HOME
QUICK SEARCH
AUTHORS
KEYWORDS
HELP
BUGSSUGI 1985 Papers - Index
212 papers. Source: SAS Users Group International conference proceedings.
| # | Title× | Authors× | Abstract× | Keywords× | Extras | Pages |
|---|---|---|---|---|---|---|
| 1 | A Customer Sampling and Reporting System Using SAS Software | Barbara J. Matthews, Sue M. Worline |
This paper discusses a customer sampling and reporting system which permits greater accessibility to the contents of an electric utility's customer billing file for the purpose of data analysis and research. The system standardizes the process of obtaining customer extracts, random samples, statistical summary reports, and labels from the customer billing file. SAS is used to manipulate files and data, perform statistical analysis, and generate reports. The system is interactively driven by TSO CLIST's and SAS/FSP to provide an easy-to-use fl11-in-the-blank process to select criteria for the extraction, sample, or report. The SAS Macro facility generates SAS code which produces the extract of the population, samples r and reports of the customer file. The primary objective of the system is to transform a billing file into a marketing research tool |
sample, dataset, proc step, macro facility, proc means | source | 6 |
| 2 | A Dynamic Work Station Loading Program | Paul H. Grant Jr. | output delivery system, proc format | source | ||
| 3 | A Generic Data Entry Quality Assurance Tool | Paul Kanciruk, Aaron E. Rosen |
COMPARE is an important SAS quality assurance tool for data base management that is especially useful when the SAS Full-Screen Product (SAS/FSP) is used for data entry. A good way to check input data for typographical errors is to have the same information entered by two different people and compare the versions for differences. This has been standard practice for data entered via keypunch machine on CARDS; however, it has not been as easily accomplished for data entered using SAS/FSP, via terminal, directly into SAS data sets. To facilitate this procedure, a SAS macro, COMPARE, has been developed at Oak Ridge National Laboratory. as part of the U.S. Environmental Protection Agency's National Surface Water Survey Data Base Management Project. Oata are directly and independently entered into two SAS data sets. COMPARE then automatically compares the data sets and prints out the observation number, variable name, and values for any nonmatching observations |
dataset, quality assurance, quality control, macro definition | source | 3 |
| 4 | A Laser Printer Can Enhance the Appearance of SAS Procedures | David J. O'Sullivan | It is widely known t~at laser printers produce high quality printouts. However, they can also be used to print special characters which are self-designed. Therefore, with this capability one can produce more professional looking tables by highlighting specific words in TITLES, LABELS and individual data points. In addition, one can alter SYMBOL, FORMCHAR and LABEL options to greatly enhance the appearance of PRoe | proc step, proc freq | source | 6 |
| 5 | A Macro Library Facility for SAS Software | Ross Merlin | To use a stored SAS macro requi res two operations: the macro must be loaded (by %INCLUDE), then it must be invoked. Once a macro has been loaded, it remains available throughout the session and should not be reloaded; however, if many macros are being used, it is not reasonable to expect the user to keep track of which macros have been loaded .and which have not. Dther programming 1 anguages 1 cad macros only when needed -- why not the SAS system? The primary purpose of this macro library facility is to let the software manage a collection of macros; i.e. to load macros only when necessary, to make macros readily available through a lfbrary, and to provide a framework for the standardization of parameter names |
dataset, macro call, macro definition | source | 4 |
| 6 | A Monte Carlo Comparison of Extended Mantel-Haenszel Procedures | Daniel W. Griffin | The Mantel-Haenszel (MH) statistic (1) is commonly used to summarize results from sets of two by two contingency tables in case control studies. The test statistic can be motivated as a design-based test directly from finite population sampling theory, in which formal conclusions are restricted to the population of subjects actually observed, and has been recommended for use whether stratification of the control variable was undertake prior to randomization (by design) or after randomization as an analytic strategy, i6e., poststratification. A post stratified (PS) test similar to the MH test has been developed (2). The two tests differ with respect to their variance. Depending on the study design, either the MH or the PS test may be preferred. The simple dichotomization of a risk variable, however, often obscures the full range of association between exposure and disease. The MH procedure has been extended in form by Mantel (3) to 2xc tables. The | randomization, sample, covariance, monte carlo, permutation test, proc freq | source | 6 |
| 7 | A New Tool for the Information Center: SAS/AF Software | Philip James Busby | source | 8 | ||
| 8 | A Parameter Passing Utility for SAS Programs | Paul M. Gilbert, Mark Klompas, David J. O'Sullivan |
The format of the SPARM command is: SPARM, a parameter passing utility, was developed to run standardized SAS programs over different populations of data. This combination of VM/CMS EXEC'S and XEDIT MACRO'S works by preprocessing and then executing a SAS program. Now used as a standard SAS execution utility, SPARM increases programming efficiency and programmer organization | clinical data management, clinical data, data governance | source | 5 |
| 9 | A Personnel Budget System Utilizing the SAS Macro Language and SASIFSP Software | Joseph Buonome, Patricia L. Hines |
At Berlex Laboratories, an ethical pharma ceut i ca 1 company, emp 1 oyee l 5 performance eva 1 ua tions and salary increases are determined on an individual basis. The amount of a raise and the number of months unti 1 eligible again are deter mined by performance rating and salary position within the limits defined for a job grade. Manual tabulation of salary transactions for a workforce of some 700 employees became an over whelming task for annual budget reports and pro- jections. In response to this burgeoning pro- blem, a system was developed for calculating, reporting, and summarizing this information more efficiently. This system, PEBUDY (Personnel Budget by Year), is a macro-driven program with an interactive front end menu. The multi-level menu interface uses SAS/FSP to first prompt the users for the type of run to be made (e.g. fi Ie input/updates, detail, or summary reports) then asks for the file names, parameters, etc. nec essary to complete the requested task |
proc step, dataset | source | 6 |
| 10 | A Project Management System Written in SAS Software | Andrew J. L. Cary, Martin J. Rosenberg |
An easy to use project management sys tem was written in SAS software to track the analysis of clinical trials in a pharmaceutical company environment. This system uses SAS/FSP as an inter active tool to build and review a SAS database and uses SAS programs to produce monthly activity reports, Gantt charts, and calendars indicating start and finish dates for ongoing projects. The system was written and is operated in an environment without SAS/OR* |
