Midterm 2

Midterm 1 Review: Midterm 1 Study Guide

Section 2.2: The Inverse of a Matrix
Definitions
● Invertible Matrix
○ An n x n matrix A where
● Inverse of a Matrix
○ A-1 where
● Singular
○ Not invertible
● Determinant of a 2x2 matrix
○ ad - bc
● Elementary Matrix (E)
○ Matrix obtained by performing a single row operation on an identity matrix
○ Are invertible: inverse of an elementary matrix, E, is another elementary
matrix of the same type that transforms E back to I
○ All elementary matrices are invertible
● Row Equivalent Matrices
○ Matrices that can transform into each other through a sequence of
elementary row operations

Key Notes
● Invertible = nonsingular
● Not invertible = singular
● Inverse of a 2x2 Matrix

○
■ If (ad - bc) = 0, then A is not invertible
● Ax = b can be rewritten using inverses only if A is invertible

○
○ Of course, you can still use the row reduction method to solve Ax = b
○ For all b in Rn, x = A-1b is a unique solution
■ Invertible matrices have no free variables
■ Unique solution
● Product of n x n invertible matrices is invertible
○ Inverse of product is the product of the inverses in reverse order
 ● When an elementary row operations is performed on an m x n matrix A, the
resulting matrix can be written as EA
○ What if we had multiple elementary row operations on A?
■
● Method to find the inverse
○ Row reduce A to the identity matrix while performing the same row
operations on the identity matrix at the same time
○ [A | I] => [I | A-1]
● A matrix is invertible if and only if it is row equivalent to the identity
○ Pivots in every row and column (onto & one-to-one)

Section 2.3: Characterizations of Invertible Matrices

Definitions
● Linear Transformation
○ Mapping between two vector spaces (R n’s) that preserves all vector addition
& scalar properties
● Invertible Linear Transformation
○ Linear transformation T: Rn -> Rn is invertible if there is another linear
transformation S: Rn -> Rn such that:
■ S(T(x)) = x for all x in Rn
■ T(S(x)) = x for all x in Rn
○ Equivalent to saying:
■ A-1Ax = (I)x

Key Notes
● The IMT
 ○
● Let A and B be square matrices:
○ If AB = I, then A and B are both invertible
○ B = A-1 & A = B-1
● How to determine if a linear transformation is invertible?
○ Let a matrix A represent the linear transformation
○ If A is invertible, then the linear transformation is invertible

Section 2.4: Partitioned Matrices

Definitions
● Partitioned Matrix
○ Matrix divided up into separate blocks
● Block Diagonal Matrix
○ A partitioned matrix where all blocks except the main diagonal are 0’s
○ Is invertible if the main diagonal blocks are invertible

Key Notes
● Adding 2 partitioned matrices A and B
○ A and B must be the same size and partitioned in the exact same way
■ Add block by block
● Scaling partitioned matrices
 ○ Scale block by block
● Multiplying 2 partitioned matrices A and B (A*B)
○ Column partition of A must equal row partition of B
○ Number of columns in partition A = number of rows in partition B
■ Just like multiplying regular matrices
■ (2 x 2) * (2 x 1) => (2 x 1)
■ (3 x 4) * (4 x 1) => (3 x 1)
● Inverses of Partitioned Matrices

Section 2.5: Matrix Factorizations

Definitions
● Factorization of a matrix
○ Expression of a matrix as the product of two or more matrices
● Row interchanges
○ Swapping rows when row reducing
● Lower triangular matrix
○ Entries above the main diagonal are all 0’s
● Upper triangular matrix
○ Entries below the main diagonal are all 0’s
● Algorithm for an LU Factorization
○ 1. Reduce A to an echelon form U by a sequence of row replacement
operations, if possible.
○ 2. Place entries in L such that the same sequence of row operations reduces L
to I.

Key Notes
● LU Factorization
○ Why do we use it?
■ More efficient to solve a sequence of equations with the same
coefficient matrix (Ax = b1, Ax = b2, ... , Ax = bn) by LU factorization than
row reducing the equations every single time
● Let A be an m x n matrix that can be row reduced to echelon form without row
exchanges, then:
 ○
■ L: m x m lower triangular matrix with one’s on the main diagonal
■ U: m x n echelon form of A
● Rewriting Ax = b using A = LU

○
○ Ax = b -> L(Ux) = b
● The LU Factorization Algorithm
○ How do we get U?
■ Row reduce A to echelon form using only row replacements that add a
multiple of one row to another below it
○ How do we get L?
■ Take the row replacement operations you did on A when getting
echelon form
● Basically: find the elementary matrices that transform A into U
■ Then, reverse the signs and input them in their respective spots in
the m x m identity matrix
● Replace the 0’s below the main diagonal with the row
replacement “coefficients”
● Basically: after finding all the elementary matrices, take their
inverses

