SlideShare a Scribd company logo
1 of 18
Download to read offline
Annals of Operations Research 5(1985/6)613 -630 613
A SUMMARY OF OR/MS SOFTWARE ON MICROCOMPUTERS
R. SHARDA
Department of Management, Collegeof BusinessAdministration,
OklahomaState University,Stillwater, OK 74078, USA
Abstract
This paper provides a summary of some OR/MS software for microcomputers.
Lists of packages in many management science techniques are included. A biblio-
graphy of other papers is also included. We also summarize the general features of
packages in each category. Finally, we mention a few sources to help the reader
keep up with the developments in this area.
Keywords and phrases
Microcomputers, computer programs, mathematical
optimization, management science, operations research.
programming, software,
1. Introduction
The explosive growth of the use of microcomputers in the last five years has
the potential of changing the way OR professionals perform and present the results of
analysis. The April 1981 special issue of the Journal of OperationalResearchSociety
on microcomputers was the first indicator of the recognition of the importance of
micros to OR analysts and users. Belshaw [ 1] suggests that micros offer a new oppor-
tunity for the distribution of OR models. The spreadsheet interfaces to OR/MS soft-
ware and their links to graphics packages can help an OR analyst present results in a
different form. The graphics capability of Macintosh-like computers can be exploited
to develop better network-oriented software.
Ranyard [16] discusses general advantages and disadvantages of using a micro
in OR applications. Bischoff and Dowsland [2] describe a successful application of
micros for product design and distribution. The programs had originally been developed
for the mainframe and later transferred to a microcomputer. The use of a micro in
this case brought 'the speed of the computer together with the skill and experience of
the end user'.
9J.C. Baltzer A.G., Scientific Publishing Company
614 R. Sharda,OR~MSsoftwareon microcomputers
Geoffrion [5] discusses many aspects of the use the micros in OR/MS. He
suggests that by offering better economics of computing and a 'splendid' learning
environment, the mass market of microcomputers is creating a large group of poten-
tial users of OR/MS technology. Users experienced in building models on spreadsheets
to perform 'what if' analysis are willing to build models for 'what is best' analyses.
This has created a tremendous opportunity for software developers to publish OR/MS
software. A great variety of packages is now available.
New packages are being introduced practically every week. Each issue of
Interfaces or OR/MS Today contains advertisements for new software. Some attempts
have been made to keep the OR/MS community aware of the developments in this
field. For example, OR/MS Today publishes a column on personal computing, edited
by Terry P. Harrison. Management Science Roundtable has commissioned two studies
in this area. A special issue of Computers and OperationsResearch (1986) will be
devoted to OR/MS software.
One objective of this volume is to provide information on state-of-the-art of
algorithms and software for optimization. Proliferation of optimization software for
microcomputers suggests that the readers of this issue would be interested in a summary
of currently available OR/MS software for microcomputers. This paper provides lists
of OR/MS software compiled by techniques. We also summarize the general features
of packages in each category. Finally, we mention a few pointers to help the reader
keep up with the developments in this area.
The scope of this compendium is limited. The paper includes software offering
the most commonly used OR/MS techniques. It also summarizes the forecasting and
simulation software. Some OR/MS techniques are not represented here. This is either
because the author is not aware of any software offering those techniques, or de-
tailed surveys have already been published elsewhere.
The information presented here was gathered from a previous survey, OR/MS
journals, popular microcomputer journals and personal communication. But the list
is by no means exhaustive. It only represents the extent of the author's knowledge of
such software. Any omission is unintentional and not an indicator of the quality of
the package. Further, project management, statistical analysis and simulation software
has received considerable coverage elsewhere and the summaries are readily available.
This paper mentions the references without reproducing the tables from those papers.
Thus, an analyst looking for a package in these areas should consult the original
studies for a more complete list. The paper also does not include the financial modeling
software sold as spreadsheet or decision support software. The wide publicity already
available for such software allows us to concentrate on more specific packages. Finally,
the paper summarizes mostly the software available for the IBM-PC or compatible
machines running MS-DOS operating systems. These machines have the greatest variety
of software available for them. The extra RAM capacity and math 8087-coprocessor
availability have made large-scale number crunching on a micro feasible. These features
R. Sharda, OR/MS software on microcomputers 615
were not available on the earlier micros, so the software base did not develop at the
same level. We do indicate the availability of a package on other types of machines,
where known.
2. Evaluation criteria
Erikson and Turban [4] list several criteria to be considered in evaluating
software for classroom use. Golden et al. [8] also mention the following criteria:
program environment, user interfaces and support, program scope and capabilities
and its effectiveness and efficiency. The basic features to be compared in selecting
OR/MS software can be restated in the following terms: functional, capacity, input
and output related features. A brief explanation of each follows.
FUNCTIONALITY
The packages differ in their ability to solve problems. For example, not all
linear programming packages have sensitivity analysis capabilities. Even when a package
has sensitivity analysis available, it may not permit addition of a constraint/variable.
Further, it may not save the inverted matrix of the optimal solution for a quick post
optimality analysis, but instead require solving the problem all over again after each
change. Forecasting programs also differ in the methods available to the user. Some
offer a smorgasbord of methods, whereas others include only regression or moving
average and exponential smoothing. The also differ in graphical presentation of data.
As a third example, project management software has a wide variation in resource
management options. Thus a thorough check of functional features of the available
software is desirable.
CAPACITY/SIZE LIMITATIONS
A major determinant in selecting a package is the size of the problem it can
solve. Most packages describe the size of the dataset they can handle. For example,
in LP problems the number of constraints, variables and non-negative elements are
limiting factors. In forecasting packages, the number of data items in a series, as well
as the number of time series, may be restricted. In selecting a PERT package, one
would consider the number of activities for each project and the allowed number of
subprojects,
A related issue is the speed of execution. Given that a lot of number crunching
is involved in most OR/MS software, the program has to be written well to execute
fast. One way for programs to improve speed (as well as accuracy) on MS-DOS com-
puters is to use an Intel 8087 math coprocessor. Use of the coprocessor can speed
execution up by 15-500%. The software must be written to take advantage of the
616 R. Sharda, OR/MS software on microcomputers
chip. Fortunately, a number of OR/MS packages support this coprocessor, but not all
do so. It should be an important criterion in selecting a package for any large-scale
application.
INPUTFEATURES
The ease of data and command input for software can sometimes determine
its eventual acceptance by the user. Some programs are menu driven, whereas others
are strictly command driven. Menu driven programs are easy to learn but they can
become boring for an experienced user. A combination of the two modes is ideal for
command entry. A program should also be able to accept data in a number of ways.
For a beginner, the program should provide a data editor/manager with appropriate
prompts. For an experienced user, as well as for solving larger problems, the program
should be able to accept data input created externally, such as in an ASCII file, by a
word processor. To offer better portability, the program should be able to accept data
in a standard format. For example, many spreadsheet packages can output data in
the DIF (data interchange format) so that it can be read by other packages. A good
OR/MS package should be able to read such files. For linear programming applications,
a program should be able to accept problem input in the popular MPS format. This
would permit easy transfer of a problem from mainframe to micro.
OUTPUTFEATURES
One reason microcomputers are becoming so popular is the ability of a user to
control the periPherals of the computer. A user should be able to decide which part of
the output goes on screen and which goes to the printer. For example, an LP package
could allow one to see only the most important output on screen and print the rest on
the printer, if requested. Further, one should be able to access this output through a
data file. This access should be twofold. If the solution is readable by a spreadsheet
package such as Lotus 1-2-3, one will be able to present some of the results graphically.
Another paper in this volume [19] explains this is detail. Second, if the solution file
can be accessed by a word processing package, the report can incorporate the results
of the analysis efficiently without the need of retyping all the numbers.
The program should be able to use graphics where appropriate. For example,
data plots should be available in performing statistical or time series analysis. One
should be able to print these plots not only on the printer but also on popular plotters.
If a project management package is being used, it should be able to produce a network
diagram, Gantt chart, etc.
OTHERCONSIDERATIONS
Some of the other obvious considerations include minimum required hardware,
and price. Most programs can use additional memory. Given that the price of memory
R. Sharda, OR/MS software on microcomputers 617
is falling, it is usually a worthwhile investment and can result in a significant improve-
ment in performance. Even those programs which can not use additional memory will
execute faster if the floppy disk data is copied into a RAM disk. Many utilities are now
available to define part of the computer's RAM as a disk drive. In fact, most sellers of
memory boards include such software,
The price of the software also varies widely. And it does not seem to correlate
positively with performance. Of course, the potential benefits of using any of these
packages can far outweigh the cost of the software. It is, however, worthwhile to com-
pare costs as well as performance.
One other consideration should be user support. Many companies offering
these packages are quite small. One should consider whether the company is strong
enough to update the software and help correct any problems. While no guidelines
can be offered, one can perhaps ask for references to check on the strength of the
software publisher.
3. What is available?
We now summarize software for each of the major categories. Where the
author is aware of a more detailed summary published elsewhere, the reference is
cited. For a complete list, the reader should also consult these references.
The information in the following tables is current as of this writing. The
prices listed here are for single copies for commercial users. Many of the software
publishers have educational prices and/or site license agreements available. Space does
not permit listing all such details.
As mentioned earlier, the software included in these tables runs on IBM-PC
or compatible microcomputers. Some of the programs also run on other types of