proc step, clinical trial | source | 6 |
| 11 | A Quadratic Program for Determining Efficient Frontier Portfolio Compositions Using the SAS Language | Alex Anckonie III | covariance, standard deviation, correlation, macro facility | source | 6 | |
| 12 | A Quality Assurance System for Manufacturing Applications using Base SAS Software on a Minicomputer | Sandra Heard, Judith H. Mopsik |
quality assurance, proc step, proc freq, bar chart, quality control, dataset, put statement, sas base | source | 8 | |
| 13 | A Rotation Scheduling Model Using PROC LP | Duane Walker | This paper will explain a scheduling model that applies to the Circuit Pack rotation update processes. The model will give a complete demand schedule for seed stock circuit packs by cycle to the manufacturing division, minimizing the total cost of investment while meeting the projected schedule required. The results will show that a smaller number of seed stock circuit packs will be required | customer satisfaction, linear programming | source | 6 |
| 14 | A SAS Macro for Logistic Discrimination for Three or More Populations | James E. Dunn, Eileen C. King |
Logistic discrimination can be used to classify an observation into one of several populations. The observation can be qualitative or quantitative. The estimate of the probabil ity of population membership as a function of the independent variables is represented by the posterior probabilities. The SAS procedure LOGIST will give the logistic discrimination function and posterior probabilities when the number of populations is two or the model is ordinal. J. A. Ander son (1972) extended the logistic regression | likelihood function, logistic distribution, maximum likelihood estimation, posterior probability, sample, estimate, predictor variable, regression analysis, covariance, response variable | source | 6 |
| 15 | A SAS Procedure for Evaluating the Performance of Token Ring Local Area Networks | Richard L. Gimarc, James H. Smith |
In this paper the authors present a SAS procedure for evaluating and predicting the performance of token ring-baeed local area networks. The description and parameters of the network are all input through the procedure control carda, then an analytic model of the ring is constructed and executed. A SAS data set is created which contains the important performance metrics of the network, notably predicted queueing times for each station and overall utilization of the network. Examples are provided to show how this procedure can be used to predict the behavior of a token ring in scenarios or projected increases in message traffic. Use of the procedure to perform sensitivity analyses on the network is also illustrated | proc step, sensitivity analysis, standard deviation, key performance indicator, dataset, poisson distribution | source | 8 |
| 16 | A SAS Procedure for Smoothing Generalized Linear Models | William J. Raynor Jr. | regression analysis, likelihood function, binomial distribution, logistic distribution, data cleaning, poisson distribution, logistic regression | source | 6 | |
| 17 | A SAS Routine to Compute Minimum Sample Sizes for Single Attribute Sampling Plans | Melvin T. Alexander | This SASe routine computes the minimum sample size and acceptance numbers required for single attribute plans which are based on the binomial, or Poisson distributions. Using a modified search procedure, the program also determines the operating characteristics and average outgo ing quality curves designed to meet the desired protection levels for both producer and con sumer. Receiving inspectors who must inspect shipped lots of goods may utilize this routine interactively without having to resort to pub lished tables, graphs or formulas found in most acceptance sampling plans. Program description, listing, and examples are provided | sample, poisson distribution, sample size, binomial distribution | source | 4 |
| 18 | A SAS Software to Lotus 1-2-3 Macro | Thomas A. Miron | dataset, proc contents, proc step, macro definition | source | 6 | |
| 19 | A SAS System to Support a Corporate Health and Lifestyle Assessment Program | William M. Taylor | Designed for a corporate environment to determine health trends and potential health problems among workers. and to provide employees with specific personal guidelines to improve their health. over 300 parameters of health are analyzed. Health profile data, including personal information, medical history, current symptoms, medications. use of alcohol and tobacco, diet and exercise and many other items are completed by the employee on an annual basis. This data is combined with a physician's e~amlnation. blood laboratory work up. electrocardiogram. x-ray. vision and audiometric tests plus occupationally-related tests to establish a comprehensive databank for each person. The software, dependent upon tables and computation formulas, relies extensively on the use of macros. It is well structured to allow for easy update and revision. Computations and recommendations are based on responses to specific questions controlling for sex, age group, race and vital measurements | laboratory data, management reporting | source | 4 |
| 20 | A SAS-based LOG System for Keeping Track of Computerized Clinical Trial Data | Ruth F. Quah | crf, sas log, clinical trial, dataset | source | 4 | |
| 21 | A Simple Time-tracking System Using SAS Software | Richard A. Henkle | proc step, dataset, sas library, proc format | source | 6 | |
| 22 | A Successful Formula for SAS Education on a Rock Bottom Budget | Roger A. Chapin | proc tabulate, proc step | source | 1 | |
| 23 | A Tests Result Trend Analysis System Using SAS/GRAPH Software | Susan S. Duckworth | trend analysis, regression analysis | source | 6 | |
| 24 | A Time Series Approach to Modeling Daily Peak Electricity Demands | Michael F. Jacob | Statistical models of peak electricity demand provide the basis for short-term load forecasting and weather normalization applications. Due to the weather sensitive nature of peak demand, a modeling methodology is developed that explicitly accounts for the relationship between peak demand and temperature. By focusing on daily data, this relationship can be identified on a seasonal basis for any given year. A multi-regression framework is initiaHy formulated that provides the basic structur~ for a time series modeling approach. The SAS/ETS ARIMA procedure is used to identify and estimate a Box-Jenkins model that accounts for the non-linear relationship between daily peak demand and temperature through a specification of piecewise-linear transfer functions. Also, since electricity usage varies by day types, pulse intervention variables are introduced to capture the effects of weekends and holidays |
time series analysis, regression analysis, proc step, residual, goodness of fit, proc autoreg, proc arima, mse, null hypothesis, proc reg, normalization, binary variable, cohen's kappa | source | 6 |
| 25 | ACDAS - An Automated Clinical Data Analysis System | Lynn T. Julian, Donald D.M. Tong |
class variable, data validation rules, proc step | source | 6 | |