○
● Using the LU Decomposition
○ After constructing A = LU, solve Ax = LUx = b by:
■ 1. Forward solve for y in Ly = b
● R1(x) + R2 -> R2
● Modify rows below using above rows
 ■ 2. Backwards solve for x in Ux = y
● R2(x) + R1 -> R1
● Modify rows above using below rows

Section 2.6: The Leontief Input-Output Model

Definitions
● Production vector in Rn (x)
○ Lists the output of each sector for one year
● Final demand vector (d)
○ Lists the value of goods and services produced for the consumers
(nonproductive part of the economy)
● Intermediate demand (Cx)
○ The demand for goods and services that the producers (sectors) need as
inputs for their own production
■ Ex: electricity sector needs inputs from the water sector and vice
versa
● Consumption matrix (C)
○ How much each sector consumes from other sectors in terms of percentages
● Column sum
○ The sum of the entries in a column

Key Notes
● The Leontief Input-Output Model (Production Equation)

○
○ Can be rewritten as:
■ (I - C)x = d
● Solve for x (amount produced) by row reduction
■ x = (I - C)-1 * d
● Solve for x (amount produced) by multiplying
● For a good economy, the column sum of each sector should be less than 1
○ A sector should in general require less than one unit’s worth of inputs to
produce one unit of output
● Output vector (x)
○ xi: entry i of vector x
■ Number of units produced by sector i
 ● Internal consumption (C)
○ 2 equivalent ways of defining entries of C where an entry is c i, j:
■ Sector i sends a proportion of its units to sector j
■ Sector j requires a proportion of the units created by sector i
● Consumption matrix (Cx)
○ Total output for each sector (per one unit) is the sum of the columns for each
sector
● A Formula for (I - C)-1
○ As an economy is introduced to a demand vector, the equation starts off as:
■ x=d
○ However, production will require intermediate demand from other sectors,
and then that intermediate demand will require more inputs from even more
sectors
■ x = d + Cd + C2d + C3d + …
● => (I + C + C2 + C3)d

○
■ We can approximate (I - C)-1 by making m as large as possible
● Add as many intermediate demands as we can
● Economic Importance of Entries in (I - C)-1
○ Entries used to predict how the production x will have to change when the
final demand d changes
■ Remember: x = (I - C)-1 * d
○ The entries in each column of (I - C)-1 are the increased amounts each sector
has to produce in order to satisfy an increase of 1 unit in the final demand

Section 2.7: Applications to Computer Graphics

Definitions
● Homogeneous coordinates
○ Each point (x, y) in R2 can be identified with the point (x, y, 1) on the plane in
R3 that lies one unit above the xy - plane
● Composite transformations
○ Multiplication of 2 or more basic transformations

Key Notes
● Why do we use homogeneous coordinates?
○ Translations are not linear transformations
 ● Homogeneous coordinates are allowed to be scalars
○ (3, 5, 1) = (6, 10, 2)
● (x, y) -> (x + h, y + k)
○ Translation cannot be represented by an R 2 matrix multiplication
○ (x, y, 1) -> (x + h, y + k, 1)

○
■ Translation not possible if we used a 2x2 identity matrix
● Linear transformations in R2 represented with homogeneous coordinates are
written as partitioned matrices:

○ where A is a 2x2 matrix

○ Examples

■
● Composite Transformations
○ “Add” on more transformation matrices to the left of the other
transformations
■ First transformation is always the rightmost (modifies the x vector
first)
● Homogeneous 3D Coordinates
○ (X, Y, Z, H) are homogeneous coordinates for (x, y, z) if H ≠ 0
and

Section 2.8: Subspaces of Rn

Definitions
● Subset of Rn
○ Any collection of vectors that are in R n
● Subspace of Rn
○ A subset H in Rn that has 3 properties:
■ The zero vector is in H
 ■ (closed under addition)
■ (closed under scalar multiplication)
○ Subspace can be written as the Span{} of some amount of linearly
independent vectors
● Column Space of a Matrix A (m x n)
○ Col A: the subspace of Rm spanned by {a1 , … , an}
○ Essentially all the pivot columns
● Null Space of a Matrix A (m x n)
○ Null A: the subspace of Rn spanned by the set of all vectors x that solve Ax = 0
● Basis for a Subspace H of Rn
○ A linearly independent set in H that spans H
■ DOES NOT CONTAIN THE ZERO VECTOR (BECAUSE IT IS
LINEARLY INDEPENDENT) UNLIKE THE SPAN
● Standard Basis for Rn
○ {e1 , … , en}

