Topic 1: ALGEBRA
Matrices and Matrix Operations
Determinants and Inverse Matrices
Vectors
System of linear equations
Rouch-Frbenius Theorem
Cramers Rule
Gausss Rule
MATRICES AND MATRIX OPERATIONS
We will call Matrix of order mxn over the set of the Real
numbers (R,+,.) a set of mxn elements of R ordered in m rows
and n columns:
a11
a
A = 21
am1
a12 ... a1n
a22 ... a2 n
am 2 amn
We say that a matrix is square if m=n. In this case, we will call
trace
Tr(A)= a11+a22+...+ann
We say that a matrix is diagonal if aij=0, if ij. A matrix is null if
aij=0, i,j
A row is a matrix 1xn and a column is a matrix mx1. We say
that a matrix is symmetric if aij=aji
We will denote the set of the matrices mxn by Mmxn.
Given a Matrix A we call its transpose
a11
a
At = 12
a1n
a21
a22
a2 n
... am1
... am 2
amn
If A, B Mmxn, it is possible to calculate A+B:
Given A= (aij), B=(bij), the A+B=(cij), where cij= aij+bij,
i=1,2,...,m; j=1,2,...,n
If R
.A=(cij), where cij= aij , i,j
Given a matrix A Mmxn and B Mnxp, the product of A and B is
denoted by A.B=(cij)mxp, where
n
cij = aik bkj
k =1
Properties:
1.A.(B.C)=(A.B).C
2.A.(B+C)=A.B+A.C
[Link].A
4. If Anxn is a square matrix, I.A=A=A.I, where I is the diagonal
matrix, with aii=1, i=1,2,..., n and aij=0, i j
Definition: The inverse matrix A-1 of a square matrix A is a matrix
such that
A-1.A=I=A.A-1
Definition: Given a square matrix Anxn, we call its Determinant,
and we will denote as |A|,
A = (1)[ p1 , p2 ... pn ] a1 p1 ...anpn
[ p1 , p2 ... pn ]
where
by {p1,...,pn} we denote the possible permutations of the
number 1,2,... n and by [p1,...,pn] the number of translations used to
arrive from 1,2,,n to this permutation.
In particular
Sarruss Rule:
a11
a21
a31
a12
a22
a32
a13
a23 = a11a22 a33 + a12 a23 a31 + a13 a21a32 a11a23 a32 a12 a21a33 a13 a22 a31
a33
We will call the cofactor of the element aij , and denote as
Cij= Co(aij)=(-1)i+j.|Aij|,
where |Aij| is the determinant of order (n-1)x(n-1) matrix obtained
by canceling the row i and the column j.
The Adjugate matrix (adjoint or adjunct) of A is the transpose of
T
the Cofactor Matrix of A
C
C
...
11
1n
Adj ( A) = = C T
Cm1 Cmn
LAPLACE EXPANSION AND HIGHER ORDER DETERMINANTS.
Suppose B = (bij) is an n n matrix and fix any i, j {1, 2, ..., n}.
Then its determinant |B| is given by:
B = bi1Ci1 + bi 2Ci 2 + ... + bin Cin = b1 j C1 j + b2 j C2 j + ... + bnj Cnj
It is a method for ev aluating determinants in terms of cofactors.
It permits higher-order determinants to be established in terms of lowerorder determinants.
The matrix inverse of A is obtained by the formula
A1 =
T
1
1
adj ( A) = Co( A)
A
A
Properties:
(A.B)-1= B-1.A-1
(A-1)-1=A
(At)-1=(A-1)t
Given a matrix A we call minors of order k (km,n) of A the
determinants of the sub-matrices constructed with elements
belonging to k rows and k columns of A.
We call Rank of A, rk(A) the order of the biggest minor of A
different from zero.
Determinant Properties
1.
2.
3.
4.
5.