| 26 | Airline Connection Point Analysis | Dave Baird | This paper describes a SAS Institute, Inc. based system for computing connection cities to be used for Trans World Airlines', Inc. World Wide Availability System. This system utilizes the worlds airline schedules to compute roughly 35,000 markets that contain direct (no plane change) flight service. The direct markets are then processed to find all combinations of connection points, by utilizing spherical trigonometry (with world latitudes and longitudes), market and trip mileages may be used as a base for elliptical analysis. This program will aid us in identifying the best connect points and routings for our airlines' construction of trips in markets that may have no direct service or due to traffic "flow", require connection trips as well as direct trips | dataset, sample | source | 5 |
| 27 | Alternative PROC FACTOR Output | P. Greg Bonner, Jeffrey L. Hancock |
SAS code is presented which provides an alternate reporting scheme for factor analysis output. This alternate format is intended to be more "user friendly"; that is, directly inter pretable by the non-technical analyst. An example is included for review | factor loading, factor analysis, r-squared | source | 6 |
| 28 | An Alternative Method of Passing Data Between SAS Software and ISPF Software | Alan A. Parrow | Interfaces between SAS and ISPF (IBM's Interactive System Productivity Facility) have been the topic of several papers presented at previous SUGI conferences (e.g., Chen, 1984, Svirsky, 19841 Prague, 1983). The topic is likely to remain lively for some time as in creasing numbers of users come to see the great complementarity achieved through conjunctive use of these systems. Most systems which have interfaced SAS and ISPF do not do so at the level of data storage. That is, most often I,sPF generated menus pro vide front-end or back-end processing to call SAS routines for data storage, analysis and maintenance. More recently, attention has been given to interfaces between these software prodUcts which capitalize on the different data-storage facilities used by the products. The present paper falls in the latter category in that it describes a method of passing data between SPF tables and SAS datasets | data step, input statement | source | 6 |
| 29 | An Application of PROC FUNCAT to Randomized Response Data | Martin J. Rosenberg | Randomized response is a technique for collecting survey research data of a sensitive nature. The transformations used in randomized response may also be used to contaminate pre-existing data sets so as to allow their release to external investigators, while protecting the confidentiality of the respondents. A major drawback to randomized response has been the inability to perform multivariable analyses that examine the relationships among two or more variables. This paper presents the necessary methodology for performing multivariable analyses of categorical randomized response data utilizing the SAS· procedure FUNCAT | transition matrix, proc freq, anova, confidence interval, proc step | source | 6 |
| 30 | An Application of the SAS Macro Language in a Research Environment: A Post-processor for PROC LOGIST | Erika Mares Stone | The program presented here was written in response to the request from researchers to manipulate the output from PROe LOGIST in various ways. Many features of the SAS macro language were implemented to make this possible and to make the program flexible and easy to use. The user first directs the output from his PROC LOGIST run to an OS data set with PROC PRINTTO. The post-processor program extracts, from the saved output for each MODEL statement, the names of the independent variables, the beta values, p values, means, standard errors, and other data and prints them in a rearranged |
proc step, proc calis, predictor variable, p-value, macro variable | source | 6 |
| 31 | An Enhanced SAS Macro for Testing Homogeneity of Variance | Dennis W. King, John Henry King, Kevin C. Thompson |
sample, degrees of freedom | source | 6 | |
| 32 | An Environmental Data System using SAS Software on the VAX 11/780 | Dewey Blaylock, Larry Salvo |
dataset, data dictionary, sample | source | 4 | |
| 33 | An Indexing Approach to Ranking Risk in Consumer Credit Portfolios | Marshall Constantino, Richard D. Heberlee |
delinquency, sas base | source | 6 | |
| 34 | An Interactive Interface for Scheduling Your Work Force | Mike Cybrynski | source | 7 | ||
| 35 | An Interactive Technique for Copying and Modifying SAS/FSP Screen Data Set Definitions | Phil Brennan, Michael Cushman |
proc step, out=, dataset, character variable, %if-%then-%else | source | ||
| 36 | An Interactive, Menu-Driven SAS System for Sample Size and Power Computations in Clinical Trial Design | Gregory G. Enas, Walter W. Offen, John D. Taulbee |
sample, sample size, clinical trial, treatment effect | source | 6 | |
| 37 | An Interface with SAS Software and MUMPS | H. William Cooley, Michelle Gallagher-Nelson |
standard deviation, resource allocation | source | 1 | |
| 38 | An Introduction to Macro Processing | Lynn Hannah | macro facility, macro function, &variable, macro variable, macro definition | source | 9 | |
| 39 | Analyzing Tables with Ordered Categories | Ginger Caldwell, John D. Smith |
different models are fitted to this data set using our SAS %MACROs. This paper describes SAS %MACROs which create design matrices for nominal, ordinal, or mixed nominal-ordinal log-linear models, and fit the resulting design matrices using the Newton-Raphson algorithm. We compare these macros with facilities currently available in SAS, and discuss the ba.o;ic characteristics of ordinal models. An example is given using admis sions data for a large research university | degrees of freedom, proc step, main effect, likelihood function, odds ratio, ordinal variable, response variable, macro variable, do over, macro definition, proc freq | source | 6 |
| 40 | Application of the SAS System to Regression Analysis of Tumor Prevalence Data | Saeid B. Amini, Debra L. Palka |
Trend tests play an important role in the analysis of tumorigenicity experiments. Recently, a logistic regression model for comparing treatment groups with respect to tumor prevalence has been proposed and discussed by Dinse, G.E. & Lagakos, S.W. (1983, JRSS, series C, volume 32, No.3, pp 236-248). Unlike the other available methods used to compare treatments in animal carcinogenicity experiments (e.g. Hoel & Walburg, 1972; Peto et (11.,1980; and Gaylor & Kodell, 1983), this method provides direct incorporation and analysis of covariables. A general SAS program is developed to estimate the parameters from a regression analysis of the tumor prevalence model and to compute the overall dose related trend tests using the likelihood ratio, Wald, adjusted and unadjusted score test statistics | logistic distribution, likelihood function, null hypothesis, regression analysis, logistic regression, dose, proc logistic, standard deviation, maximum likelihood estimation, proc step | source | 7 |
| 41 | Applications in Bayesian Analysis | Arnold Zellner | sample, regression analysis, likelihood function, time series analysis, bayesian statistics, posterior probability, expected loss, prior probability, sample size | source | 7 | |
| 42 | Assessing the Impact of VM System Modifications using VMAP and the SAS System | Harriet E. Brichta | proc means, sample, covariance | source | 6 | |