microcomputers. The reader should check with the publishers for such details.
LINEARPROGRAMMING
One of the most important OR application tools, linear programming (LP),
has been made available on the PC by a number of software publishers. The author is
aware of at least two dozen packages on the market. Many of these packages were
summarized by Sharda [18]. Additional packages are summarized in table 1. The
reader should also consult the original paper for a complete list. The information in
table 1 is based only on the available literature and not on actual usage. Some items
were not discernible from the brochures.
The linear programming software on the market varies widely in price, size of
the problem solvable by the software, and other features. The largest LP problem one
can solve on a PC appears to be 2000 rows and 2500 columns, at least thus far. Of
course, as problem size increases, the speed of execution is affected adversely. Having
an 8087 math coprocessor can help. But, the software has to be able to take advantage
of the math coprocessor.
618 R. Sharda, OR~MSsoftware on microcomputers
0
o
0
o~
t~
"a
o
o
o
~z
Z
%
~o~
oo
0
~.~ ~ ~.~ ~ o ~ o
o o o ~ o
Z Z Z ~ Z ;~
o o ~ o
0
;~ Z
I
0
0 o
0 0 0 0
0000 0,-,
V/ N V/V/
~ Z ~
0 O 0 O 0
O 0 O 0
(.J
0 0
d
~o ~d_~o
o
,N
0
0
E
II
N
Z
a
0
E
II
2
oo
II
R. Sharda, OR/MS software on microcomputers 619
These packages handle input/output in many ways. Some allow the data entry
only one coefficient at a time, whereas others can accept input from a file created by
using an editor/word processor/matrix generation program. A few packages accept
data in MPS format as well, which permits portability of a model between micro and
mainframe, as well as between different LP packages offering this feature. Some even
interface with packages such as Lotus 1-2-3, Multiplan or Visicalc to accept a problem
formulation using that package. This interface may also permit one to process the LP
output using the spreadsheet package to prepare better reports.
While OR/MS software has not proliferated on Macintosh yet, the author is
aware of one code running on Macintosh. This code is being developed by Professor
James Ho of the University of Tennessee. According to Professor Ho, the code is able
to solve problems with 500 rows and 5000 non-zeroes on the 512 K Mac.
OTHER MATHEMATICALPROGRAMMINGSOFTWARE
Some of the LP packages also offer integer programming as an option. This
allows one to solve mixed integer programming (MIP) problems. The author is aware
of five packages which include significant IP or MIP features. In addition, an instruc-
tional package, MICROSOLVE/OR by Jensen [10] also includes an integer program-
ming routine. Table 2 lists major packages offering MIP, nonlinear programming and
transportation method.
The integer programming capabilities of the five packages vary. LINDO is able
to solve a problem with 200 integer variables, but these are restricted to be 0/1 vari-
ables. The other four packages can accept a combination of 0/1 as well as general
integer variables, but the problem size is restrained somewhat. |LP-88 assumes all
integer variables to be general. If a variable is to be a 0/1 variable, one must enter two
constraints, x ~< 1 and x/> 0. This can add to the number of constraints considerably.
Further, it allows for 45 integer variables and 255 constraints (including the con-
straints for 0/1 variables). MIP-83, an option of LP-83, also works on a mixed integer
programming problem. It is possible to specify 0/1 variables without including the two
bound constraints. The maximum number of integer variables is not specified in the
manual. However, the author is aware of one application containing 40 integer vari-
ables where MIP-83 could not obtain a solution in 'reasonable' time. The number of
integer variables is not a sufficient indicator of the power of any MIP program. The
formulation of the problem can affect the number of nodes to be searched. As this
number increases, the storage space can also be a bottleneck, affecting the overall
performance of the program. As in the case with IP software in general, there is room
for improvement in this area.
Nonlinear programming software for micros is just beginning to appear. This
author has used only one package and is aware of one other. These are also listed in
table 2. GINO is an adaption of GRG [12] in the microcomputer environment. It
620 R. Sharda, OR/MS software on microcomputers
e~
N
~
o
~ 9
~ ..~
0
0
9
~ o
o ~
o .~ ~ .~ ~ ~ ~ ~.~
~x ~x
0
0
o
r..)
z~
.q.
.~ •
O 0
0 0
,-1
r
,~ d
o ~ ~
~.~ ~
~.~ ~
~m
0
0
0~o
0
X
0
Z
0
0
d
2
0
Z
R. Sharda, OR/MS software on microcomputers 621
0
O,
[.-,
N
~
<
,.c:
,.0
0
~ ~o
0 0
X )<
0 0
0 0
~ ~ ~
~ =~ ~
~o~ "~
~
9
~__, ~r~
:~ ~.~ ~
o
0
0 ~, 0
O0 ,~"
9
~ ~ ~.= "~ ~.~
. ~
9
,~ ~ ,~
0 ~ '~ "~
~ ~,~ ~o~.~
9
~~ ~ ~-~ ~
oo
622 R. Sharda,OR/MSsoftwareon microcomputers
is designed to work in the interactive mode similar to LINDO. The maximum problem
size on a PC is limited to 30 rows and 50 columns. While it may not be able to solve
very large problems on a PC, it can serve as an excellent instructional tool.
SYSOPT is designed for unconstrained as well as constrained optimization.
It is reported to include a number of search routines such as Fibonacci search in one
dimension and PoweU, Fletcher and Reeves' search in n dimensions. It also includes
a penalty function method for constrained optimization. Wasil and Golden [9] provide
more comments on this program, as well as many other nonlinear programming
packages. They also summarize some software capable of solving a system of non-
linear equations.
Two other packages of interest are MPPII and TSA-88. MPPII is actually a
linear programming system, but it also offers a transportation simplex implementation.
It is able to solve a problem with 70 rows and 100 columns. TSA-88 is able to solve
problems with up to 510 sources and sinks. The publisher of TSA-88 has also an-
nounced a microcomputer implementation of Karmarkar's algorithm for solving LP
problems. A set of programs is available in MP. This package can be purchased in full,
or the modules can be obtained at a lower price.
SIMULATION
Many of the popular simulation packages on the mainframe are now available
for microcomputers. The March 1984 issue of Byte summarizes simulation on micros.
Pratt [15] lists some simulation packages running on various microcomputers.
Table 3 lists some simulation packages available for a PC. Law and Kelton [13]
provide a comparison of the mainframe versions of GPSS, SIMSCRIPT and SLAM.
GPSS is a discrete event simulation language following the process interaction approach.
Its block structure makes model development relatively easy. A major criticism of the
mainframe version of GPSS has been that its clock is integer-valued and random vari-
ates from different distributions are not generated automatically. However, informa-
tion on the PC version claims to have fixed the latter problem.
A continuous simulation language following the systems dynamics approach
is Dynamo. Its micro version is now available as Micro-Dynamo. It appears to be a
reasonably complete implementation of mainframe Dynamo. All the other simulation
languages listed in table 3 offer both discrete and continuous simulation features.
SIMSCRIPT/PC is sold as SIMLAB. The mainframe SIMSCRIPT II.5 is known to
offer natural language-like syntax for easy modeling. It is also able to perform sta-
tistical analysis of simulation output data. The PC version is supposedly a full imple-
mentation of the mainframe version.
SLAM II/PC is also a PC version of SLAM II running on mainframe. It offers
an easy network modeling structure as well as discrete and continuous modeling. The
PC version can output the data to a DIF file, so that further processing, plotting, etc.
can be performed on the PC. This is a very convenient feature of this package.
R. Sharda, OR/MS software on microcomputers 623
Table 3
Simulation software
Software Publisher Price Type of
(USA) (U.S.$) simulation
GPSS/PC Minuteman Software 900 Discrete
P.O. Box 171
Stow, MA 01775
(617) 897-5662
Micro Dynamo Addison-Wesley Publishing 295 Continuous
Company, Inc.
Jacob Way
Reading, MA 01867
(617) 944-3700
C.A.C.I 6500
3344 N. Terry Pines Ct.
La Jolla, CA 92037
(619) 457-9681
Systems Modelling Corp. 1500
P.O. Box 10074
State College, PA 16805
(814) 238-5919
Pritsker & Associates 975
P.O. Box 2413
West Lafayette, IN 47906
(317) 463-5557
Professor C.C. Barnett 125
Walla WaUaCollege
College Place, WA 99324
(509) 522-1346
Mitchell & Gauthier Associates
290 Baker Ave.
Concord, MA 01742
(617) 369-5115
SIMSCRIPT/PC Combined
SIMAN Combined
SLAM II/PC Combined
Micro Passim Combined
ACSL/PC Continuous
Micro-Passim is written in Turbo Pascal and takes advantage of the advanced
features of the language. It is a relatively inexpensive package offering capabilities
similar to most other simulation software.
624 R. Sharda, OR/MS software on microcomputers
Table 4
Some project management packages
Name of package
Company name & address Price
(USA) (U.S.$)
Number of
activities
Resource
loading
MICROPERT
MICRO GANTT
PATHFINDER
HARVARD
PROJECT
MANAGER
QUICKPLAN
Primavera
Advanced
Project 6
Sheppard Software Company 350
4750 Clougli Creek Road
Redding, CA 96002
(916) 222-1553
Earth Data Corporation 395
P.O. Box 13168
Richmond, VA 23225
(804) 231-0300
Morgan Computer Co., Inc. 100
10400 N. Central Expressway
Suite 200
Dallas, TX 75231
(214) 739-5895
Harvard Software, Inc. 395
521 Great Road
Littleton, MA 01460
(617) 486-8431
Mitchell Management 995
Systems, Inc.
2000 W. Park Drive
Westboro, MA 01581
(617) 36649800
Primavera System, Inc. 2500
29 BalaAvenue
Bala Cynwyd, PA 19104
(215) 667-8500
SoftCorp Inc. 199
2340 State Road 580
Suite 244
Clearwater, FL 33575
(813) 799-3984
220
NL
3160
NL
250
10000
250
No
Yes
No
Yes
Yes
Yes
Yes
NL = limited by disk space.
R. Sharda, OR~MSsoftware on microcomputers 625
PROJECT MANAGEMENT
Project management seems to be the only technique in an OR/MS analyst's
toolbox that has received significant attention on the micro market. Most trade
publications and magazines have surveyed this group of software and printed com-
parative tables. For example, Ruby [17] compared 38 project management systems.
A recent book by Gido [6] compares 120 project management packages. Some of
these are listed in table 4. As before, the inclusion here is only to provide a sample,
not an endorsement. It will be worthwhile to read one of the comparative articles
before selecting a project management system.
A wide variety of project management software is on the market. The packages
available can handle projects with as few as less than 100 activities to upwards of 1500
activities. The programs vary in their resource allocation and cost management features.
Some programs allow up to 100 resources per project. Others offer no resource alloca-
tion. A few packages also allow linkages between different projects for optimal re-
source management.
The programs also offer varied degree of cost management. Some programs
allow one to compare actual expenditures to budgeted amounts. A few have no cost-
related features. For project management applications requiring complete cost tracking,
one will probably have to invest in the more expensive packages of the group.
The reporting features of micro project management systems also vary widely.
While almost all of the packages prepare a Gantt chart, a few are also able to print or
plot a network chart. This may be a plus for reporting purposes. One may also wish to
consider those packages which are able to print task lists as well as a loading report for
each individual resource. Some packages are able to prepare reports aggregated at
various levels. This feature may be useful for preparing reports for managers at different
levels.
The ease of use of these packages varies as well. Some are quite technical as
they are mere downloads of mainframe packages. Others take advantage of the micro-
computer technology fully. A few have been developed using the command/menu
structure of packages such as Lotus 1-2-3. Overall, a large variety of software exists
in this area. One should compare the features and possibly examine a demonstration
copy before purchasing, since these packages are usually quite expensive.
STATISTICALANALYSISAND FORECASTING
