain the matrix parametric equation

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 4 / 29

In matrix form, we have

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 5 / 29

In matrix form, we have

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 5 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 6 / 29

a) Let A01  An be an affine line with director subspace

the scalar parametric equations of the affine line A01 are

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 6 / 29

a) Let A01  An be an affine line with director subspace

the scalar parametric equations of the affine line A01 are

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 6 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 7 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 8 / 29

xk − xk0 = t1 uk1 + t2 uk2 + · · · + tn−1 uk,n−1 , k = 1, n

express exactly the fact that the columns X − X0 , U1 , U2 , . . . , Un−1 of the

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 9 / 29

xk − xk0 = t1 uk1 + t2 uk2 + · · · + tn−1 uk,n−1 , k = 1, n

express exactly the fact that the columns X − X0 , U1 , U2 , . . . , Un−1 of the

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 9 / 29

This equation is equivalent to

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 10 / 29

This equation is equivalent to

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 10 / 29

τ (αP + βQ) = ατ (P) + βτ (Q) , (∀)P, Q ∈ A, α, β ∈ K : α + β = 1 .

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 11 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 12 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 13 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 14 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 14 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 15 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 15 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 15 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 15 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 15 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 15 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 15 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 15 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 16 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 16 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 17 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 18 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 18 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 18 / 29

Lect.dr. Mihai Chiş () Lecture 6 24.III.2009 18 / 29

a) Let A01 An be an affine line with director subspace

a) Let A01 An be an affine line with director subspace