| 43 | Bootstrap Computations | Robert Tibshirani | bootstrap, proc step, correlation, sample, macro definition | source | 5 | |
| 44 | Bounded Influence Regression Using SAS Software | David Giltinan | regression analysis, proc reg, covariance, outlier | source | 6 | |
| 45 | Box Plots and More Box Plots - A Generalized SAS Macro | Donna L. Stock | Boxplots are a useful tool for exploratory data analysis. This paper describes a macro written with the SASR (SAS Institute, Inc., 1982) macro language which produces multiple pages of boxplots appropriately labeled for inclusion into reports or memos. The macro has extended capabilities over the previously documented BOXPLOT procedure PROC SPLOT (Gerig, 1983) such as automatic fence plotting and a capacity for 22 boxplots per page. Labels for each plotted page as well as for each box may be easily changed to allow for production of report quality plots, and plots are routed to a user-specified device such as graphics terminal or graphics plotter |
box plot, proc step, outlier, central tendency, macro variable | source | 6 |
| 46 | Box Plots using SAS/GRAPH Software | Ann Olmsted | confidence interval, macro definition, %if-%then-%else, box plot, proc step, macro variable, proc univariate, sample, proc gplot, call symput, confidence level | source | 7 | |
| 47 | Bring Microcomputers into a Small, SAS Software-oriented Computer Laboratory | Gerald V. Tangren | anova, txt, sample | source | 4 | |
| 48 | Build Your Own Query Language | David Burrage | macro variable, %if-%then-%else, macro definition | source | 2 | |
| 49 | Building an Executive Chartbook Using the SAS System with Examples of New SAS/GRAPH Software Features | Jack Bulkley | Many end-users, such as executives, do not wish to spend the time necessary to learn to produce their own g'raphs. Instead they are only interested in viewing the selected data in graphical form. An executive chartbook can provide this capability. An executive chartbook is a catalog of predefined graphs. Executive chartbook users simply find which graph they want to see and enter the catalog name to view the latest data in the chosen form. This paper is a guide of how to develop an executive chartbook using the SAS System. This paper assumes that the reader is familiar with the SAS System and SAS/GRAPH Software. The examples include enhancements and new features of Version 5 SAS/GRAPH software |
sample, proc step, proc greplay, macro conditional, macro definition | source | 9 |
| 50 | Building Hybrid Systems, Using the SAS System and Microcomputers-A User Survey | J. Harvey Trimble Jr. | Coupling the low cost interactive power of microcomputers and the data management~ analysis, and display capabilities of the SAS System is a tantalizing possibility many SAS users have long contemplated. Several organizations have been able to combine these capabilities and have built such systems. However~ the interchange of information about alternatives and successful appt-oaches this dynamic area limited. between users in has been quite In order to measure the effectiveness, applicability, and operational requirements of building Hybrid Systems which use the mainframe computational pOwer of the SAS System and the interactive power of microcomputers, over fifty organizations who have shown an interest is such developments were surveyed. The results of that survey and an analysis of its implications show what hardware, software, and operational considerations are needed to best build Hybrid systems | source | 4 | |
| 51 | Cartographic Data Bases for SAS/GRAPH Software: A Historical Survey | Donald F. Cooke, Muriel M. Farrington |
Most SAS/GRAPH@mapping is done at the state and county levels of geography using cartographic data bases (boundary coordinates) supplied by the SAS Institute. However, SAS/GRAPH has a comp letely general mapping capability limited only by the kind of data and cartographic databases supplied to it. This paper briefly outlines the history of cartographic database development as the basis for thematic mapping and describes options for SAS users desiring to map data at various levels of geography |
source | 4 | |
| 52 | Changes and Enhancements to PROC TABULATE | Alan Eaton | There are several enhancements to PROC TABULATE available in the next major release of the SAS System. This tutorial begins with a review of the fundamental language features of TABULATE. The rest of the tutorial describes enhancements including correct interleaving of crossed subgroups application of user-defined formats in table cells a CONDENSE option to put mUltiple logical pages on a single physical page a BOX option to put page dimension text, a variable label, or a character string in the empty box over the row titles using analysis variables in denominator defi n ltion s. The new algorithm for interleaving crossed subgroups is most useful for locating subtotals of classification groups immediately following the levels of each group. Using analysis variables in denominator definitions allows better handling of multiple response variables. Improved algorithms for handling percent calculations now allow the presentation of row, column, and data set percentages in the same table |
class variable, response variable, proc step, proc tabulate, user-defined format, label statement | source | 15 |
| 53 | Changes and Enhancements to SAS/ETSSoftware | David M. DeLong, Leigh A. Ihnen, Mark Little |
proc step, residual, missing value, autoregressive model, confidence interval, predictor variable, maximum likelihood estimation, regression analysis, time series analysis, covariance, likelihood function, proc model, monte carlo, variance, dataset, moving average, proc forecast, sample, expected loss, sas library, proc autoreg | source | 12 | |
| 54 | Changes and Enhancements to SAS/FSP Software | Patti McAlister | This paper discusses the changes and enhancements to the SAS/FSP software. Some changes are common to all SAS/FSP procedures, while others affect only a single procedure or a particular screen within a procedure. All the screens you create with the FSCALC and FSEDIT procedures, letters and forms created with PROC FSLETTER, and function key definition screens are stored in catalogs. Catalogs are not SAS data sets, which can be accessed in the same manner as data sets are cu rrently accessed. Catalog management involves understanding how the catalogs are set up and manipulated. A common text editor is now available for FSCALC program screens, PROC FSEDIT screen modification, and FSLETTER forms and letters. In addition, PROC FSEDIT has changed significantly in order to protect screens and to regularly store data entered without having to exit the procedure. PROC FSCALC has changed considerably to make it more user friendly and to enable programmers to set up spreadsheets for novice users |
proc step | source | 5 |