6.
|A.B| = |A|.|B|; kN, k0, |Ak|=|A|k ; |A-1|= 1/|A|
|aA | = an|A|
If A has a row (or a column) with all numbers equal to 0, |A|=0
If A has two rows (or two columns) proportional or equal, then
|A|=0
Given a matrix A, when two rows (or two columns) are
exchanged, then the determinant of the obtained matrix is
equal to -|A|
If in a matrix A, to one row (or column) we add a linear
combination of other parallel rows (or columns), the
determinant of the obtained matrix is |A|
Definition of Vector
A matrix A with only one row is also called a Row Vector, and a
matrix with one column is called a Column Vector. Often they
are denoted by small bold letters
a = [a1 , a2 ,..., an ] = a
The numbers ai are called the components (or coordinates) of
the vector..
k vectors a1 = [a11 , a12 ,..., a1n ], a2 = [a21 , a22 ,..., a2 n ],..., ak = [ak1 , ak 2 ,..., akn ]
are said to be independent if there is a minor of order k of the
matrix
different from 0.
a11 a12 ... a1n
a
A = 21
ak 1
a22 ... a2 n
ak 2 akn
Norm or Length of a vector.
[
]
v
v
v
v
v
=
,
,...,
=
Given a vector
n
1 2
we define the Length or the Norm of
V, that we will denote as V as
v = v12 + v22 + ... + vn2
In fact, the norm V is the distance
from the the origin (0,0,,0) to the
Point (v1 , v2 ,..., vn )
The inner product of two vectors
a = [a1 , a2 ,..., an ] and b = [b , b ,..., b ]
1 2
n
is defined as
a.b = [a1 , a2 ,..., an ][b1 , b2 ,..., bn ] = a1b1 + a2b2 + ... + anbn
SYSTEMS of LINEAR EQUATIONS
We will call a System of m Linear Equation with n
unknowns a system like
a11 x1 + a12 x2 + ... + a1n xn = b1
...................
am1 x1 + am 2 x2 + ... + amn xn = bm
This system can be written as
a11
a
21
am1
a12 ... a1m x1 b1
a22 ... a2 m x2 b2
=
am 2 amn xn bm
or
A.X=B
where A is the coefficients matrix, X is the variable matrix,
and B is the independent term matrix.
Classification:
Determined Compatible System: One solution
Undetermined Compatible System: More than one solution
Incompatible System: No solutions.
If we call M= (A|B) the amplified matrix,
THEOREM (ROUCH-FRBENIUS)
The system AX=B has solution iff rg(A) = rg(M)
If rg(A)rg(M) then the system is Incompatible.
CASES:
1. rg(A)=rg(M) = m= n. Compatible system. The solution is
unique and we can apply the Cramers Rule:
x1
X = where xi =
xn
a11 ... b1 ... a1n
... ...
an1 ... bn ... ann
A
where the determinant in the numerator is obtained by substituting
the column i of A by the column of the independent terms.
2. rg(A)=rg(M)=n<m. It is possible to cancel the equations which
are linearly dependent of the others, and then we work as in the
case 1.
3. rg(A)=rg(M)=r<[Link] terms corresponding to variables that
we have not used in the study of the range pass to the second
term of the equations. Then, we will proceed as in cases 1 or 2.
GAUSSIAN ELIMINATION
METHOD
Make a staircase with 1 as the coefficient for each
non-zero leading entry.
Produce 0s below each leading entry.
The general solution is found by expressing the
unknowns that occur as leading entries in terms of
those unknowns that do not. The latter unknowns (if
there are any) can be chosen freely. The number of
unknowns that can be chosen freely (possibly 0) is
called the number of degrees of freedom for the
system
Solve the system
5x + 4y + 2z + w = 5
2x + 3y + z 2w = 3
5x 7 y 3z + 9w = 16
x 2y z + 4w = 10
Sol (1,0,-1,2)
Example:
2 x2 x3 =
7
2
x1 + x2 + 3 x3 =
3 x + 2 x + 2 x =
10
2
3
1
4
x + 2y + z =
5
x y+z =
2 x + 3 y z =
1
Sol: (2,-3,1)