Key Notes
● If v1 and v2 are in Rn and H = Span{v1 , v2}, then H is a subspace of Rn
○ v1 and v2 must be in Rn for this relation to work
● For v1, … , vp in Rn, the set of all linear combinations of v1, … , vp is a subspace of Rn
○ Span{v1 , … , vp} = subspace spanned by v1 , … , vp
● Is b in the column space of A?
○ Same as : Is b a linear combination of A?
○ Same as : Is b in the Span of A?
● Is H a subspace of Rn?
○ Basically asking if H has n linearly independent columns
○ Does H have no free variables?
● Subspaces vs. Bases
○ Subspaces => Span{v1 , … , vn}
■ Includes the 0 vector
○ Bases => {v1 , … , vn}
● Defining a basis for column space A
○ Number of entries for each vector = number of rows in matrix A
○ Number of vectors in the basis = number of pivot columns
○ What vectors can you include in the basis?
■ Scalar multiples
■ The identity matrix columns only if every column is pivotal in A
● Finding the Column Space
○ Row reduce the matrix
 ■ Row operations do not affect linear dependence relations
○ Determine the pivot columns
○ Create a basis/subspace using the pivot columns in the original matrix
■ Not the row reduced one
○ If every column is linearly independent, then the elementary vectors are
included in the column space
■ Linear combinations of elementary vectors can get you any column of
the original matrix
● Finding the Null Space
○ Determine all the free variables
○ Rewrite system in parametric vector form
○ Vectors created in parametric vector form generate the null space

Section 2.9: Dimension and Rank

Definitions
● Coordinates
○ Weights that map our vectors to get to some point in the span of the vectors
● Coordinate Vector

○
● Dimension of a Subspace
○ dim H: the number of vectors in a basis of H
○ dim{0} = 0
● Rank of a Matrix A
○ Dimension of the column space of A
○ Number of pivots in A

Key Notes
● Why we choose to write bases:
○ Each vector in H can be written in only one way as a linear combination of
the basis vectors
● A plane through 0 in R3 is two-dimensional
○ 3x3 matrix A has 2 pivots
● A line through 0 in R2 is one-dimensional
○ 2x2 matrix A has one pivot
● Any two choices of bases of a non-zero subspace H have the same dimension
 ○ dim Rn = n
○ dim(Col A) = number of pivots
○ dim(Null A) = number of free variables
● dim(Col A) = rank A
● Rank Theorem
○ If A has n columns, then:
■ rank A + dim(Null A) = n
○ Number of pivots + number of free variables = number of columns
● Basis Theorem
○ Any two bases for a subspace have the same dimension (cardinality)
○ Many choices for the basis of a subspace
● Continuation of the Invertible Matrix Theorem with Rank

Section 3.1: Introduction to Determinants

Definitions
● Ai j submatrix
○ Delete the ith row and jth column of matrix A
○ Remaining elements will form the new submatrix
● Determinant for a 2x2

○ A= -> det A = ad - bc
● Cofactor expansion
○ A way to solve determinants for square matrices that are 3x3 and greater

Key Notes
● Signs of cofactor expansions
○ Depends on position of element ai j in the matrix
 ○
● Shortcut for finding the determinant
○ Row reduce to REF
■ Effects of row operations on determinant covered in 3.2
○ Multiply all the numbers on the main diagonal

Section 3.2: Properties of Determinants

Definitions
● Column Operations
○ Same effect on determinants as row operations
○ This is true because the determinant of A = determinant of A T (transpose)

Key Notes
● Row operations on determinants
○ Row replacement: nothing
○ Row swap: multiply determinant by negative one
○ Row scale: multiply determinant by scale
● Summary of elementary matrices’ determinants

○
● More specific example of row scaling on determinants

○
○ Row divided by k
■ Determinant is multiplied by 1/k
● If A is invertible (every column is pivotal)
○ det A ≠ 0
● If A is not invertible
○ det A = 0
○ At least one entry on the main diagonal of REF is 0
 ●
● When A is not invertible, the rows are linearly dependent
○ If A is square, then the columns are also linearly dependent
● det A = det AT
● det AB = (det A)(det B)
● det A-1 = 1 / (det A)

Section 3.3: Volume and Linear Transformations

Definitions
● Parallelepiped: a parallelogram in Rn where n > 2

Key Notes
● If A is a 2x2 matrix:
○ Area of the parallelogram determined by the columns of A is | det A |
● If A is a 3x3 matrix:
○ Area of the parallelepiped determined by the columns of A is | det A |
● Row/column swaps and replacements do not affect the absolute value of the
determinant
● Linear transformations on a parallelepiped
○ Area of T(S) = | det A | * { area of S }
■ T: linear transformation determined by matrix A
■ S: parallelogram

