MS&E 121
Introduction to Stochastic Modeling
January 3, 2005
EXCEL TUTORIAL
This tutorial will introduce you to some essential features of Excel that we will be using throughout
MS&E 121 to solve linear systems of equations, analyzing data, and simulating random variables. You
will learn how to
• use some of the functions commonly used in math/statistics,
• represent a linear system of equations in an Excel worksheet,
• use the matrix functions MMULT and MINVERSE to solve linear systems of equations, and
• generate various types of random variables.
For additional Excel assistance, try consulting the Excel Help files or numerous online resources.
Parts of this tutorial borrow from Albert Whangbo’s Excel Solver Tutorial.
1 Commonly used Excel functions
In many probability applications we will commonly need to analyze data by computing quantities such
as mean, variance, standard deviation, correlation, confidence intervals, etc., and Excel provides us with
easy to use functions to do such things. All the functions can be accessed through the insert function
button, fx, or directly by typing them on the formula bar.
Naming Vectors and Matrices in Excel
Excel allows you to name ranges of cells in the worksheet. This feature is especially convenient for doing
matrix calculations and for setting up linear systems of equations. To name a range of cells, select the
entire range with the mouse and use the Insert → Name → Define... menu. Alternatively, select the
range and enter its name in the Name Box adjacent to the Formula Bar.
General use functions
One of the most useful functions is SUM(range), which sums the numbers in an array of cells. This
array can either be a vector (vertical or horizontal) or a matrix. This function can also be used by
highlighting first the desired vector of numbers and then pressing the Σ button, in which case the
answer is automatically placed at the end of the vector.
Two other useful functions are SUMPRODUCT(range1, range2) and SUMSQ(range). The arguments of
SUM, SUMPRODUCT, and SUMSQ can be either ranges of cells or range names. For a pair of ranges named
S and T, entering SUMPRODUCT(S,T) into the Formula Bar returns
∑
i,j
sijtij ,
Excel Tutorial
1
MS&E 121
Introduction to Stochastic Modeling
January 3, 2005
i.e., the sum of the products of corresponding elements of S and T. Note that SUMPRODUCT will return
an error if the two arrays do not have the same dimensions. Entering SUMSQ(S) returns
∑
i,j
s2ij ,
i.e., the sum of the squared elements of S.
Statistical functions
The function AVERAGE(range) computes the arithmetic mean of the range of numbers, i.e.,
x̄ =
1
n
n∑
i=1
xi.
The function VAR(range) computes the sample variance of the range of numbers, i.e.,
S2 =
1
n− 1
n∑
i=1
(xi − x̄)2.
And the function STDEV(range) the sample standard deviation, i.e.,
S =
√
S2.
To compute the correlation between two data sets, use the function CORREL(range1, range2). If X
and Y are two vectors with elements xi and yi, respectively, then CORREL(X,Y) returns
n∑
i=1
(xi − x̄)(yi − ȳ)
√√√√ n∑
i=1
(xi − x̄)2
√√√√ n∑
i=1
(yi − ȳ)2
.
To close this section, it is important to mention that Excel counts with several cumulative distribution
functions, CDFs, and inverse CDFs for some of the most commonly used probability distributions. The
inverse CDFs are especially useful for computing confidence intervals, quantiles, and simulating random
variables through the inverse transformation method.
Excel Tutorial
2
MS&E 121
Introduction to Stochastic Modeling
January 3, 2005
Prob. Distribution
CDF
inverse CDF
Beta
BETADIST
BETAINV
Binomial∗
BINOMDIST
Chi Square
CHIDIST
CHIINV
Exponential∗
EXPONDIST
F
FDIST
FINV
Gamma∗
GAMMADIST
GAMMAINV
Hypergeometric
HYPGEOMDIST
Lognormal
LOGNORMDIST
LOGINV
Negative Binomial NEGBINOMDIST
Normal∗
NORMDIST
NORMINV