Statistical analysis has also received a lot of attention in the microcomputer
market. Carpenter et al. [3] summarized 24 statistical analysis packages running on
various micros. More recently, Littman [14] compared 19 statistical analysis packages
for the IBM-PC and compatibles. This section will not repeat that comparison, but will
cover only the forecasting software. While some of the comprehensive statistical
626 R. Sharda, OR~MSsoftware on microcomputers
Table 5
Some forecasting packages
Name of package Company name & address Price
(USA) (U.S.$)
Comments
SmartForecasts
SIBYL/RUNNER
AUTOBJ
ESP
SSI-SSF
4CAST/1
WG/ARIMA
Forecast Plus
Smart Software 495
392 Concord Avenue
Belmont, MA 02178
(617) 489-2743
Applied Decision Systems 995
Temple, Barker & Sloane, Inc.
33 Hayden Avenue
Lexington, MA 02173
(617) 861-7580
Automatic Forecasting 795
Systems, Inc.
Statistical Consultants
P.O. Box 563
Hatboro, PA 19040
(215) 675-0652
ALPHA Software Corp. 795
30 B. Street
Burlington, MA 01803
(617) 229-2924
Scientific Systems, Inc. 495
54 Rindge Avenue Ext.
Cambridge, MA 02140
(617) 661-6364
Heurix Computer Products 525
P.O. Box 9227
Morristown, NJ 07960
(201) 267-2806
The Winchendon Group 350
3970 Lakota Road
P.O. Box 10114
Alexandria, VA 22310
(703) 960-2587
Walonick Associates 400
5624 Grand Avenue South
Minneapolis, MN 55419
(612) 340-1515
Has own operating system or can
be run with MS-DOS. Graphics
adaptor required.
PC/MULTBJ available for multi-
variate analysis. Links to Lotus
1-2-3
8087 coprocessor recommended.
AUTOBOXX available for multi-
variate analysis
8087 coprocessor recommended.
Econometric forecastingpackage
State-space modeling package
Supports color plotter
Has own operating system. Auto-
matic version-EASI/ARIMA
available for US$ 350. Also sells
Econometric Linear Forecasting
(ELF) for US$ 350.
Also sells a general statistical
analysis package, StatPac.
R. Sharda, OR/MS software on microcomputers 627
analysis packages also include forecasting routines, those packages are not included
here. Packages which emphasize forecasting only are summarized. A list of major
systems is presented in table 5.
The forecasting software is reasonably homogeneous in its data management
offerings. Most packages are able to accept data from external sources in the form of
DIF or ASCII fries. They also offer a graphical presentation of historical as well as
forecast data, though not necessarily in color. Only one package, 4CAST/1,can plot
data on a color plotter. However, others can also accomplish this by creating a DIF
file and using a graphics package with the file. All of the packages also offer commonly
used data transformations.
The software varies in the methods available for forecasting. Five packages,
SYBIL/RUNNER, ESP, Forecast Plus, 4Cast/1 and SmartForecasts, offer a number of
methods for forecasting. All of these packages cover the commonly used methods
such as moving average, exponential smoothing, filtering, regression and decomposition.
SmartForecasts and 4Cast/1 make excellent use of the color graphics features of a
microcomputer. Only 4Cast/1 offers the Bayesian technique. SmartForecasts offers
a similar technique called judgmental forecasting whereby the user is allowed to eye-
ball the forecast and modify it, based on earlier experience.
ESP is a subset of mainframe software. ESP offers two-stage and three-stage
regressions using various options. It also offers nonlinear estimations as well as uni-
variate Box-Jenkins modeling. SIBYL/RUNNER offers eighteen techniques for uni-
variate analysis plus multiple regression. A companion package, PC/MULTBJ is avail-
able for multivariate modeling. It also makes initial suggestions for techniques to be
used, based upon preliminary data analysis.
A few packages have taken this concept further and offer a somewhat auto-
mated version of forecasting. SmartForecasts, in its automatic mode, conducts a
'tournament' among competing methods and selects the best method for forecasts.
It does not include the Box-Jenkins technique (B-J) in its evaluation, thus leaving a
major tool out. EASI/ARIMA, a complement of TWG-ARIMA, claims to be able to
automate the univariate Box-Jenkins modeling methodology. Another package,
AUTOBJ, is reportedly able to build B-J models automatically. A second product by
the same publisher, AUTOBOXX, offers automatic multivariate Box-Jenkins modeling
including transfer-function methodology. It also claims to include an intervention
detection system. Automatic forecasting option is also available in SSI-SSF, which
uses the state-space forecasting technique. Of course, all of the 'automatic' packages
can be used to perform forecasting based on the analyst's skill in modeling.
There is some concem about the ability of a program to mimic the iterative
process of modeling by a forecaster. More tests of these packages are necessary to
validate their claims. It appears that the structure of Box-Jenkins modeling can be
exploited to come up with a tentative model automatically, to be further refined
by the analyst. Perhaps this is an example of a rule-based expert system. In any case,
the performance of these automatic packages needs to be tested thoroughly.
628 R. Sharda, OR/MS software on microcomputers
Overall, the state of forecasting software appears to be good and improving.
Many comprehensive statistical packages not reported here also include at least some
forecasting options, thus making the selection even wider. This software will probably
see even more growth if the 'automatic' software packages can make statistical analysis
easier for micro users.
OTHER PACKAGESOF INTEREST
A list of other software of interest to the OR/MS user is provided in table 6.
Erikson and Turban [4] summarize most of the educational software in their working
Table 6
Other packages of interest to OR/MS analysts
Software Publisher Price Techniques Comments
(USA) (U.S.$)
Expert-Ease Expert Systems 695 Expert system Spreadsheet-type
868 W. End Avenue generator input
New York, NY 10025
Exsys Exsys 300 Expert system Rule based
P.O. Box 75158 generator
Contr. Street 14
Albuquerque, NM 87194
(505) 83645676
Insight Level5 Research 95 Expert system Rule based, simple
4980 South A-1-A generator system, best as an
Melbourne Beach, FL 32951 educational tool
(305) 676-5810
HR-FLOW HR Management Associates 795 Markov Human resource
P.O. Box 64070 analysis planning system
Tacoma, WA98464
(206) 564-9704
paper. A lot of attention is being paid to expert systems software. For example,
Management Science Roundtable has commissioned two studies in this area. These
reports will be available through ORSA/TIMS later this year. Two packages which are
advertised extensively are Expert-Ease and Exsys. These are not expert systems but
expert systems generators. Expert-Ease is a spreadsheet-type system which is applicable
to classification-type problems requiring expert opinion. It is able to generate rules
using the examples provided by the user. Exsys is an IF-THEN rule based system for
R. Sharda, OR/MS software on microcomputers 629
similar problems, but also includes a probabilistic element. Insight is also a simple
rule-based expert system generator. A number of other packages will probably be
available within the next few months.
An implementation of Markov analysis of manpower supply and demand is
available in HR-FLOW. This is a straightforward application of Markov analysis. The
user is required to enter the number of job levels, time periods, a transition matrix,
the starting number of employees and projected demand at each job level. The pro-
gram is able to calculate total projected supply and excess/shortage in each category.
It requires the user to enter the data by modifying the program written in MBASIC
(by using DATA statements).
Vehicle routing software has also been available for the PC. Golden and Bodin
[7] present a survey of such software. Golden et al. [8] also summarize much of the
decision analysis software available for the micros. The reader should refer to these
papers for tables similar to the ones presented here.
HOWTO KEEPUPTO DATE
Microcomputers in general and OR/MS software in particular are growing
industries. Many new packages are being introduced and existing packages are being
upgraded. Here are some pointers on how to stay in touch with the latest developments.
First of all, Interfaces and OR/MS Today are good sources for new product
advertisements. A number of products listed in this paper were advertised in these two
journals. Interfaces also publishes papers in this area occasionally. A column on
personal computing and OR/MS appears regularly in OR~MSToday. It usually includes
new product announcements and product reviews. The column also identifies papers
of general interest to the OR/MS community. Two other journals which have pub-
lished microcomputer-related articles are Computers and Operations Research and
Journal of Operational Research Society. More are sure to follow. A special issue
of Computer and OperationsResearch in early 1986 will include comparison of many
OR/MS packages.
There are two major sources of on-line information. International Software
Database Corporation, Fort Collins, Colorado, publishes MENU - The International
Software Database. This database contains short descriptions of software with in-
dexing by broad application categories, techniques, compatible computers, and so on.
It is updated monthly. A related source is Microcomputer Index, published by Data-
base Services, Los Altos, California. This index is a subject and abstract guide to
articles from over fifty microcomputer journals. It is updated quarterly. Both of these
databases are accessible through Dialog (Dialog Information Services, Palo Alto,
CA (800) 227-1927). A diskette version of Microcomputer Index has just been
announced.
630 R. Sharda, OR/MS software on microcomputers
4. Conclusion
This paper has presented a listing of some OR/MS software on microcomputers.
A large variety of software is now available. The advantages of an independent CPU
with direct control of peripherals offers a new way of delivering OR/MS models to
clients. This paper provides the analyst with at least partial information about these
opportunities.
References
[1] P.N. Belshaw, Microcomputers and management science, Interfaces 12, 6(1982)105.
[2] E. Bischoff and W.B. Dowsland, An application of the micro to product design and distri-
bution, J. of Opl. Res. Soc. 33, 3(1982)271.
[3] J. Carpenter, D. Deloria and D. Morganstein, Statistical software for microcomputers,
Byte 9, 5(1984)234.
[4] W.J. Erikson and E. Turban, Teaching operations research on microcomputers, Research
Report DS-84-09, School of Business Administration, University of Southern California,
Los Angeles (1984).
[5] A.M. Geoffrion, Can MS/OR evolve fast enough? Interfaces 13, 1(1983)10.
[6] J. Gido, ProjectManagementSoftwares (Industrial Press, Inc., New York, NY,
1985).
[7] B. Golden and L. Bodin, Vehicle routing software for microcomputers: A survey, Com-
puters and Operations Research (1986) forthcoming.
[8] B. Golden, A. Hevner and D. Power, Decision insight systems for microcomputers: A critical
evaluation, Computers and Operations Research (1986) forthcoming.
[9] B. Golden and E. Wasil, A survey of nonlinear programming software, Computers and
Operations Research (1986) forthcoming.
[10] P. Jensen,Microsolve/Operat~onsResearch(Holden-Day, Oakland, CA, 1983).
[11] J.A. Kastner and S.J. Hong, A review of expert systems, Eur. J. Oper. Res. 18, 3(1984)285.
[12] L.S. Lasdon, A.D. Warren, A. Jain and M. Ratner, Design and testing of a generalized re-
duced gradient code for nonlinear programming, ACM Transactions on Mathematical Soft-
ware 4(1978)34.
[13 ] A.M. Law and W.D. Kelton, SimulatT"onModelingand Analysis(McGraw-Hill, New York,
NY, 1982).
[141 J. Littman, Statistical software: Making it simple, PC Week 2, 3(1985)45.
[15 ] C.A. Pratt, Going further: A compendium of conferences, organizations, books and software
for simulationists, Byte 9, 3(1984)204.
[161 J.C. Ranyard, Introducing the microcomputer into the OR department, J. of Opl. Res. Soc.
32,4(1981)277.
[17] D. Ruby, On the critical path, PC Week 2, 3(1985)51.
[18] R. Sharda, Linear programming on microcomputers: A survey, Interfaces 14, 6(1984)27.
[ 19] R. Sharda, Optimization using spreadsheets on a microcomputer, Ann. Oper. Res. 5(1985/6)
599.