| 55 | Combining SAS/FSP and the New Macro Facility to Create a Full-screen Interface to a SAS Program | Mark J. Holdbrook | Presented here is a technique for creating generalized report writing software which uses SAS/FSP as a user interface. The user, with 1 ittl e or no know1 edge of SAS, can create a report quality listing or table from a SAS dataset by following the instructions provided by a SAS/FSP screen dataset. SAS/FSP is used to create a "controll! dataset (or datasets). The information from this dataset can be passed to macro variables with the CALL SYMPUT function. Additional control information can be obtained from the input dataset which is to be processed and passed to macro variables as well. SAS code can then utilize these macro variables to adapt to var iations in the input dataset and/or final re port format. Such a software system may have several applica- tions. It could be utilized by novice or in- experienced SAS users to meet specific data processing needs, such as generating a report or listing |
macro variable, in=, dataset, call symput, macro facility, laboratory data, variable, proc step | source | 4 |
| 56 | Commonwealth Edison SAS Data Set Facility | John Fitterer | dataset, sas library, macro definition, options statement, macro facility, proc step | source | 5 | |
| 57 | Comparison of Multivariate Regression Analysis between Logistic Model and the Least Square Model Using SAS Software | Jane R. Bratz, J.S. Kim, George K. Tokuhata |
regression analysis, response variable, predictor variable, logistic distribution, logistic regression, relative risk, stepwise selection, multiple regression, p-value | source | 4 | |
| 58 | Comparison of Variance Estimation Methods for Complex Sample Designs Under Extreme Conditions | Edward L. Spitznagel | Modern survey sampling designs frequently involve complex strategies far removed from simple random sampling. Three popular methods for approximating the variances of estimators Taylor series linearization, balanced repeated replications, and jackknife repeated replications - are currently in common use. All three methods have been studied through simulation and found to give approximately the same results. However, conditions involving weighting and cluster sizes have been fairly tame in comparison with extremes sometimes encountered in practice. This paper examines these methods under extreme condit-ions, for the purpose of estimating rates that themselves are extreme, in the one to five percent range. The results, generally, are that all three methods remain nearly identical in terms of bias and standard error, under all single extremes and combinations thereof used to challenge them. They are, found, however, to be inferior in terms of standard error to the jackknife | proc calis, design effect | source | 5 |
| 59 | Computerized Material Loading Schedules: An Aid in Implementing High Density Storage Systems | Darryl W. Kolojaco | data step, lookup, put statement | source | 6 | |
| 60 | Conditional Logistic Regression in SAS Software | Gregg A. Drube, Dale F. Kraemer, Stephanie Van Lier |
Logistic regression models are a commonly used form of analysis for categorical data. For matched designs, however, the usual logistic model becomes quite unwieldy. The conditional logistic model is a computationally more effi cient way of handling such matched designs. Three SAS MACROs are presented that perform conditional logistic regression in a step-up scheme. An epidemiologic case-control study example will be discussed | logistic distribution, logistic regression, regression analysis, likelihood function, conditional logistic regression, macro definition, binary variable | source | 7 |
| 61 | Conference Program Management Using SAS/FSPSoftware | H. Pat Artis | One of the most significant problems facing the organizers of a professional conference is the management of the paper selection and agenda creation processes, Briefly described, the program manager must acknowledge the receipt of abstracts and papers, send various reminder letters, and send letters of acceptance or rejection to each of the contributing authors. In addition, the program chairman must communicate session times, locations, and abstracts to the session chairmen and produce the preliminary and final agendas for the conference. This paper discusses the implementation of a SAS/FSP database application designed to perform these functions. The application maintains file structures for papers, referees/chairmen, and sessions and makes extensive use of FS/LETIER to generate standard letters for all of these individuals |
sas batch, format statement, sample, unique key | source | 6 |
| 62 | Converting CLINFO Data Sets into SAS Data Sets | Philip R. Khoury, Peter Laskarzewski |
The General Clinical Research Center (GCRC) at University Hospital in Cincinnati, Ohio, is one of 7S GCRCs in the country, 28 of which have access to the CLINFO Data Management and Analysis System. In addition, there are another 10 CLINFO sites worldwide. CLINFO was developed specifi cally to aid clinical investigators in a research environment and provides for data entry, manipu lation, and simple analyses. Since CLINFO was being run on a VAX-11/7S0, we decided to augment CLINFO with the base SAS product to aid sta tisticians who undertook more sophisticated ana lyses of existing data bases. We received the test version of the base SAS product under VMS in late 1983 and developed a simple procedure to convert CLINFO files into a form which could easily be used with SAS. The procedure uses a DCL command file and a CLINFO response file to convert CLINFO worksheets into SAS data sets, requires almost no user interaction, and can be easily adapted to other installations | dataset, sas base | source | 5 |
| 63 | Creating a Summary Table of Normalized(Z) Scores | Walter W. Owen Jr. | Data from the behavioral sciences are often analyzed by normalizing the scores for individuals in experimental subgroups to a reference population. Normalized scores, called Z-scores, may then be used to compare performance relative to the | standard deviation, label statement, macro variable, variable | source | 5 |
| 64 | Cumulative Sum (CUSUM) Control Schemes | J.M. Lucas | Cumulative Sum (CUSUM) quality control schemes are becoming widely used in industry because they are powerful, versatile, and easy to use. They cumul ate recent process data to quickly detect out-of-control situations. They also serve as a powerful diagnostic tool. There are now more than 10,000 CUSUM control schemes in use daily ;n Du Pont. This talk describes design and implementation procedures for CUSUM control schemes with emphasis on properties that are recorded as counts. The talk will describe recent de velopments which make CUSUM procedures more useful and more powerful. The recent developments described are: o Fast Initial Response (FIR) CUSUM which gives extra sensitivity to out-of control situations at start up or after a (possibly ineffective) control action. o A Combined Shewhart-CUSUM which com bines the key features of CUSUM schemes and Shewhart Schemes by adding Shewhart Control Limits to a CUSUM scheme | poisson distribution, sample, outlier, quality control | source | 6 |
| 65 | Customized Tables Using the DATA Step | Charles W. Kelly, Richard Roach |
This paper describes a technique that uses the SAS DATA STEP to produce customized tables. The table can have any form l either rectangular or otherwise, and the data displayed within the table can be of any desired type. The program used to generate the table is simple in concept |
if-then-else, proc step | source | 4 |
| 66 | DB2 Simulation Using SAS/ISPF | Doug Korthof | macro variable, order by | source | 3 | |