Section 4.9: Applications to Markov Chains

Definitions
● Probability vector
○ A vector with nonnegative entries that sum to 1
● Stochastic matrix
○ A square matrix whose columns are probability vectors
● Markov Chain
○ A sequence of probability vectors {x0, x1, x2, …} together with a stochastic
matrix {P} such that:
■
● Steady State Vector
○ A probability q such that Pq = q
 ○ Every stochastic matrix has a steady state vector
● Regular stochastic matrix
○ Stochastic matrix where some power of it will contain only strictly positive
entries
■ Pk where all entries > 0

Key Notes
● How to find the next outcome of a Markov Chain?
○ Simply multiply P by xk to find xk+1
● How to find a steady state vector?

○
○ After finding a basis for the null space of (P - I) q = 0, remember to make sure
that the column sum is 1
■ Steady state vector is a probability vector
● The initial state has no effect on the long term behavior of the Markov Chain

Section 5.1: Eigenvectors and Eigenvalues

Definitions
● Eigenvector of an n x n matrix A:
○ Nonzero vector x such that Ax = λx for some scalar λ
● Eigenvalue of A:
○ A scalar λ where there is a nontrivial solution x of Ax = λx
● Eigenspace of an eigenvalue
○ Contains the zero vector and all eigenvectors corresponding to λ

Key Notes
● Determine if a vector x is an eigenvector
○ A*x => see if product is a scalar multiple of x
● Finding the eigenvector from an eigenvalue (7)
○ Solve (A - 7I)x = 0
○ Then, do the parametric vector form of what you have left
● Finding the eigenvalue λ
○ Solve (A - λI)x = 0 for a nontrivial solution
○ Find the set of all solutions to the null space of (A - λI)
● Eigenvalues of a triangular matrix are the entries on the main diagonal
● 0 is an eigenvalue of A if and only if A is not invertible
○ Ax = 0x
 ○ Ax = 0: x is a nontrivial solution if A is not invertible
● Eigenvectors that correspond to distinct eigenvalues are linearly independent
○ Opposite is not always true

■ : eigenvectors are linearly independent but have the same

eigenvalue

Section 5.2: The Characteristic Equation

Definitions
● The Characteristic Polynomial:
○ det(A - λI)
● The Characteristic Equation
○ det(A - λI) = 0
● Trace
○ Sum of the diagonal entries in a matrix
● Algebraic Multiplicity of an Eigenvalue
○ The number of times the eigenvalue shows up as roots of the characteristic
polynomial
● Geometric Multiplicity of an Eigenvalue
○ The dimension of Null (A - λI) for a given eigenvalue λ

Key Notes
● How to find eigenvalues?
○ Solve (A - λI)x = 0 for a nontrivial solution
○ Find the set of all solutions to the null space of (A - λI)
● Continuation of IMT
○ For A: n x n matrix, A is invertible if and only if:
■ The number 0 is not an eigenvalue of A
■ The determinant of A is not 0
● Finding the characteristic polynomial using trace and determinant for a
characteristic polynomial of 2
○ λ2 - λ(trace) + det A
● Warnings:
○ Cannot determine the eigenvalues of a matrix from its reduced from
○ Row operations change the eigenvalues

Theorems
Chapter 2
 Theorem 4: Finding the Inverse of a 2x2 Matrix

Theorem 5: Alternate Method of Finding the Solution Set

Theorem 6: Properties of Invertible Matrices

Theorem 7: Finding the Inverse of a Matrix

Theorem 8: The Invertible Matrix Theorem

Theorem 9: Invertible Linear Transformations

Theorem 10: Column-Row Expansion of AB

 Theorem 11: Solving the Output Vector (x)

Theorem 12: Finding the Null Space of Matrix A

Theorem 13: Determining the Column Space of Matrix A

Theorem 14: The Rank Theorem

Theorem 15: The Basis Theorem

Chapter 3
Theorem 1: Cofactor Expansion to find Determinants
 Theorem 2: Shortcut to Computing Determinant

Theorem 3: Row Operations on Determinants

Theorem 4: IMT DLC: Determinant

Theorem 5: Transpose Equivalence for Determinants

Theorem 6: Multiplicative Property of Determinants

Theorem 9: Determinants as Area and Volume

Theorem 10: Linear Transformations on Area/Volume

Chapter 4
Theorem 18: Long-term Behavior of a Markov Chain

Chapter 5
Theorem 1: Eigenvalues of a Triangular Matrix

Theorem 2: Eigenvectors for Distinct Eigenvalues