More Related Content

Similar to A Summary Of OR MS Software On Microcomputers

Basic-Project-Estimation-1999
Basic-Project-Estimation-1999Basic-Project-Estimation-1999
Basic-Project-Estimation-1999
Michael Wigley
 
Stat Tech Reportv1
Stat Tech Reportv1Stat Tech Reportv1
Stat Tech Reportv1
stat
 
Review on Algorithmic and Non Algorithmic Software Cost Estimation Techniques
Review on Algorithmic and Non Algorithmic Software Cost Estimation TechniquesReview on Algorithmic and Non Algorithmic Software Cost Estimation Techniques
Review on Algorithmic and Non Algorithmic Software Cost Estimation Techniques
ijtsrd
 
Proceedings of the 2015 Industrial and Systems Engineering Res.docx
Proceedings of the 2015 Industrial and Systems Engineering Res.docxProceedings of the 2015 Industrial and Systems Engineering Res.docx
Proceedings of the 2015 Industrial and Systems Engineering Res.docx
wkyra78
 
Analyzing the solutions of DEA through information visualization and data min...
Analyzing the solutions of DEA through information visualization and data min...Analyzing the solutions of DEA through information visualization and data min...
Analyzing the solutions of DEA through information visualization and data min...
Gurdal Ertek
 
Zelditchetal workbookgeomorphoanalyses
Zelditchetal workbookgeomorphoanalysesZelditchetal workbookgeomorphoanalyses
Zelditchetal workbookgeomorphoanalyses
Wagner M. S. Sampaio
 
10.1.1.104.5038
10.1.1.104.503810.1.1.104.5038
10.1.1.104.5038
96565
 
Software metric analysis methods for product development
Software metric analysis methods for product developmentSoftware metric analysis methods for product development
Software metric analysis methods for product development
iaemedu
 
Software metric analysis methods for product development
Software metric analysis methods for product developmentSoftware metric analysis methods for product development
Software metric analysis methods for product development
iaemedu
 
Software metric analysis methods for product development maintenance projects
Software metric analysis methods for product development  maintenance projectsSoftware metric analysis methods for product development  maintenance projects
Software metric analysis methods for product development maintenance projects
IAEME Publication
 

Similar to A Summary Of OR MS Software On Microcomputers (20)

Practical Machine Learning
Practical Machine LearningPractical Machine Learning
Practical Machine Learning
 
Basic-Project-Estimation-1999
Basic-Project-Estimation-1999Basic-Project-Estimation-1999
Basic-Project-Estimation-1999
 
Stat Tech Reportv1
Stat Tech Reportv1Stat Tech Reportv1
Stat Tech Reportv1
 
International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)International Journal of Engineering and Science Invention (IJESI)
International Journal of Engineering and Science Invention (IJESI)
 
Information technology
Information technologyInformation technology
Information technology
 
Role of Computers in Research, Data Processing, Data Analysis
Role of Computers in Research, Data Processing, Data AnalysisRole of Computers in Research, Data Processing, Data Analysis
Role of Computers in Research, Data Processing, Data Analysis
 
Sugi-82-26 Dolkart
Sugi-82-26 DolkartSugi-82-26 Dolkart
Sugi-82-26 Dolkart
 
Review on Algorithmic and Non Algorithmic Software Cost Estimation Techniques
Review on Algorithmic and Non Algorithmic Software Cost Estimation TechniquesReview on Algorithmic and Non Algorithmic Software Cost Estimation Techniques
Review on Algorithmic and Non Algorithmic Software Cost Estimation Techniques
 
Proceedings of the 2015 Industrial and Systems Engineering Res.docx
Proceedings of the 2015 Industrial and Systems Engineering Res.docxProceedings of the 2015 Industrial and Systems Engineering Res.docx
Proceedings of the 2015 Industrial and Systems Engineering Res.docx
 
[IJCT-V3I2P32] Authors: Amarbir Singh, Palwinder Singh
[IJCT-V3I2P32] Authors: Amarbir Singh, Palwinder Singh[IJCT-V3I2P32] Authors: Amarbir Singh, Palwinder Singh
[IJCT-V3I2P32] Authors: Amarbir Singh, Palwinder Singh
 
IRJET- Determining Document Relevance using Keyword Extraction
IRJET-  	  Determining Document Relevance using Keyword ExtractionIRJET-  	  Determining Document Relevance using Keyword Extraction
IRJET- Determining Document Relevance using Keyword Extraction
 
Analyzing the solutions of DEA through information visualization and data min...
Analyzing the solutions of DEA through information visualization and data min...Analyzing the solutions of DEA through information visualization and data min...
Analyzing the solutions of DEA through information visualization and data min...
 
IRJET- Towards Efficient Framework for Semantic Query Search Engine in Large-...
IRJET- Towards Efficient Framework for Semantic Query Search Engine in Large-...IRJET- Towards Efficient Framework for Semantic Query Search Engine in Large-...
IRJET- Towards Efficient Framework for Semantic Query Search Engine in Large-...
 
Zelditchetal workbookgeomorphoanalyses
Zelditchetal workbookgeomorphoanalysesZelditchetal workbookgeomorphoanalyses
Zelditchetal workbookgeomorphoanalyses
 
A Study of Software Size Estimation with use Case Points
A Study of Software Size Estimation with use Case PointsA Study of Software Size Estimation with use Case Points
A Study of Software Size Estimation with use Case Points
 
10.1.1.104.5038
10.1.1.104.503810.1.1.104.5038
10.1.1.104.5038
 
Software metric analysis methods for product development
Software metric analysis methods for product developmentSoftware metric analysis methods for product development
Software metric analysis methods for product development
 
Software metric analysis methods for product development
Software metric analysis methods for product developmentSoftware metric analysis methods for product development
Software metric analysis methods for product development
 
Software metric analysis methods for product development maintenance projects
Software metric analysis methods for product development  maintenance projectsSoftware metric analysis methods for product development  maintenance projects
Software metric analysis methods for product development maintenance projects
 
Association Rule Mining Scheme for Software Failure Analysis
Association Rule Mining Scheme for Software Failure AnalysisAssociation Rule Mining Scheme for Software Failure Analysis
Association Rule Mining Scheme for Software Failure Analysis
 

More from Christine Williams

More from Christine Williams (20)

Why A College Education Is Important Essay.Docx - Wh
Why A College Education Is Important Essay.Docx - WhWhy A College Education Is Important Essay.Docx - Wh
Why A College Education Is Important Essay.Docx - Wh
 
About Me Paper Example Awesom
About Me Paper Example AwesomAbout Me Paper Example Awesom
About Me Paper Example Awesom
 