| 67 | dBASE III and the SAS System | P. L. Olympia | This paper describes a unique implementation of the Federal Government's Small Business Innovation Research (SBIR) Program that takes advantage of the best that micros and mainframes have to offer. The SBIR system. developed around dBASE III. freely exchanges data with other microcomputer software as well as with SAS and SAS/GRAPH on the mainframe. The system's initial data load is carried out from proposers' hardcopy submissions with the aid of an optical character reader. SBIR summary -statistics and graphs are routinely produced by SAS under control of the micro which can emulate a TEK4010 device to receive SAS/GRAPH output. The procedure has many advantages including o eQ..6e 0' U06e - a. U6eJL need not be ,am .. Le..iQJt w.ith ~theJL dBASE III o~ |
source | 3 | |
| 68 | Descriptive Statistics: Using Indicator Variables | Tamara R. Fischell, Juliana M. Ma |
binary variable, proc step, proc means, proc freq, bar chart, nominal variable, class variable | source | 5 | |
| 69 | Designing Interactive Instruction | Janice Sibley | You have been asked to develop a computer based training course (CBT) for your organization. Where do you start? What design principles do you apply to make your CBT course both effective and motivating? Or maybe you are responsible for acquiring computer-based training courses for your company. How do you recognize a CBT course that is welt designed? High quality CBT courses are developed through careful planning and structured design. Structured design promotes CST that is interactive and tailored to an individual's needs. Structured design also helps prevent the production of computerized "page-turners" where pressing the ENTER key is the only required interaction built into a course. This paper provides you with the specific steps to follow in a structured design approach to computer-based training. If you use authoring systems. to develop CBT courses, you will find that the structured design approach works particularly well | source | 4 | |
| 70 | Designing Macro-based Systems | Jeffrey Phillips | This tutorial is a discussion of the SAS~ Macro Facility as a tool to enhance SAS' performance as a programming language. The focus is on using Macro to construct modular SAS-based software systems. Modular structure, conditional execution of SAS code, and the use of Macro variables as software system parameters are the key elements of the tutorial | macro definition, proc step, macro put, macro facility, macro call, macro variable | source | 11 |
| 71 | Developing a Standardized SAS Software Library for Analysis of Numerous Large Scale Clinical Trials | Debra L. Bingham-Barnes, Betty J. Nelson, David E. Tappe |
A library of standardized SAS programs was developed to perform the statistical analysis and data reporting for a large project of clinical trials. Numerous studies with varying designs and large amounts of data needed to be analyzed. Due to regulatory requirements and a strict time schedule, the analyses needed to be performed and reported consistently and quickly. Since several programmers from different projects would be doing the analyses, the software needed to be easily understood and modified to accommodate different study designs. This paper describes the methods and SAS features used in completing this project, emphasizing those that allow for rapid analyses and consistent output. These include the macro language for conditional execution of code and for variable substitution, format libraries, and Full Screen Product. A composite program illustrating several techniques is included | macro variable, dataset, clinical trial, date variable, %if-%then-%else, macro definition, sample, macro facility, proc genmod, proc step | source | 6 |
| 72 | Developing and Using Macro Utility Libraries | Katie A. Hubbell, David S. Septoff |
proc step, proc print, macro put, macro definition, macro variable, %if-%then-%else | source | 8 | |
| 73 | Development of a Data Management System for Long-term Ecological Research | Marvin F. Marozas, Robert A. McLaughlin, William K. Michener |
quality assurance, raw data, data governance | source | 6 | |
| 74 | Development of Graphical Tools for Enhancements to SAS/GRAPH Software | Kikuo Nakazawa, Hiroshi Takaya, Michiya Tomita, Masafumi Yoshida |
This paper describes techniques to edit pictures produced by SAS/GRAPH procedures, employing the various functions of different graphic devices, such as the CAD system, in order to create more attractive pictures. For this purpose an interface system was developed at the Information Processing Center of Kajima Corporation. This system transfers graphic data from one device or the mainframe computer, where SAS is installed, to other devices. As a result, manipulations, such as magnification and reduction, overlapping, coloring with as many as 4,913 colors, shading, use of Japanese characters, and combining pictures produced by different programs are realized. It can be used to produce more intelligible graphs containing more information |
data science, proc princomp, histogram | source | 6 |
| 75 | Editing Questionnaire-Data Skip Patterns with SAS Macros or How SAS Macros Saved My Neck | Susan Bolotin | keyword parameter, %macro parameters, macro definition | source | 6 | |
| 76 | ENHANCEMENTS TO PROC GREPLAY | Morse F. Kalt | proc step, proc greplay, proc gplot | source | 5 | |
| 77 | ENHANCEMENTS TO SAS/GRAPH DEVICE DRIVERS | Morse F. Kalt | source | 3 | ||
| 78 | Enhancements to SAS/OR Software | Marc-David Cohen | source | 12 | ||
| 79 | Enhancing Printout of SAS Data Sets Using Macros | Gary Vair | Techniques for improving printouts of SAS 1 data sets using macros are discussed in this report. Data sets with large fields such as a 200 character field or an array of 200 character fields, II wrap aroundll when using PROC PRINT, making reading of the data very difficult. This paper presents a macro which greatly simplifies the printing of data in column format when used with PUT state ments in a DATA STEP |
macro variable, proc step, character variable, sample, data science, proc print, macro definition, %if-%then-%else | source | 6 |
| 80 | Enhancing SAS Software Skills through a Better Understanding of the SAS Log | Anne M. Asher, Judith H. Mopsik |
sas log, dataset, proc step, input statement, by group processing | source | 9 | |
| 81 | Estimating Autocorrelations and Crosscorrelations where Missing Values are Present: A SAS Macro | Pan-Yu Lai, Kenneth M. Portier |
correlation, sample, missing value, time series analysis, covariance | source | 5 | |
| 82 | Executing SAS Utilities from a Menu-Driven System | Janet C. Lind | proc step, variable, dataset, sas library | source | 6 | |
| 83 | Fast Reporter | Mark Freeman | proc step, numeric variable | source | 5 | |
| 84 | Fitting Log-linear Models Using SAS Software | Amy Ko, W. J. Shih |
proc freq, likelihood function, sample | source | 5 | |
| 85 | Forecasting and Simulation Using SAS/ETS Software | Mark Little | source | 10 | ||
| 86 | Forecasting Residential, Commercial, and Industrial Gas Demand | Jeff Brand, Jim Funk |
This paper shows how the Wisconsin Gas Company used SAS@. SAS/ETSU. SAS/GRAPH@. and SAS/FSP@ to develop a gas sales forecasting system. It emphasizes which SAS Institute Inc. software PROC's were used to complete each model development stage. In particular, the Insti tute's PROC's that were used to build and evaluate the forecasting system are outlined. Although this paper addresses the process for building and evaluating an econometric model, the process can be adapted to most forecasting techniques |