Standard Normal
NORMSDIST
NORMSINV
Poisson∗
POISSON
Student’s T
TDIST
TINV
Weibull∗
WEIBULL
The distributions marked with ∗ have the added feature that their CDF functions have an option to
evaluate both the CDF and the PDF (probability density function). For example,
EXPONDIST(3, 2, TRUE) = 1− e−2(3) → the CDF
EXPONDIST(3, 2, FALSE) = 2e−2(3)
→ the PDF.
The PDFs are specially useful for simulating random variables through the Acceptance/Rejection method.
2 Setting Up Linear Systems of Equations In Excel
In several applications throughout the quarter, we will be dealing with the problem of finding a solution
to a linear system of equations. A linear system of equations can be represented mathematically as
Ax = b
where the matrices A and b comprise the problem data. Our goal is to find a solution x to the above
equation.
Such systems of equations can have either
• one unique solution,
• infinitely many solutions, or
• no solution.
Excel Tutorial
3
MS&E 121
Introduction to Stochastic Modeling
January 3, 2005
In the examples that we will be dealing with, we will typically have n equations and n unknowns, that
is, the matrix A will have n rows and n columns. If the matrix A is invertible, then the solution to the
system of equations is unique and is given by
x = A−1b.
Entering Matrices into Excel
It is easy to display a linear system of equations in their natural way in Excel, i.e., as a range of cells.
For example, if A is a 5× 5 matrix, then each element of A can be entered into its corresponding cell
in a 5× 5 range of cells.
a11
a12
a13
a14
a15
a21
a22
a23
a24
a25
a31
a32
a33
a34
a35
a41
a42
a43
a44
a45
a51
a52
a53
a54
a55
Each matrix element occupies a cell in the worksheet.
To keep your worksheet organized, you might consider arranging the matrices as shown below.
A
x
b
A sample layout of a linear system of equations in an Excel worksheet.
Matrix Calculations: MMULT and MINVERSE
We will frequently use the MMULT and MINVERSE commands in Excel when solving systems of linear
equations.
Syntax
Description
MMULT(range1, range2 ) ordinary matrix multiplication
MINVERSE(range )
matrix inversion
Excel Tutorial
4
MS&E 121
Introduction to Stochastic Modeling
January 3, 2005
The arguments of MMULT, and MINVERSE can be either ranges of cells or range names. It is more concise
and descriptive to use names. For a pair of ranges named S and T, entering MMULT(S,T) into the
Formula Bar returns the matrix whose ijth element is
∑
k siktkj , i.e., normal matrix multiplication; and
entering MINVERSE(S) will return S−1.
MMULT(S,T) will return an error unless the number of columns of S equals the number of rows of T.
If you need to take the transpose of a matrix, use the TRANSPOSE function. Similarly, MINVERSE will
return an error if the range of cells does not constitute a square matrix, or if the matrix is not invertible.
In the first case the error will be #VALUE!, and in the second case #NUM!
Important note: If a matrix function returns a matrix larger a single element, you must first use the
mouse to select the cells that will contain the result, then type the matrix function into the Formula
Bar, and finally type CTRL SHIFT ENTER (COMMAND ENTER on a Mac) to evaluate the matrix function.
For example, if S and T are 3× 5 and 5× 4 matrices, respectively, then the result of MMULT(S,T) will
be a 3× 4 matrix. To perform this calculation and display the result in the worksheet, first highlight a
3× 4 range of cells, type MMULT(S,T) into the Formula Bar, then type CTRL SHIFT ENTER to display
the result in the highlighted area.
An Example
Consider the following system of linear equaqtions:
3x1 + 3x2 − 4x3 + x4 = 8
4x1 + x2 + 2x3 + 8x4 = 0
x1
+ 2x3 + 5x4 = 6
3x1 − x2 + 4x3
= −1
This problem can easily be represented in matrix notation as
Ax = b
where data matrices A and b are
A =