015 Guide English How To Write An Essay Perfect T
015 Guide English How To Write An Essay Perfect T015 Guide English How To Write An Essay Perfect T
015 Guide English How To Write An Essay Perfect T
 
Great Essay Titles. Top 30
Great Essay Titles. Top 30Great Essay Titles. Top 30
Great Essay Titles. Top 30
 
6 Basic Outline Templates - PDF
6 Basic Outline Templates - PDF6 Basic Outline Templates - PDF
6 Basic Outline Templates - PDF
 
Essay On Empathy For Students - EnglishGramma
Essay On Empathy For Students - EnglishGrammaEssay On Empathy For Students - EnglishGramma
Essay On Empathy For Students - EnglishGramma
 
4 Major Types Of Essays - Infographics Types Of Ess
4 Major Types Of Essays - Infographics Types Of Ess4 Major Types Of Essays - Infographics Types Of Ess
4 Major Types Of Essays - Infographics Types Of Ess
 
Opinion Writing Graphic Organizer 4Th Grade
Opinion Writing Graphic Organizer 4Th GradeOpinion Writing Graphic Organizer 4Th Grade
Opinion Writing Graphic Organizer 4Th Grade
 
45 Perfect Thesis Statement Template
45 Perfect Thesis Statement Template45 Perfect Thesis Statement Template
45 Perfect Thesis Statement Template
 
Can You Help Me Write An Essay About How T
Can You Help Me Write An Essay About How TCan You Help Me Write An Essay About How T
Can You Help Me Write An Essay About How T
 
Good Essay Editing Service
Good Essay Editing ServiceGood Essay Editing Service
Good Essay Editing Service
 
Affordable Custom Paper Writing Services Prof
Affordable Custom Paper Writing Services ProfAffordable Custom Paper Writing Services Prof
Affordable Custom Paper Writing Services Prof
 
Definition Essay Usc Dornsife Supplemental Ess
Definition Essay Usc Dornsife Supplemental EssDefinition Essay Usc Dornsife Supplemental Ess
Definition Essay Usc Dornsife Supplemental Ess
 
Mrs. Jones - Free Worksheets And Printable
Mrs. Jones - Free Worksheets And PrintableMrs. Jones - Free Worksheets And Printable
Mrs. Jones - Free Worksheets And Printable
 
Reflective Essay Examples Structure Great Tips Pr
Reflective Essay Examples Structure Great Tips PrReflective Essay Examples Structure Great Tips Pr
Reflective Essay Examples Structure Great Tips Pr
 
Cheap Research Paper Writing Cheap Essay Writing Service From Page
Cheap Research Paper Writing Cheap Essay Writing Service From PageCheap Research Paper Writing Cheap Essay Writing Service From Page
Cheap Research Paper Writing Cheap Essay Writing Service From Page
 
Practicing Writing An Argumentat
Practicing Writing An ArgumentatPracticing Writing An Argumentat
Practicing Writing An Argumentat
 
How To Write A Narrative Essay Best Guide And To
How To Write A Narrative Essay Best Guide And ToHow To Write A Narrative Essay Best Guide And To
How To Write A Narrative Essay Best Guide And To
 
How You Can Write A Great Narrative Essay Step-By-Step
How You Can Write A Great Narrative Essay Step-By-StepHow You Can Write A Great Narrative Essay Step-By-Step
How You Can Write A Great Narrative Essay Step-By-Step
 
IR Argumentative Essay Graphic Organizer By Ren
IR Argumentative Essay Graphic Organizer By RenIR Argumentative Essay Graphic Organizer By Ren
IR Argumentative Essay Graphic Organizer By Ren
 

Recently uploaded

Spellings Wk 3 English CAPS CARES Please Practise
Spellings Wk 3 English CAPS CARES Please PractiseSpellings Wk 3 English CAPS CARES Please Practise
Spellings Wk 3 English CAPS CARES Please Practise
AnaAcapella
 

Recently uploaded (20)

This PowerPoint helps students to consider the concept of infinity.
This PowerPoint helps students to consider the concept of infinity.This PowerPoint helps students to consider the concept of infinity.
This PowerPoint helps students to consider the concept of infinity.
 
Understanding Accommodations and Modifications
Understanding  Accommodations and ModificationsUnderstanding  Accommodations and Modifications
Understanding Accommodations and Modifications
 
microwave assisted reaction. General introduction
microwave assisted reaction. General introductionmicrowave assisted reaction. General introduction
microwave assisted reaction. General introduction
 
How to Give a Domain for a Field in Odoo 17
How to Give a Domain for a Field in Odoo 17How to Give a Domain for a Field in Odoo 17
How to Give a Domain for a Field in Odoo 17
 
Making communications land - Are they received and understood as intended? we...
Making communications land - Are they received and understood as intended? we...Making communications land - Are they received and understood as intended? we...
Making communications land - Are they received and understood as intended? we...
 
Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...
Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...
Explore beautiful and ugly buildings. Mathematics helps us create beautiful d...
 
TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...
TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...
TỔNG ÔN TẬP THI VÀO LỚP 10 MÔN TIẾNG ANH NĂM HỌC 2023 - 2024 CÓ ĐÁP ÁN (NGỮ Â...
 
Key note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdfKey note speaker Neum_Admir Softic_ENG.pdf
Key note speaker Neum_Admir Softic_ENG.pdf
 
Application orientated numerical on hev.ppt
Application orientated numerical on hev.pptApplication orientated numerical on hev.ppt
Application orientated numerical on hev.ppt
 
Accessible Digital Futures project (20/03/2024)
Accessible Digital Futures project (20/03/2024)Accessible Digital Futures project (20/03/2024)
Accessible Digital Futures project (20/03/2024)
 
ICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptxICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptx
 
Unit-IV- Pharma. Marketing Channels.pptx
Unit-IV- Pharma. Marketing Channels.pptxUnit-IV- Pharma. Marketing Channels.pptx
Unit-IV- Pharma. Marketing Channels.pptx
 
Unit-V; Pricing (Pharma Marketing Management).pptx
Unit-V; Pricing (Pharma Marketing Management).pptxUnit-V; Pricing (Pharma Marketing Management).pptx
Unit-V; Pricing (Pharma Marketing Management).pptx
 
How to Create and Manage Wizard in Odoo 17
How to Create and Manage Wizard in Odoo 17How to Create and Manage Wizard in Odoo 17
How to Create and Manage Wizard in Odoo 17
 
General Principles of Intellectual Property: Concepts of Intellectual Proper...
General Principles of Intellectual Property: Concepts of Intellectual  Proper...General Principles of Intellectual Property: Concepts of Intellectual  Proper...
General Principles of Intellectual Property: Concepts of Intellectual Proper...
 
Sociology 101 Demonstration of Learning Exhibit
Sociology 101 Demonstration of Learning ExhibitSociology 101 Demonstration of Learning Exhibit
Sociology 101 Demonstration of Learning Exhibit
 
Spellings Wk 3 English CAPS CARES Please Practise
Spellings Wk 3 English CAPS CARES Please PractiseSpellings Wk 3 English CAPS CARES Please Practise
Spellings Wk 3 English CAPS CARES Please Practise
 
Basic Civil Engineering first year Notes- Chapter 4 Building.pptx
Basic Civil Engineering first year Notes- Chapter 4 Building.pptxBasic Civil Engineering first year Notes- Chapter 4 Building.pptx
Basic Civil Engineering first year Notes- Chapter 4 Building.pptx
 
Grant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingGrant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy Consulting
 
Unit-IV; Professional Sales Representative (PSR).pptx
Unit-IV; Professional Sales Representative (PSR).pptxUnit-IV; Professional Sales Representative (PSR).pptx
Unit-IV; Professional Sales Representative (PSR).pptx
 