proc step, demand forecasting | source | 6 |
| 87 | FSCONTENTS-Full Screen Editing for PROC CONTENTS | Gary Vair | The need for a user tool to create and maintain SAS 1 data sets has led to the development of FSCONTENTS a full screen editor for PROC CON TENTS. Th';s tool has been implemented on VM a~d TSO using PROC FSEDIT and SAS macros. FSEDIT,s the SAS Full Screen Product to edit the contents of a SAS data set. However FSEDIT cannot change the variable attributes associated with the data set or create a new SAS data set. FSCONTENTS allows all SAS data set at tri bute s listed by PROC CONTENTS to be edited in full screen mode. New SAS data sets may be defined. Data set changes allowed include: Create a new SAS data set Delete an existing variable Add a new variable Rename a variable Convert a character variable to numeric and vice versa Change the length of a variable Change, add or delete formats, informats, and 1 abe 1 s FSCONTENTS has been designed for those who may not be familiar with the techniques necessary to cre ate or change SAS data sets |
dataset, proc step, proc contents, character variable, sas programmer | source | 4 |
| 88 | Functions and Subroutines Not Documented in the SAS User's Guide | Robert A. Frank | Usually, a programmer or computer user will make use of whatever tools are available, and likely will not make much effort to use tools whose existence or location of documentation is unknown. There are a number of routines (functions and subroutines) available in the SASR System that are not documented in the SAS User's Guide and whose documentation is scattered- elsewhere. This paper gives the programmer or user a concise list of these routines, thus making their use much more likely. Most of the functions and subroutines available to writers of DATA and PROC MATRIX steps are documented in the two volumes of the SAS User's Guide, 1982 edition. However, a variety of other functions and subroutines are documented only in other user's guides (e.g., SAS/ETSTM), techni~al reports, or issues of SAS Communications |
proc step, covariance | source | 6 |
| 89 | Generating Standardized Reports for Preclinical Pathology Research using Base SAS Software and CMS EXEC2 | Rong Q. Pan | The Preclinical Biostatistics Department at Ortho Pharmaceutical Corporation provides statistical analyses for the Drug Safety Evaluation Division (DSE) for Investigational New Drug submissions. The statistical analyses include: individual animal listings, mean value tables and parametric/non parametric methods. This paper discusses a system built to generate a series of mean value tables for DSE studies. This system utilizes Base SAS Software, CMS EXEC 2 and the Display Management System. The techn; ques used to build this system can also be used to generate any standardized tables for different fields of application such as marketing research and operations research | sas base, by group processing, dataset, proc step | source | 7 |
| 90 | Help SAS System Users with an RBBS | P. L. Olympia | Faced with an ever Increasing number of SAS system users who have discovered the power of microcomputing. a few organizations have started in-house Remote Bulletin Board Systems (RBBS). An RBBS gives organizations the facility for delivering timely information and assistance to users via the electronic messaging subsystem and the files subsystem. The RBBS Doors feature allows an authorized user to run a program remotely and is an excellent vehicle for diagnosing user problems with software. This paper describes our experiences in operating the national SUGI SIG/M RBBS and tips on starting and implementing your own RBBS | batch processing, data science | source | 4 |
| 91 | Implementing IMSL Routines as SAS Functions | Michael Conlon | This paper describes the JCL and FORTRAN program ming necessary to create an interface routine which can be called from SAS as a SAS function, which, in turn, calls an IMSL routine to perform an operation. The interface routine then returns the value calculated by the IMSL routine to the SAS user | sas function, probability distribution, binomial distribution | source | 2 |
| 92 | Individualized Reporting using PROC CHART | Alissa S. Bernholc | proc step, proc sort | source | 4 | |
| 93 | Information Center Using SAS Software-A Success Story | Henry B. Edwards | dataset, macro facility | source | 4 | |
| 94 | Interactive SAS Program Generators for Analyzing and Reporting Medical Research | Sandra B. Brizendine, James T. Crowe Jr. |
dataset, proc step, sample | source | 4 | |
| 95 | Interfacing IBM Microcomputers and SASSoftware | E. Jeffrey Hutchinson, Karin P. Hutchinson |
What happens when you mix 10% IBM PC, 30% IBM PCjXT, 10% IBM PC AT, 40% IBM 3270 PC and 10% COMPAQ PLUS portable computers with SAS Software? If done properly, you will obtain the most powerful brew currently available. As a case study this paper will cover the installation of over 100 IBM or compatible microcomputers in a corporate environment. Topics to be covered include: 1). Hardware and software configurations 2). Communications between the mainframe and the microcomputers 3). Problem areas to watch out for 4). A review of a successful SAS Software application. The microcomputer hardware that will be covered includes the IBM PC/XT, IBM PC AT, IBM 3270 PC, and the COMPAQ PLUS | source | 6 | |
| 96 | MACRO LGTREG for Logistic Regression and Associated Diagnostics | Thomas F. Devlin, Barbara J. Weeks |
Logistic regression is :'l statistical tool for describing, controlling or predicting pro portions or rates. It has found application in such diverse fields as medicine, marketing, toxicology and industrial processes. Devlin (1984) demonstrated the use of the MATRIX procedure for obtaining maximum likelihood fits and the associated diagnostics proposed by Pregibon (1981) for logistic regression models. This paper describes the SAS statement macro LGTREG which provides a more convenient, user oriented implementation of those techniques | logistic regression, dataset, regression analysis, logistic distribution, binomial distribution, likelihood function, maximum likelihood estimation, macro definition | source | 6 |
| 97 | Macro Power: A Procedure to Generate Wilcox on Rank Sum Power Curves | Michael J. Todd | %MACRO POWER generates power curves for the Wilcoxon Rank Sum test using a method described by Lehmann (1975). %POWER performs both a priori and post-hoc power calculations. The user may define a range of deltas and the alpha level or use the defaults provided by %PQHER. An output data set of the results and formatted output are also aVailable |
standard deviation, sample size, power analysis, sample, response variable, macro definition | source | 4 |
| 98 | Macro Quoting Functions | Jerry Mock | source | 7 | ||
| 99 | Macros for Paired Categorical or Continuous Scale Data Analysis | Thomas E. Bradstreet | sample, sample size, missing value, proc freq, data science, rank test | source | 8 | |
