BA 555
Fall 2006
Simulation
@Risk
@Risk Quick Start
1. Launch Excel and open a blank worksheet.
2. Load @Risk program:
Click on
(at the upper left corner) to load @Risk program. If you don’t see the icon, try
Tools / Add-Ins and click on Decision Tools Utilities. If you are asked to Enable
Macros or Disable Macros, click on Enable Macros. If @Risk is successfully loaded,
you should see @Risk Toolbar at the upper left corner. The toolbar looks like
Display List of Outputs and Inputs
Fit Distributions to Data
Report Settings
Save @Risk File
Open @Risk File
Add Output
Define Distributions
Simulation Settings
Start Simulation
3. Prepare the model for simulation
Game 1:
Game 2:
4. Define Outputs:
Move cursor to cell C2 (for Game 1) or B8 (for Game 2) and click on the Add Output icon.
This would define cells C2 and B8 as output cells. @Risk program will automatically define
Input cells (cells with @Risk functions such as RISKDISCRETE or RISKNORMAL.)
5. Start simulation by clicking the Start Simulation icon.
6. The Simulation Settings icon allows you to change the number of simulation and iterations
(the default is one simulation with 100 iterations). The Report Settings icon determines
what should be reported (e.g., brief vs. detailed summary statistics).
Hsieh, P.-H.
1
BA 555
Fall 2006
Simulation
@Risk
Commonly Used @Risk Functions
* indicates the most commonly used functions and will be discussed in the class.
Distribution Function
RiskBeta(alpha1,alpha2)
beta distribution with shape parameters alpha1 and alpha2
RiskBetaGeneral( alpha1,
alpha2,minimum, maximum)
beta distribution with defined minimum, maximum and shape
parameters alpha1 and alpha2
binomial distribution with n draws and p probability of success
on each draw
* RiskBinomial(n,p)
*
Returns
RiskChiSq(v)
Chi-Square distribution with v degrees of freedom
RiskDiscrete({X1,X2,...,Xn},
{p1,p2,...,pn})
discrete distribution with n possible outcomes with the value X
and probability weight p for each outcome
discrete uniform distribution with n outcomes valued at X1
through Xn
RiskDuniform({X1,X2,...Xn})
RiskExpon(beta)
exponential distribution with decay constant beta
extreme value (or Gumbel) distribution with location parameter
a and scale parameter b
gamma distribution with shape parameter alpha and scale
RiskGamma(alpha,beta)
parameter beta
general density function for a probability distribution ranging
RiskGeneral(minimum,maximum,
between minimum and maximum with n (x,p) pairs with value
{X1,X2,...,Xn}, {p1,p2,...,pn})
X and probability weight p for each point
RiskExtvalue(a,b)
RiskGeometric(p)
Hsieh, P.-H.
geometric distribution with probability p
2
BA 555
Fall 2006
Simulation
@Risk
RiskHistogrm(minimum,maximum,{p1,p2,...,pn})
histogram distribution with n classes between minimum and
maximum with probability weight p for each class
hypergeometric distribution with sample size n, D number of
RiskHypergeo(n,D,M)
items and M population size
RiskIntUniform(minimum,maximu uniform distribution which returns integer values only between
m)
minimum and maximum
inverse gaussian (or Wald) distribution with mean mu and
RiskInvGauss(mu,lambda)
shape parameter lambda
logistic distribution with location parameter alpha and scale
RiskLogistic(alpha,beta)
parameter beta
RiskLoglogistic(gamma,beta,
log-logistic distribution with location parameter gamma, scale
alpha)
parameter beta and shape parameter alpha
RiskLognorm(mean,standard
lognormal distribution with specified mean and standard
deviation)
deviation
negative binomial distribution with s successes and p
RiskNegbin(s,p)
probability of success on each trial
RiskNormal(mean,standard
normal distribution with given mean and standard deviation
*
deviation)
*
RiskPareto(theta,a)
pareto distribution
RiskPert(minimum,most likely,
maximum)
pert distribution with specified minimum, most likely and
maximum values
RiskPoisson(lambda)
poisson distribution
* RiskSimtable({X1,X2,...Xn})
RiskStudent(nu)
*
lists values to be used in each of a series of simulations
student's t distribution with nu degrees of freedom
RiskTriang(minimum,most likely, triangular distribution with defined minimum, most likely and
maximum)
maximum values
* RiskUniform(minimum, maximum) uniform distribution between minimum and maximum
RiskWeibull(alpha,beta)
*
RiskCorrmat(matrix cell range,
position,instance)
RiskDepC("ID",coefficient)
RiskIndepC("ID")
RiskTruncate(minimum,
maximum)
Hsieh, P.-H.
weibull distribution with shape parameter alpha and scale
parameter beta
Identifies a matrix of rank correlation coefficients and a
position in the matrix for the distribution in which the Corrmat
function is included. Instance specifies the instance of the
matrix at matrix cell range that will be used for correlating this
distribution.
Identifies dependent variable in correlated sampling pair with
rank correlation coefficient and "ID" identifier string
Identifies independent distribution in rank correlated sampling
pair — "ID" is identifier string
Minimum-maximum range allowable for samples drawn for the
distribution in which the Truncate function is included
3