A Summary Of OR MS Software On Microcomputers

  • 1. Annals of Operations Research 5(1985/6)613 -630 613 A SUMMARY OF OR/MS SOFTWARE ON MICROCOMPUTERS R. SHARDA Department of Management, Collegeof BusinessAdministration, OklahomaState University,Stillwater, OK 74078, USA Abstract This paper provides a summary of some OR/MS software for microcomputers. Lists of packages in many management science techniques are included. A biblio- graphy of other papers is also included. We also summarize the general features of packages in each category. Finally, we mention a few sources to help the reader keep up with the developments in this area. Keywords and phrases Microcomputers, computer programs, mathematical optimization, management science, operations research. programming, software, 1. Introduction The explosive growth of the use of microcomputers in the last five years has the potential of changing the way OR professionals perform and present the results of analysis. The April 1981 special issue of the Journal of OperationalResearchSociety on microcomputers was the first indicator of the recognition of the importance of micros to OR analysts and users. Belshaw [ 1] suggests that micros offer a new oppor- tunity for the distribution of OR models. The spreadsheet interfaces to OR/MS soft- ware and their links to graphics packages can help an OR analyst present results in a different form. The graphics capability of Macintosh-like computers can be exploited to develop better network-oriented software. Ranyard [16] discusses general advantages and disadvantages of using a micro in OR applications. Bischoff and Dowsland [2] describe a successful application of micros for product design and distribution. The programs had originally been developed for the mainframe and later transferred to a microcomputer. The use of a micro in this case brought 'the speed of the computer together with the skill and experience of the end user'. 9J.C. Baltzer A.G., Scientific Publishing Company
  • 2. 614 R. Sharda,OR~MSsoftwareon microcomputers Geoffrion [5] discusses many aspects of the use the micros in OR/MS. He suggests that by offering better economics of computing and a 'splendid' learning environment, the mass market of microcomputers is creating a large group of poten- tial users of OR/MS technology. Users experienced in building models on spreadsheets to perform 'what if' analysis are willing to build models for 'what is best' analyses. This has created a tremendous opportunity for software developers to publish OR/MS software. A great variety of packages is now available. New packages are being introduced practically every week. Each issue of Interfaces or OR/MS Today contains advertisements for new software. Some attempts have been made to keep the OR/MS community aware of the developments in this field. For example, OR/MS Today publishes a column on personal computing, edited by Terry P. Harrison. Management Science Roundtable has commissioned two studies in this area. A special issue of Computers and OperationsResearch (1986) will be devoted to OR/MS software. One objective of this volume is to provide information on state-of-the-art of algorithms and software for optimization. Proliferation of optimization software for microcomputers suggests that the readers of this issue would be interested in a summary of currently available OR/MS software for microcomputers. This paper provides lists of OR/MS software compiled by techniques. We also summarize the general features of packages in each category. Finally, we mention a few pointers to help the reader keep up with the developments in this area. The scope of this compendium is limited. The paper includes software offering the most commonly used OR/MS techniques. It also summarizes the forecasting and simulation software. Some OR/MS techniques are not represented here. This is either because the author is not aware of any software offering those techniques, or de- tailed surveys have already been published elsewhere. The information presented here was gathered from a previous survey, OR/MS journals, popular microcomputer journals and personal communication. But the list is by no means exhaustive. It only represents the extent of the author's knowledge of such software. Any omission is unintentional and not an indicator of the quality of the package. Further, project management, statistical analysis and simulation software has received considerable coverage elsewhere and the summaries are readily available. This paper mentions the references without reproducing the tables from those papers. Thus, an analyst looking for a package in these areas should consult the original studies for a more complete list. The paper also does not include the financial modeling software sold as spreadsheet or decision support software. The wide publicity already available for such software allows us to concentrate on more specific packages. Finally, the paper summarizes mostly the software available for the IBM-PC or compatible machines running MS-DOS operating systems. These machines have the greatest variety of software available for them. The extra RAM capacity and math 8087-coprocessor availability have made large-scale number crunching on a micro feasible. These features
  • 3. R. Sharda, OR/MS software on microcomputers 615 were not available on the earlier micros, so the software base did not develop at the same level. We do indicate the availability of a package on other types of machines, where known. 2. Evaluation criteria Erikson and Turban [4] list several criteria to be considered in evaluating software for classroom use. Golden et al. [8] also mention the following criteria: program environment, user interfaces and support, program scope and capabilities and its effectiveness and efficiency. The basic features to be compared in selecting OR/MS software can be restated in the following terms: functional, capacity, input and output related features. A brief explanation of each follows. FUNCTIONALITY The packages differ in their ability to solve problems. For example, not all linear programming packages have sensitivity analysis capabilities. Even when a package has sensitivity analysis available, it may not permit addition of a constraint/variable. Further, it may not save the inverted matrix of the optimal solution for a quick post optimality analysis, but instead require solving the problem all over again after each change. Forecasting programs also differ in the methods available to the user. Some offer a smorgasbord of methods, whereas others include only regression or moving average and exponential smoothing. The also differ in graphical presentation of data. As a third example, project management software has a wide variation in resource management options. Thus a thorough check of functional features of the available software is desirable. CAPACITY/SIZE LIMITATIONS A major determinant in selecting a package is the size of the problem it can solve. Most packages describe the size of the dataset they can handle. For example, in LP problems the number of constraints, variables and non-negative elements are limiting factors. In forecasting packages, the number of data items in a series, as well as the number of time series, may be restricted. In selecting a PERT package, one would consider the number of activities for each project and the allowed number of subprojects, A related issue is the speed of execution. Given that a lot of number crunching is involved in most OR/MS software, the program has to be written well to execute fast. One way for programs to improve speed (as well as accuracy) on MS-DOS com- puters is to use an Intel 8087 math coprocessor. Use of the coprocessor can speed execution up by 15-500%. The software must be written to take advantage of the
  • 4. 616 R. Sharda, OR/MS software on microcomputers chip. Fortunately, a number of OR/MS packages support this coprocessor, but not all do so. It should be an important criterion in selecting a package for any large-scale application. INPUTFEATURES The ease of data and command input for software can sometimes determine its eventual acceptance by the user. Some programs are menu driven, whereas others are strictly command driven. Menu driven programs are easy to learn but they can become boring for an experienced user. A combination of the two modes is ideal for command entry. A program should also be able to accept data in a number of ways. For a beginner, the program should provide a data editor/manager with appropriate prompts. For an experienced user, as well as for solving larger problems, the program should be able to accept data input created externally, such as in an ASCII file, by a word processor. To offer better portability, the program should be able to accept data in a standard format. For example, many spreadsheet packages can output data in the DIF (data interchange format) so that it can be read by other packages. A good OR/MS package should be able to read such files. For linear programming applications, a program should be able to accept problem input in the popular MPS format. This would permit easy transfer of a problem from mainframe to micro. OUTPUTFEATURES One reason microcomputers are becoming so popular is the ability of a user to control the periPherals of the computer. A user should be able to decide which part of the output goes on screen and which goes to the printer. For example, an LP package could allow one to see only the most important output on screen and print the rest on the printer, if requested. Further, one should be able to access this output through a data file. This access should be twofold. If the solution is readable by a spreadsheet package such as Lotus 1-2-3, one will be able to present some of the results graphically. Another paper in this volume [19] explains this is detail. Second, if the solution file can be accessed by a word processing package, the report can incorporate the results of the analysis efficiently without the need of retyping all the numbers. The program should be able to use graphics where appropriate. For example, data plots should be available in performing statistical or time series analysis. One should be able to print these plots not only on the printer but also on popular plotters. If a project management package is being used, it should be able to produce a network diagram, Gantt chart, etc. OTHERCONSIDERATIONS Some of the other obvious considerations include minimum required hardware, and price. Most programs can use additional memory. Given that the price of memory
  • 5. R. Sharda, OR/MS software on microcomputers 617 is falling, it is usually a worthwhile investment and can result in a significant improve- ment in performance. Even those programs which can not use additional memory will execute faster if the floppy disk data is copied into a RAM disk. Many utilities are now available to define part of the computer's RAM as a disk drive. In fact, most sellers of memory boards include such software, The price of the software also varies widely. And it does not seem to correlate positively with performance. Of course, the potential benefits of using any of these packages can far outweigh the cost of the software. It is, however, worthwhile to com- pare costs as well as performance. One other consideration should be user support. Many companies offering these packages are quite small. One should consider whether the company is strong enough to update the software and help correct any problems. While no guidelines can be offered, one can perhaps ask for references to check on the strength of the software publisher. 3. What is available? We now summarize software for each of the major categories. Where the author is aware of a more detailed summary published elsewhere, the reference is cited. For a complete list, the reader should also consult these references. The information in the following tables is current as of this writing. The prices listed here are for single copies for commercial users. Many of the software publishers have educational prices and/or site license agreements available. Space does not permit listing all such details. As mentioned earlier, the software included in these tables runs on IBM-PC or compatible microcomputers. Some of the programs also run on other types of microcomputers. The reader should check with the publishers for such details. LINEARPROGRAMMING One of the most important OR application tools, linear programming (LP), has been made available on the PC by a number of software publishers. The author is aware of at least two dozen packages on the market. Many of these packages were summarized by Sharda [18]. Additional packages are summarized in table 1. The reader should also consult the original paper for a complete list. The information in table 1 is based only on the available literature and not on actual usage. Some items were not discernible from the brochures. The linear programming software on the market varies widely in price, size of the problem solvable by the software, and other features. The largest LP problem one can solve on a PC appears to be 2000 rows and 2500 columns, at least thus far. Of course, as problem size increases, the speed of execution is affected adversely. Having an 8087 math coprocessor can help. But, the software has to be able to take advantage of the math coprocessor.
