+ All Categories
Home > Documents > MM2 Linear Algebra 2011 I 1x2

MM2 Linear Algebra 2011 I 1x2

Date post: 07-Apr-2018
Category:
Upload: aaron-lee
View: 221 times
Download: 0 times
Share this document with a friend

of 33

Transcript
  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    1/33

    Mathematics

    Part 1: Review of Matrices & Elements of Linear Algebra

    Part 2: Multivariable Calculus

    Review of Matrix Algebra

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    2/33

    What is a Matrix?

    A matrix is a rectangular array of numbers.

    A =

    1 2 09 1 2

    1 3 4

    ; M = 1 2 5 7

    4 1 0 8

    ; v =

    13

    26

    Order of a matrixA matrix has order n m if it has n rows and m columns.

    Vectors

    DefinitionA (1 m) matrix is called as a row vector

    v = 1 3 2 6 A (m 1) matrix is called as a column vector

    v =

    13

    26

    By default, when we talk about a vector we mean a column vector.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    3/33

    Matrices

    The elements of a matrix A are often written aij where

    i is the row number of the element

    j is the column number of the element

    Example

    A = 1 2 09 1 21 3 4

    = a11 a12 a13a21 a22 a23a31 a32 a33

    Matrices

    A general m n matrix will often be written

    A =

    a11 a12

    a1n

    a21 a22 a2n...

    .... . .

    ...am1 am2 amn

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    4/33

    Notations

    Square or round brackets. Different notations... same meaning.

    A =

    1 2 09 1 21 3 4

    A = 1 2 09 1 21 3 4

    Recallthe following notation has a different meaning (determinant):

    det(A) =

    1 2 09 1 21 3 4

    Matrices

    An obvious fact: Two matrices are said to be equal if they havethe same order and all their elements are equal.

    A =

    1 2 09 1 2

    1 3 4

    ; B =

    1 2 09 1 2

    1 5 4

    ; C =

    1 2 0 19 1 2 3

    1 3 4 6

    A = B and A = C.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    5/33

    Types of Matrices

    An upper triangular matrix has the form

    a11 a12 a1n0 a22 a2n

    ... ... . . . ...0 0 ann

    An lower triangular matrix has the form

    b11 0 0b21 b22 0

    ......

    . . ....

    bn1 bn2 bnn

    An diagonal matrix has the form

    d11 0

    0

    0 d22 0...

    .... . .

    ...0 0 dnn

    Matrix addition and subtraction

    Two matrices may be added or subtracted if they have the sameorder.

    2 11 0

    1 3

    +1 0

    2 13 1

    =

    3 13 12 4

    1 3 6 11 1 2 1

    1 1 1 11 1 1 1

    =

    0 2 5 0

    2 0 1 0

    2 11 0

    1 3

    +

    1 3 6 1

    1 1 2 1

    not defined

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    6/33

    Matrix addition and subtraction

    Matrices are commutative and associative under addition:

    A + B = B + A; A + (B + C) = (A + B) + C

    A zero matrix is any matrix with all elements equal to zero,and isusually written 0:

    0 00 00 0

    ;

    0 0 00 0 00 0 0

    A + 0 = 0 + A = A

    Matrix Multiplication

    It is straightforward to multiply matrices by a number k:

    k1 2 0

    9 1 21 3 4

    = k 2k 0

    9k k 2kk 3k 4k

    Multiplying two matrices together is more complicated.

    1 2 00 1 21 3 4

    1 0 11 1 32 0 2

    =?

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    7/33

    Matrix Multiplication

    We define the product of a row vector and a column vector asbeing the sum of the products of their components. e.g:

    1 2 0

    112

    = 1 1 + 2 1 + 0 2 = 1

    To multiply two matrices we multiply the rows of the first one by

    the columns of the second one.

    Matrix Multiplication

    1 2 00 1 2

    1 3 4

    1 0 11 1 3

    2 0 2

    =

    1

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    8/33

    Matrix Multiplication

    1 2 00 1 2

    1 3 4

    1 0 11 1 3

    2 0 2

    =

    1 2

    Matrix Multiplication

    1 2 00 1 2

    1 3 4

    1 0 11 1 3

    2 0 2

    =

    1 2 5

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    9/33

    Matrix Multiplication

    1 2 00 1 2

    1 3 4

    1 0 11 1 3

    2 0 2

    =

    1 2 55

    and so forth. . .

    Matrix Multiplication

    1 2 00 1 2

    1 3 4

    1 0 11 1 3

    2 0 2

    =

    1 2 55 1 7

    12 3 18

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    10/33

    Rule for matrix multiplication:

    The number of columns of the first matrix

    must equal the number of rows of the

    second.