3
3
-4 1
4
1
2
8
1
0
2
5
3
-1
4
0

b =

8
0
6
-1

In an Excel worksheet, identify ranges of cells to accommodate A, b, and x, and name the ranges
appropriately. In this example we will assume these ranges have been named A, b, and x, respectively.
Select the range where the solution x will appear, and type in the formula bar MMULT(MINVERSE(A),b)1.
1Remember that since x is a 4× 1 matrix, you will first have to highlight a 4× 1 range of cells and press CTRL SHIFT
ENTER to evaluate MMULT(MINVERSE(A),b).
Excel Tutorial
5
MS&E 121
Introduction to Stochastic Modeling
January 3, 2005
The vector
x =

-11.4
39.6
18.2
-3.8

is the unique solution to the equation Ax = b in this example.
Simulating Random Variables in Excel
Excel comes with a random number generator that can be used through the function RAND(). This
function generates a uniform random variable on the interval [0, 1], which can be used to generate many
other random variables.
For example, if we want to generate a uniform random variable on the interval [a, b], we use the formula
RAND() ∗ (b− a) + a.
Note: If you want to use RAND to generate a random number but you do not want the number to
change every time the cell is calculated, you can enter =RAND() in the formula bar and then press F9
to change the formula for a fixed random number.
To compute a random number in the set {1, 2, . . . , n} according to a discrete uniform distribution, we
can generate a random number in [0, n] as we did in the first example and then compute the ceiling of
that number (i.e., if the random number is 3.4167, then its ceiling is the integer obtained by rounding-up,
in this case, 4):
CEILING(RAND() ∗ n, 1).
If we want to generate a random number having distribution Normal(µ, σ2) we can do it by the inverse
transformation method. As we will see in class, if F (x) is the CDF of a random variable, F−1(y) is its
inverse, and U is a uniform random number in [0, 1], then F−1(U) is a random observation from the
distribution F . In Excel we can do this by typing
NORMINV(RAND(), µ, σ).
An Example Taken from Hillier, Lieberman, Introduction to Operations Research
Suppose you want to play a game that has the following rules:
1. Each play of the game involves repeatedly flipping an unbiased coin until the difference between
the number of heads tossed and the number of tails is 3.
2. If you decide to play the game, you are required to pay $1 for each flip of the coin. You are not
allowed to quit during a play of the game.
Excel Tutorial
6
MS&E 121
Introduction to Stochastic Modeling
January 3, 2005
3. You receive $8 at the end of each play of the game.
Thus, you win money if the number of flips required is fewer than 8, but you lose money if more than
8 flips are required. How would you decide whether to play this game?
We will use simulation to determine what is the probability that you will win money if you play this
game. First we need to generate a sequence of random observations to represent the tosses of a coin.
We will denote a Head with a “0” and a Tail with a “1”, and what we need is to generate a sequence
of 0s and 1s where the probability of a 0 is 1/2 and the probability of a 1 is 1/2 (since we are assuming
that we will be using a fair coin). To simulate each flip of the coin we can use the formula
CEILING(RAND() ∗ 2, 1).
Now we generate a large number of flips, say 50, and we use the following spreadsheet to record how
many flips were necessary to perform before the game ended.
Flip Result Total Heads Total Tails Difference Stop?
1
0
1
0
1
2
1
1
1
0
3
1
1
2
1
4
0
2
2
0
5
0
3
2
1
6
1
3
3
0
7
0
4
3
1
8
1
4
4
0
9
0
5
4
1
10
0
6
4
2
11
0
7
4
3
Stop
12
0
8
4
4
13
1
8
5
3
14
0
9
5
4
...
...
...
...
...
In this particular simulation, 11 flips were performed before the game ended, in which case you would
lose $3. By repeating this experiment many times and counting the number of simulations in which the
game stops before the 8th flip, you can approximate the probability of making money in this game with
the formula
# of simulations in which the game stops before the 8th flip
total # of simulations performed
.
Just remember that for any inference made out of a simulation to be reliable, you must usually repeat
the experiment many times. In class we will see how many simulations are required in general to obtain
a reliable approximation to the quantity we are computing, and we will also learn to provide confidence
intervals for such approximation.
Excel Tutorial
7