  • 6. 618 R. Sharda, OR~MSsoftware on microcomputers 0 o 0 o~ t~ "a o o o ~z Z % ~o~ oo 0 ~.~ ~ ~.~ ~ o ~ o o o o ~ o Z Z Z ~ Z ;~ o o ~ o 0 ;~ Z I 0 0 o 0 0 0 0 0000 0,-, V/ N V/V/ ~ Z ~ 0 O 0 O 0 O 0 O 0 (.J 0 0 d ~o ~d_~o o ,N 0 0 E II N Z a 0 E II 2 oo II
  • 7. R. Sharda, OR/MS software on microcomputers 619 These packages handle input/output in many ways. Some allow the data entry only one coefficient at a time, whereas others can accept input from a file created by using an editor/word processor/matrix generation program. A few packages accept data in MPS format as well, which permits portability of a model between micro and mainframe, as well as between different LP packages offering this feature. Some even interface with packages such as Lotus 1-2-3, Multiplan or Visicalc to accept a problem formulation using that package. This interface may also permit one to process the LP output using the spreadsheet package to prepare better reports. While OR/MS software has not proliferated on Macintosh yet, the author is aware of one code running on Macintosh. This code is being developed by Professor James Ho of the University of Tennessee. According to Professor Ho, the code is able to solve problems with 500 rows and 5000 non-zeroes on the 512 K Mac. OTHER MATHEMATICALPROGRAMMINGSOFTWARE Some of the LP packages also offer integer programming as an option. This allows one to solve mixed integer programming (MIP) problems. The author is aware of five packages which include significant IP or MIP features. In addition, an instruc- tional package, MICROSOLVE/OR by Jensen [10] also includes an integer program- ming routine. Table 2 lists major packages offering MIP, nonlinear programming and transportation method. The integer programming capabilities of the five packages vary. LINDO is able to solve a problem with 200 integer variables, but these are restricted to be 0/1 vari- ables. The other four packages can accept a combination of 0/1 as well as general integer variables, but the problem size is restrained somewhat. |LP-88 assumes all integer variables to be general. If a variable is to be a 0/1 variable, one must enter two constraints, x ~< 1 and x/> 0. This can add to the number of constraints considerably. Further, it allows for 45 integer variables and 255 constraints (including the con- straints for 0/1 variables). MIP-83, an option of LP-83, also works on a mixed integer programming problem. It is possible to specify 0/1 variables without including the two bound constraints. The maximum number of integer variables is not specified in the manual. However, the author is aware of one application containing 40 integer vari- ables where MIP-83 could not obtain a solution in 'reasonable' time. The number of integer variables is not a sufficient indicator of the power of any MIP program. The formulation of the problem can affect the number of nodes to be searched. As this number increases, the storage space can also be a bottleneck, affecting the overall performance of the program. As in the case with IP software in general, there is room for improvement in this area. Nonlinear programming software for micros is just beginning to appear. This author has used only one package and is aware of one other. These are also listed in table 2. GINO is an adaption of GRG [12] in the microcomputer environment. It
  • 8. 620 R. Sharda, OR/MS software on microcomputers e~ N ~ o ~ 9 ~ ..~ 0 0 9 ~ o o ~ o .~ ~ .~ ~ ~ ~ ~.~ ~x ~x 0 0 o r..) z~ .q. .~ • O 0 0 0 ,-1 r ,~ d o ~ ~ ~.~ ~ ~.~ ~ ~m 0 0 0~o 0 X 0 Z 0 0 d 2 0 Z
  • 9. R. Sharda, OR/MS software on microcomputers 621 0 O, [.-, N ~ < ,.c: ,.0 0 ~ ~o 0 0 X )< 0 0 0 0 ~ ~ ~ ~ =~ ~ ~o~ "~ ~ 9 ~__, ~r~ :~ ~.~ ~ o 0 0 ~, 0 O0 ,~" 9 ~ ~ ~.= "~ ~.~ . ~ 9 ,~ ~ ,~ 0 ~ '~ "~ ~ ~,~ ~o~.~ 9 ~~ ~ ~-~ ~ oo
  • 10. 622 R. Sharda,OR/MSsoftwareon microcomputers is designed to work in the interactive mode similar to LINDO. The maximum problem size on a PC is limited to 30 rows and 50 columns. While it may not be able to solve very large problems on a PC, it can serve as an excellent instructional tool. SYSOPT is designed for unconstrained as well as constrained optimization. It is reported to include a number of search routines such as Fibonacci search in one dimension and PoweU, Fletcher and Reeves' search in n dimensions. It also includes a penalty function method for constrained optimization. Wasil and Golden [9] provide more comments on this program, as well as many other nonlinear programming packages. They also summarize some software capable of solving a system of non- linear equations. Two other packages of interest are MPPII and TSA-88. MPPII is actually a linear programming system, but it also offers a transportation simplex implementation. It is able to solve a problem with 70 rows and 100 columns. TSA-88 is able to solve problems with up to 510 sources and sinks. The publisher of TSA-88 has also an- nounced a microcomputer implementation of Karmarkar's algorithm for solving LP problems. A set of programs is available in MP. This package can be purchased in full, or the modules can be obtained at a lower price. SIMULATION Many of the popular simulation packages on the mainframe are now available for microcomputers. The March 1984 issue of Byte summarizes simulation on micros. Pratt [15] lists some simulation packages running on various microcomputers. Table 3 lists some simulation packages available for a PC. Law and Kelton [13] provide a comparison of the mainframe versions of GPSS, SIMSCRIPT and SLAM. GPSS is a discrete event simulation language following the process interaction approach. Its block structure makes model development relatively easy. A major criticism of the mainframe version of GPSS has been that its clock is integer-valued and random vari- ates from different distributions are not generated automatically. However, informa- tion on the PC version claims to have fixed the latter problem. A continuous simulation language following the systems dynamics approach is Dynamo. Its micro version is now available as Micro-Dynamo. It appears to be a reasonably complete implementation of mainframe Dynamo. All the other simulation languages listed in table 3 offer both discrete and continuous simulation features. SIMSCRIPT/PC is sold as SIMLAB. The mainframe SIMSCRIPT II.5 is known to offer natural language-like syntax for easy modeling. It is also able to perform sta- tistical analysis of simulation output data. The PC version is supposedly a full imple- mentation of the mainframe version. SLAM II/PC is also a PC version of SLAM II running on mainframe. It offers an easy network modeling structure as well as discrete and continuous modeling. The PC version can output the data to a DIF file, so that further processing, plotting, etc. can be performed on the PC. This is a very convenient feature of this package.
  • 11. R. Sharda, OR/MS software on microcomputers 623 Table 3 Simulation software Software Publisher Price Type of (USA) (U.S.$) simulation GPSS/PC Minuteman Software 900 Discrete P.O. Box 171 Stow, MA 01775 (617) 897-5662 Micro Dynamo Addison-Wesley Publishing 295 Continuous Company, Inc. Jacob Way Reading, MA 01867 (617) 944-3700 C.A.C.I 6500 3344 N. Terry Pines Ct. La Jolla, CA 92037 (619) 457-9681 Systems Modelling Corp. 1500 P.O. Box 10074 State College, PA 16805 (814) 238-5919 Pritsker & Associates 975 P.O. Box 2413 West Lafayette, IN 47906 (317) 463-5557 Professor C.C. Barnett 125 Walla WaUaCollege College Place, WA 99324 (509) 522-1346 Mitchell & Gauthier Associates 290 Baker Ave. Concord, MA 01742 (617) 369-5115 SIMSCRIPT/PC Combined SIMAN Combined SLAM II/PC Combined Micro Passim Combined ACSL/PC Continuous Micro-Passim is written in Turbo Pascal and takes advantage of the advanced features of the language. It is a relatively inexpensive package offering capabilities similar to most other simulation software.
  • 12. 624 R. Sharda, OR/MS software on microcomputers Table 4 Some project management packages Name of package Company name & address Price (USA) (U.S.$) Number of activities Resource loading MICROPERT MICRO GANTT PATHFINDER HARVARD PROJECT MANAGER QUICKPLAN Primavera Advanced Project 6 Sheppard Software Company 350 4750 Clougli Creek Road Redding, CA 96002 (916) 222-1553 Earth Data Corporation 395 P.O. Box 13168 Richmond, VA 23225 (804) 231-0300 Morgan Computer Co., Inc. 100 10400 N. Central Expressway Suite 200 Dallas, TX 75231 (214) 739-5895 Harvard Software, Inc. 395 521 Great Road Littleton, MA 01460 (617) 486-8431 Mitchell Management 995 Systems, Inc. 2000 W. Park Drive Westboro, MA 01581 (617) 36649800 Primavera System, Inc. 2500 29 BalaAvenue Bala Cynwyd, PA 19104 (215) 667-8500 SoftCorp Inc. 199 2340 State Road 580 Suite 244 Clearwater, FL 33575 (813) 799-3984 220 NL 3160 NL 250 10000 250 No Yes No Yes Yes Yes Yes NL = limited by disk space.
  • 13. R. Sharda, OR~MSsoftware on microcomputers 625 PROJECT MANAGEMENT Project management seems to be the only technique in an OR/MS analyst's toolbox that has received significant attention on the micro market. Most trade publications and magazines have surveyed this group of software and printed com- parative tables. For example, Ruby [17] compared 38 project management systems. A recent book by Gido [6] compares 120 project management packages. Some of these are listed in table 4. As before, the inclusion here is only to provide a sample, not an endorsement. It will be worthwhile to read one of the comparative articles before selecting a project management system. A wide variety of project management software is on the market. The packages available can handle projects with as few as less than 100 activities to upwards of 1500 activities. The programs vary in their resource allocation and cost management features. Some programs allow up to 100 resources per project. Others offer no resource alloca- tion. A few packages also allow linkages between different projects for optimal re- source management. The programs also offer varied degree of cost management. Some programs allow one to compare actual expenditures to budgeted amounts. A few have no cost- related features. For project management applications requiring complete cost tracking, one will probably have to invest in the more expensive packages of the group. The reporting features of micro project management systems also vary widely. While almost all of the packages prepare a Gantt chart, a few are also able to print or plot a network chart. This may be a plus for reporting purposes. One may also wish to consider those packages which are able to print task lists as well as a loading report for each individual resource. Some packages are able to prepare reports aggregated at various levels. This feature may be useful for preparing reports for managers at different levels. The ease of use of these packages varies as well. Some are quite technical as they are mere downloads of mainframe packages. Others take advantage of the micro- computer technology fully. A few have been developed using the command/menu structure of packages such as Lotus 1-2-3. Overall, a large variety of software exists in this area. One should compare the features and possibly examine a demonstration copy before purchasing, since these packages are usually quite expensive. STATISTICALANALYSISAND FORECASTING Statistical analysis has also received a lot of attention in the microcomputer market. Carpenter et al. [3] summarized 24 statistical analysis packages running on various micros. More recently, Littman [14] compared 19 statistical analysis packages for the IBM-PC and compatibles. This section will not repeat that comparison, but will cover only the forecasting software. While some of the comprehensive statistical