    Matrix Multiplication

    When you multiply 2 matrices the order matters

    Matrix multiplication is associative

    A(BC) = (AB)C

    but is non-commutative, i.e., in general

    AB = BA

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    11/33

    Matrix Multiplication

    Because in general AB = BA, we must always specify on whichside we are doing the matrix multiplication:

    X =

    2 01 1

    A =

    1 03 1

    Multiplying X on the left by A:

    AX =

    1 03 1

    2 01 1

    =

    2 05 1

    Multiplying X on the right by A:

    XA =

    2 01 1

    1 03 1

    =

    2 04 1

    Matrix Multiplication

    Matrix multiplication has some other strange properties:

    AB = 0 does not necessarily mean that A = 0 or B = 01 10 0

    2 0

    2 0

    =

    0 00 0

    AD = AC does not necessarily mean that D = C

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    12/33

    Matrix Multiplication

    The product of two matrices is a matrix whose elements can bewritten in a compact form as:

    [AB]ij =Nk=1

    aikbkj

    where N is the number of columns of A and the number of rows

    of B.

    Identity Matrix

    The identity matrix I is the matrix with the property

    AI = IA = A

    I =

    1 00 1

    for 2 2 matrices

    I =

    1 0 00 1 0

    0 0 1

    for 3 3 matrices

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    13/33

    Inverse of a Matrix

    The inverse of a square matrix A is written A1 and has theproperty

    A1A = AA1 = I

    A matrix only has one inverse

    This means that if it exists, the inverse of a matrix is unique.

    Inverse of a Matrix

    Inverse of a 2 2 matrix.

    if A =

    a bc d

    , A1 =

    1

    ad bc

    d b

    c a

    Remarks:

    ad bc is the determinant of the matrix.

    The inverse is defined only if det(A) = ad bc = 0.

    If det(A) = 0, A is said to be a singular matrix.

    For larger square matrices there is a more general way tocompute inverses. Well see that shortly.

    The sum of the diagonal elements of A is called the trace ofA: tr(A) = a + d.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    14/33

    Transpose of a Matrix

    The transpose of a matrix is obtained by interchanging the rowsand the columns:

    If A =

    1 3 6 1

    1 1 2 1

    , AT =

    1 13 16 21 1

    If A and B have the same order

    (A + B)T

    = AT

    + BT

    (AT)T = A

    (AB)T = BTAT

    Matrix Algebra

    Using the rules for addition, subtraction, multiplication andinverses, as well as the special matrices I and 0, we can re-arrangematrix equations.

    Example

    Given

    A =

    1 02 1

    ; B =

    1 22 1

    ; C =

    0 00 1

    Rearange the equation

    A + 2BX = C

    to find X.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    15/33

    Matrix Algebra

    Formally we have

    A + 2BX = C

    2BX = C A

    BX =1

    2(C A)

    X =1

    2B1(C A) (multiply on the left)

    B =1 2

    2 1

    B1

    =1

    1 4 1 2

    2 1

    = 1

    5 1 2

    2 1

    C A =

    1 02 0

    X =

    1

    10

    1 2

    2 1

    1 02 0

    =

    1

    10

    3 04 0

    Matrix Algebra

    Example

    Given

    B =

    1 11 1

    ; C =

    4 20 1

    Rearange the equation

    X+ XB = C

    to find X.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    16/33

    Matrix Algebra

    X+ XB = C

    X(I+ B) = C

    X = C(I + B)1 (not (I + B)1C)

    I + B =

    2 11 2

    (I + B)1 =1

    3 2 1

    1 2 X =

    1

    3

    4 20 1

    2 1

    1 2

    =

    1

    3

    6 01 2

    Linear Systems

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    17/33

    Linear system of equations

    A system of equations is a set of mathematical statementscontaining several unknown quantities.

    If unknowns can be found so that all the statements aresatisfied then we say that the system has a solution.

    There are three possibilities when trying to find the solution:

    1 There is too little information to find the unknowns.

    2

    There is too much information to find the unknowns.3 There is exactly the right amount of information.

    Linear system of equations

    Example

    From the two statements

    1

    There are twice as many sheep as people in Australia.2 There are 20 million more sheep than people in Australia.

    we can construct the system of equationss = 2p

    s = p + 20 p = 20, s = 40 (in millions)

    Unique solution

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    18/33

    Linear system of equations

    Example

    From the two statements

    1 There are 20 million more sheep than people in Australia

    2 There is the same number of sheep as people in Australia

    we can construct the system of equations

    s = p + 20s = p

    No solution The two statements are inconsistent

    Linear system of equations Graphical interpretation

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    19/33

    Linear system of equations

    An equation for several unknowns x1, x2 . . . xn is linear if it can beput in the form

    a1x1 + a2x2 + + anxn = b

    where the coefficients a1 . . . an and b are constants.A set of linear equations is known as a linear system.