  • 14. 626 R. Sharda, OR~MSsoftware on microcomputers Table 5 Some forecasting packages Name of package Company name & address Price (USA) (U.S.$) Comments SmartForecasts SIBYL/RUNNER AUTOBJ ESP SSI-SSF 4CAST/1 WG/ARIMA Forecast Plus Smart Software 495 392 Concord Avenue Belmont, MA 02178 (617) 489-2743 Applied Decision Systems 995 Temple, Barker & Sloane, Inc. 33 Hayden Avenue Lexington, MA 02173 (617) 861-7580 Automatic Forecasting 795 Systems, Inc. Statistical Consultants P.O. Box 563 Hatboro, PA 19040 (215) 675-0652 ALPHA Software Corp. 795 30 B. Street Burlington, MA 01803 (617) 229-2924 Scientific Systems, Inc. 495 54 Rindge Avenue Ext. Cambridge, MA 02140 (617) 661-6364 Heurix Computer Products 525 P.O. Box 9227 Morristown, NJ 07960 (201) 267-2806 The Winchendon Group 350 3970 Lakota Road P.O. Box 10114 Alexandria, VA 22310 (703) 960-2587 Walonick Associates 400 5624 Grand Avenue South Minneapolis, MN 55419 (612) 340-1515 Has own operating system or can be run with MS-DOS. Graphics adaptor required. PC/MULTBJ available for multi- variate analysis. Links to Lotus 1-2-3 8087 coprocessor recommended. AUTOBOXX available for multi- variate analysis 8087 coprocessor recommended. Econometric forecastingpackage State-space modeling package Supports color plotter Has own operating system. Auto- matic version-EASI/ARIMA available for US$ 350. Also sells Econometric Linear Forecasting (ELF) for US$ 350. Also sells a general statistical analysis package, StatPac.
  • 15. R. Sharda, OR/MS software on microcomputers 627 analysis packages also include forecasting routines, those packages are not included here. Packages which emphasize forecasting only are summarized. A list of major systems is presented in table 5. The forecasting software is reasonably homogeneous in its data management offerings. Most packages are able to accept data from external sources in the form of DIF or ASCII fries. They also offer a graphical presentation of historical as well as forecast data, though not necessarily in color. Only one package, 4CAST/1,can plot data on a color plotter. However, others can also accomplish this by creating a DIF file and using a graphics package with the file. All of the packages also offer commonly used data transformations. The software varies in the methods available for forecasting. Five packages, SYBIL/RUNNER, ESP, Forecast Plus, 4Cast/1 and SmartForecasts, offer a number of methods for forecasting. All of these packages cover the commonly used methods such as moving average, exponential smoothing, filtering, regression and decomposition. SmartForecasts and 4Cast/1 make excellent use of the color graphics features of a microcomputer. Only 4Cast/1 offers the Bayesian technique. SmartForecasts offers a similar technique called judgmental forecasting whereby the user is allowed to eye- ball the forecast and modify it, based on earlier experience. ESP is a subset of mainframe software. ESP offers two-stage and three-stage regressions using various options. It also offers nonlinear estimations as well as uni- variate Box-Jenkins modeling. SIBYL/RUNNER offers eighteen techniques for uni- variate analysis plus multiple regression. A companion package, PC/MULTBJ is avail- able for multivariate modeling. It also makes initial suggestions for techniques to be used, based upon preliminary data analysis. A few packages have taken this concept further and offer a somewhat auto- mated version of forecasting. SmartForecasts, in its automatic mode, conducts a 'tournament' among competing methods and selects the best method for forecasts. It does not include the Box-Jenkins technique (B-J) in its evaluation, thus leaving a major tool out. EASI/ARIMA, a complement of TWG-ARIMA, claims to be able to automate the univariate Box-Jenkins modeling methodology. Another package, AUTOBJ, is reportedly able to build B-J models automatically. A second product by the same publisher, AUTOBOXX, offers automatic multivariate Box-Jenkins modeling including transfer-function methodology. It also claims to include an intervention detection system. Automatic forecasting option is also available in SSI-SSF, which uses the state-space forecasting technique. Of course, all of the 'automatic' packages can be used to perform forecasting based on the analyst's skill in modeling. There is some concem about the ability of a program to mimic the iterative process of modeling by a forecaster. More tests of these packages are necessary to validate their claims. It appears that the structure of Box-Jenkins modeling can be exploited to come up with a tentative model automatically, to be further refined by the analyst. Perhaps this is an example of a rule-based expert system. In any case, the performance of these automatic packages needs to be tested thoroughly.
  • 16. 628 R. Sharda, OR/MS software on microcomputers Overall, the state of forecasting software appears to be good and improving. Many comprehensive statistical packages not reported here also include at least some forecasting options, thus making the selection even wider. This software will probably see even more growth if the 'automatic' software packages can make statistical analysis easier for micro users. OTHER PACKAGESOF INTEREST A list of other software of interest to the OR/MS user is provided in table 6. Erikson and Turban [4] summarize most of the educational software in their working Table 6 Other packages of interest to OR/MS analysts Software Publisher Price Techniques Comments (USA) (U.S.$) Expert-Ease Expert Systems 695 Expert system Spreadsheet-type 868 W. End Avenue generator input New York, NY 10025 Exsys Exsys 300 Expert system Rule based P.O. Box 75158 generator Contr. Street 14 Albuquerque, NM 87194 (505) 83645676 Insight Level5 Research 95 Expert system Rule based, simple 4980 South A-1-A generator system, best as an Melbourne Beach, FL 32951 educational tool (305) 676-5810 HR-FLOW HR Management Associates 795 Markov Human resource P.O. Box 64070 analysis planning system Tacoma, WA98464 (206) 564-9704 paper. A lot of attention is being paid to expert systems software. For example, Management Science Roundtable has commissioned two studies in this area. These reports will be available through ORSA/TIMS later this year. Two packages which are advertised extensively are Expert-Ease and Exsys. These are not expert systems but expert systems generators. Expert-Ease is a spreadsheet-type system which is applicable to classification-type problems requiring expert opinion. It is able to generate rules using the examples provided by the user. Exsys is an IF-THEN rule based system for
  • 17. R. Sharda, OR/MS software on microcomputers 629 similar problems, but also includes a probabilistic element. Insight is also a simple rule-based expert system generator. A number of other packages will probably be available within the next few months. An implementation of Markov analysis of manpower supply and demand is available in HR-FLOW. This is a straightforward application of Markov analysis. The user is required to enter the number of job levels, time periods, a transition matrix, the starting number of employees and projected demand at each job level. The pro- gram is able to calculate total projected supply and excess/shortage in each category. It requires the user to enter the data by modifying the program written in MBASIC (by using DATA statements). Vehicle routing software has also been available for the PC. Golden and Bodin [7] present a survey of such software. Golden et al. [8] also summarize much of the decision analysis software available for the micros. The reader should refer to these papers for tables similar to the ones presented here. HOWTO KEEPUPTO DATE Microcomputers in general and OR/MS software in particular are growing industries. Many new packages are being introduced and existing packages are being upgraded. Here are some pointers on how to stay in touch with the latest developments. First of all, Interfaces and OR/MS Today are good sources for new product advertisements. A number of products listed in this paper were advertised in these two journals. Interfaces also publishes papers in this area occasionally. A column on personal computing and OR/MS appears regularly in OR~MSToday. It usually includes new product announcements and product reviews. The column also identifies papers of general interest to the OR/MS community. Two other journals which have pub- lished microcomputer-related articles are Computers and Operations Research and Journal of Operational Research Society. More are sure to follow. A special issue of Computer and OperationsResearch in early 1986 will include comparison of many OR/MS packages. There are two major sources of on-line information. International Software Database Corporation, Fort Collins, Colorado, publishes MENU - The International Software Database. This database contains short descriptions of software with in- dexing by broad application categories, techniques, compatible computers, and so on. It is updated monthly. A related source is Microcomputer Index, published by Data- base Services, Los Altos, California. This index is a subject and abstract guide to articles from over fifty microcomputer journals. It is updated quarterly. Both of these databases are accessible through Dialog (Dialog Information Services, Palo Alto, CA (800) 227-1927). A diskette version of Microcomputer Index has just been announced.
  • 18. 630 R. Sharda, OR/MS software on microcomputers 4. Conclusion This paper has presented a listing of some OR/MS software on microcomputers. A large variety of software is now available. The advantages of an independent CPU with direct control of peripherals offers a new way of delivering OR/MS models to clients. This paper provides the analyst with at least partial information about these opportunities. References [1] P.N. Belshaw, Microcomputers and management science, Interfaces 12, 6(1982)105. [2] E. Bischoff and W.B. Dowsland, An application of the micro to product design and distri- bution, J. of Opl. Res. Soc. 33, 3(1982)271. [3] J. Carpenter, D. Deloria and D. Morganstein, Statistical software for microcomputers, Byte 9, 5(1984)234. [4] W.J. Erikson and E. Turban, Teaching operations research on microcomputers, Research Report DS-84-09, School of Business Administration, University of Southern California, Los Angeles (1984). [5] A.M. Geoffrion, Can MS/OR evolve fast enough? Interfaces 13, 1(1983)10. [6] J. Gido, ProjectManagementSoftwares (Industrial Press, Inc., New York, NY, 1985). [7] B. Golden and L. Bodin, Vehicle routing software for microcomputers: A survey, Com- puters and Operations Research (1986) forthcoming. [8] B. Golden, A. Hevner and D. Power, Decision insight systems for microcomputers: A critical evaluation, Computers and Operations Research (1986) forthcoming. [9] B. Golden and E. Wasil, A survey of nonlinear programming software, Computers and Operations Research (1986) forthcoming. [10] P. Jensen,Microsolve/Operat~onsResearch(Holden-Day, Oakland, CA, 1983). [11] J.A. Kastner and S.J. Hong, A review of expert systems, Eur. J. Oper. Res. 18, 3(1984)285. [12] L.S. Lasdon, A.D. Warren, A. Jain and M. Ratner, Design and testing of a generalized re- duced gradient code for nonlinear programming, ACM Transactions on Mathematical Soft- ware 4(1978)34. [13 ] A.M. Law and W.D. Kelton, SimulatT"onModelingand Analysis(McGraw-Hill, New York, NY, 1982). [141 J. Littman, Statistical software: Making it simple, PC Week 2, 3(1985)45. [15 ] C.A. Pratt, Going further: A compendium of conferences, organizations, books and software for simulationists, Byte 9, 3(1984)204. [161 J.C. Ranyard, Introducing the microcomputer into the OR department, J. of Opl. Res. Soc. 32,4(1981)277. [17] D. Ruby, On the critical path, PC Week 2, 3(1985)51. [18] R. Sharda, Linear programming on microcomputers: A survey, Interfaces 14, 6(1984)27. [ 19] R. Sharda, Optimization using spreadsheets on a microcomputer, Ann. Oper. Res. 5(1985/6) 599.