    Linear system of equations

    A linear system of m equations in n unknowns is a set of equationsof the form

    a11x1 + a12x2 + + a1nxn = b1

    a21x1 + a22x2 +

    + a2nxn = b2...

    am1x1 + am2x2 + + amnxn = bm

    A solution is a set of numbers (x1, x2, . . . , xn) that satisfiesall m equations.

    This set of numbers can be gathered into a vector.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    20/33

    Linear system of equations

    The augmented matrix associated with the system is

    a11 a12 a1n b1a21 a22 a2n b2

    ......

    . . ....

    am1 am2 amn bm

    A system with zeros below the main diagonal is said to be in

    row-echelon form.A system in row-echelon form can be solved easily byback-substitution.

    Linear system of equations

    Example

    1 2 1 110 3 1 250 0 2 70

    x1 + 2x2 + x3 = 11

    3x2 + x3 = 252x3 = 70

    From the third row we get: 2x3 = 70 x3 = 35

    From row 2 we get: 3x2 + x3 = 25 x2 =10

    3Finally we get: x1 + 2x2 + x3 = 11

    x1 + 203

    + 35 = 11 x1 = 923

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    21/33

    Row-echelon form

    A matrix is in row echelon form ifAll rows that contain only zeros are grouped at the bottom ofthe matrix.

    In any two consecutive non-zero rows, the leftmost non-zeroentry in the lower row occurs farther to the right than theleftmost nonzero entry in the upper row.

    We call a row (column) nonzero if it contains at least one nonzero

    entry.

    Row-echelon form

    A matrix in row-echelon form has a staircase pattern like thefollowing matrix:

    0 0 0 0 0 0 0 0 0 0 0 0

    where the greek letters correspond to non-zero values and the starsmay or may not be zero.

    For the augmented matrix in row echelon form, each leftmostnonzero entry corresponds to a basic variable.

    For instance, for the matrix above the basic variables are x1,x2 and x5.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    22/33

    Row-echelon form

    The row-echelon form can give us some important informationabout a system. If a system is consistent (i.e., if it has a solution)

    if all variables are basic variables the solution is unique.

    if there is at least one non-basic variable there are infinitelymany solutions.

    Row-echelon formThe number of solutions (unique, none, infinitely many) can befound by looking at the augmented matrix in row-echelon form:

    a11 a12 a1n b1a21 a22 a2n b2

    ......

    . . ....

    0 0

    amn bm

    unique solution if a11, a22, . . . , amn = 0

    a11 a12 a1n b1a21 a22 a2n b2

    ......

    . . ....

    0 0 0 bm

    no solution if bm = 0

    a11 a12 a1n b1a21 a22 a2n b2

    .

    .....

    .. .

    .

    ..0 0 0 0

    infinitely many solutions

    The number of non-zero rows in row-echelon form is the rank of the

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    23/33

    Gauss elimination method for solving a linear system of

    equations

    Any system of equations is left unchanged by:1 Multiplying a row by a non-zero scalar.

    2 Adding or subtracting two rows.

    3 Swapping any two rows.

    These are known as row operations.

    In Gauss Elimination, we use row-operations to change the

    system to row-echelon form.We can then solve the system by back-substitution.

    Gauss elimination

    Use Gauss elimination to solve the system

    x1 + x2 + 2x3 = 23x1 x2 + x3 = 6

    x1 + 3x2 + 4x3 = 4

    The augmented matrix is 1 1 2 23 1 1 6

    1 3 4 4

    We need to use row operations to put the augmented matrix inrow-echelon form.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    24/33

    Gauss elimination

    1 1 2 23 1 1 6

    1 3 4 4

    Example of a linear system: Balancing a chemical reaction

    Consider the following chemical reaction:

    C2H6 + O2 CO2 + H2O

    Balancing the reaction means finding positive integers n1, n2, n3,and n4 such that

    n1C2H6 + n2O2 n3CO2 + n4H2O

    has the same number of atoms of each element on each side of theequation.

    We can cast the problem as a linear system.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    25/33

    Example of a linear system: Balancing a chemical reaction

    n1C2H6 + n2O2 n3CO2 + n4H2O

    Inverse of a Matrix

    We saw how to find the inverse of a 2 2 matrix.

    If we can find the inverse of a n n matrix we could solve ageneral system

    AX = b X = A1b

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    26/33

    Inverse of a 3 3 Matrix. Method 1: row operations

    1 Write the LHS of the system in matrix form.

    1 2 21 2 13 2 1

    2 Augment the matrix with the identity matrix.

    1 2 2 1 0 01 2 1 0 1 0

    3 2 1 0 0 1

    3 Use row operations to transform the left half into the identity.The right half will then be the inverse.

    Inverse of a 3 3 Matrix. Method 1: row operations

    Example

    Solve

    x1 2x2 + 2x3 = 1

    x1 2x2 + x3 = 1

    3x1 2x2 + x3 = 1

    by computing the inverse of the matrix associated with the linearsystem.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    27/33

    Inverse of a 3 3 Matrix. Method 1: row operations

    The augmented matrix is

    1 2 2 1 0 01 2 1 0 1 03 2 1 0 0 1

    R1 R1 2R3R3 R1 R2

    5 2 0 1 0 21 2 1 0 1 00 0 1 1 1 0

    R2 R2 R3

    5 2 0 1 0 21 2 0 1 2 0

    0 0 1 1 1 0

    R1 R1 + R2R2 5R2 +R1

    4 0 0 0 2 20 8 0 4 10 2

    0 0 1 1 1 0

    R1 R1/4R2 R2/8

    1 0 0 0 1/2 1/20 1 0 1/2 5/4 1/4

    0 0 1 1 1 0

    Inverse of a 3 3 Matrix. Method 1: row operations

    Therefore the inverse is

    0 1/2 1/21/2

    5/4 1/41 1 0

    and the solution of the system is

    x1x2x3

    =

    0 1/2 1/21/2 5/4 1/4

    1 1 0

    11

    1

    =

    01/2

    0

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    28/33

    Cramers rule for solving a linear system

    Given the system

    a11 a12a21 a22

    x1x2

    =

    b1b2

    the solution can be expressed in terms of determinants:

    x1 =

    b1 a12b2 a22

    a11 a12a21 a22

    ; x2 =

    a11 b1a21 b2

    a11 a12a21 a22

    Of course this requires that the determinant of the system isnon-zero.

    Cramers rule for solving a linear system

    Cramers rule works for higher order systems:

    a11 a12 a13a21 a22 a23

    a31 a32 a33

    x1x2

    x3

    =

    b1b2

    b3

    x1 =

    b1 a12 a13b2 a22 a23b3 a32 a33

    a11 a12 a13a21 a22 a23a31 a32 a33

    ; x2 =

    a11 b1 a13a21 b2 a23a31 b3 a33

    a11 a12 a13a21 a22 a23a31 a32 a33

    ; x3 =

    a11 a12 b1a21 a22 b2a31 a32 b3

    a11 a12 a13a21 a22 a23a31 a32 a33

    But how do we compute a general determinant?

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    29/33

    Determinant

    There are two (main) methods for calculating the determinant of an n matrix:

    1 Cofactor expansion.

    2 Row operations.

    Determinant Cofactor expansion

    Steps to expand using cofactors with respect to a given row:

    1 Multiply each coefficients of this row by (1)i+j where i is itsrow number and j its column number.

    2 Multiply each coefficients of this row with the determinant of

    the minor matrix i.e., the matrix obtained by deleting thecoefficients row and column.a11 a12 a13a21 a22 a23a31 a32 a33

    = a11a22 a23a32 a33

    a12a21 a23a31 a33

    + a13a21 a22a31 a32

    The cofactors are the signed determinants of the small matrices.You can perform the cofactor expansion with respect to any row orcolumn.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    30/33

    Determinant Cofactor expansion

    Calculate the determinant of

    1 2 21 2 1

    3 2 1

    Inverse of a 3 3 Matrix. Method 2: Cofactors

    1 Calculate the determinant of the matrix det(A).

    2 If det(A) = 0 the inverse is given by

    A1 =1

    det(A)CT

    where C is the matrix of the cofactors of A.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    31/33

    Determinant Row operations

    Interchanging of two rows multiplies the determinant by -1.

    Adding a multiple of one row to another leaves thedeterminant unchanged.

    Multiplying an entire row by a number k multiplies thedeterminant by k.

    These rules also hold for column operations.

    Determinant Useful tips

    Things to look for before trying to calculate a determinant

    If any row or column contains only zeros then the determinantis zero.

    If any two rows or columns are identical then the determinantis zero.

    If any two rows or columns are multiples of each other thenthe determinant is zero.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    32/33

    Determinant Special case

    To find the determinant of a matrix in row-echelon form, multiplydown the diagonal

    1 2 30 5 20 0 1

    = 1 5 1 = 5

    General properties of determinant

    det(A + B) = det(A)+det(B).

    det(AB) = det(A)det(B).

    det(A)det(A1)=1.A has an inverse if and only if det(A) = 0.

    Steps to evaluate determinants using row reduction:

    1 Reduce the determinant to row-echelon form (keeping track ofrow-swaps and multiplication by numbers).

    2 Multiply the coefficients along the diagonal.

  • 8/4/2019 MM2 Linear Algebra 2011 I 1x2

    33/33

    Linear Systems Summary

    We have seen 3 methods to solve a linear system

    1 Gauss elimination.

    2 Using the inverse matrix.

    3 Cramers rule.


Recommended