| 100 | Macros that Emulate Macro Functions | Warren F. Kuhfeld | Programming in the SAS macro language, like other computer languages, is more efficient after' a library of utilities that per'for'm commonly done tasks is compiled. The macro language does not explicitly give the user the ability to write library subroutines or functions, however it is possible to wr'ite macl'Os that can subsequently be used in other macros as if they wer'e functions or subroutines. Such macros should contain no generic SAS code so that they may be "called" from anywhere. This poster will illustt'ate this capability with six macros that can be °oINCLUDE'ed and then used as if they were macro functions. These "functions" check their arguments, which will typically be the parameters of the calling macro, to see if their value is a valid SAS name, SAS variable list, dataset name, integer, signed integer, or floating point constant, possibly expressed in "E" format. The functions resolve to 1 (true) if the ar'gument is valid and 0 (false) otherwise | macro facility, macro function, macro variable, %if-%then-%else, decimal, variable list, macro definition | source | 6 |
| 101 | MADAM: An Example of Hierarchical Data Set Management | Richard L. Gimarc | dataset, length statement | source | 5 | |
| 102 | Maintaining Quality Control over Input Data with SAS Software | R.A. Whitfield | proc step, quality control, proc print, sas programmer, quality assurance, macro variable | source | 3 | |
| 103 | Managing the Human Resource with the SAS System | Cory Scott | Purchased or in-house designed human resource systems seldom have the flexibility to adequate ly manage the human resource from a data stand point. However, SAS, with its report, statis tical, and database capabilities can aid the human resource professional in managing the resource with ease, flexibtlity and security. This paper demonstrates ways that SAS can tx! used to develop a user friendly database structured for the human resource function. In the field of human resources, personnel information is normally obtained from one of three systems: a software package from a human resource information system vendor, a personnel system designed by the company data-processing department, or from a large bank of file cabinets in the personnel clerk's office. Personnel data systems are the lifeblood of the human resource environment | dataset, data wrangling | source | 11 |
| 104 | Material Management with SAS Software | Sy Aslan, Ethan Matlaw |
The paper describes the application of SAS software In an inventory control system at Montefiore Medical Center. an 1825 acute and long term care beds institution. In the past, batch segments of an accounting software program were utilized to control its 2500 inventory items and proved inadequate in providing the management information necessary to control the inventory operation effectively and efficiently. SAS software was introduced by the Department of Management Science to provide the analysis and control data necessary to properly maintain inventory levels, improve the physical layout, and forecast and plan for the future | management reporting, sample | source | 6 |
| 105 | Menu-driven Reporting | James M. Watts | dataset, macro variable, batch job | source | 8 | |
| 106 | Methods of Generating Model Estimates in the PRINQUAL MACRO | Warren F. Kuhfeld, Warren S. Sarle, Forrest W. Young |
The PRINQUAL (PRINcipal components of QUAlitative data) algorithm implemented in the PRINQUAL SAS macro generates model estimates for rescaling variables in one of two ways. The maximum total variance method maximizes the variance accounted for by a stated number of components. The minimum generalized variance | proc princomp, covariance | source | 10 |
| 107 | Modules Versus Segments: A Performance Comparison of the SAS System under the CMS Operating System | Simon Griffiths | The YM/CMS SAS system can be installed as modules or as saved segments. Use of segments | proc step, cpu time, proc genmod | source | 6 |
| 108 | More Efficient Data Storage when PROC TRANSPOSE is Used for Displaying the Output | Lynn T. Julian, Donald D.M. Tong |
proc step, proc transpose, time variable, _eof_, missing value | source | 6 | |
| 109 | Multidimensional Scaling Models and Methods for Two-way and Three-way Proximity and Preference Data | J. Douglas Carroll | euclidean distance, proc princomp | source | ||
| 110 | Multivariate Data Analysis Using the SAS System | E. James Harner | The uses of CAND1SC, D1SCRIM, STEPDJ5C, and other multivariate SAS S Procedures are illustrared using a bird habitat data set. The objective of the ana!ysis is to discriminate among four sparrow species basl2d on l2ight habitat variables. Emphasis is given to assessing the assumptions of the discriminant model. DISCR1M and PRINCOMP are used to Bl!amine the assumption of equality of covariance matrices. Collinearity is evaluated by PR1NCOMP and STEPDISC. Outliers are determined from the Mahalanobis distances, which are computed from the output of D!5CRIM. Normal ity is thl2n assessed by constructing gamma prObability plots from these Mahalanobis distances. Both weighted (based on Mahalanobis distances) and unweighted canonical discriminant analyses are performed using CANDISC. The important variables are "selected | covariance, outlier, discriminant analysis, correlation, mahalanobis distance, proc step, sample, chi-square distribution, scatter plot, proc cancorr, proc princomp | source | 6 |
| 111 | New Developments in Describing and Visualizing Multivariate Data | Forrest W. Young | A software development project is currently in progress at the Univer"sity of Nor'th Car"olina at Chapel Hill (UNC). This project has two major pur'poses: 1) The development of SAS procedures to perform a var'iety of popular descriptive multivariate data analyses. 2) The development of a "flight-simulator"-like SAS procedur'e to enable visual exploration of multivariate data. This paper' describes the design of sever'al of the pr'ocedur"es being developed. Disclaimers These procedures are being developed at UNC, not by SAS Institute. There is no commitment by the Institute or UNC to support nor to distribute these procedures. This paper repor'ts on the potential design of several procedures. Some of these procedures are in prototype form, some are in the design stage. No promise is made that these procedures will be developed, and, if they are, that they will be in accord with the designs described herein | data science, proc step | source | 9 |
| 112 | Non parametric Sequential Trend Test | Young C. Park | Experiments using a series of doses of a compound, including a zero dose control, are often used to assess the effect of a test compound. A nonparametric sequential regression analysis, which is an extension of the NOSTASOT procedure (Tukey, et al 1983, 1985), is proposed to assess trends in the response variable. The proposed trend test handles progressiveness of response with increasing doses, providing more selectivity and sensitivity than some alternative procedures, with proper a error and distribution free characteristics. A program of SAS PROC GLM and BMDPIV is presented using a numerical example |
regression analysis, anova, raw data, response variable | source | 5 |
| 113 | Obtaining P-values using PROC PRINTTO and PROC TABULATE | Michael M. Brown, Karl Konnerth |
SAS procedures can produce copi ous amounts of data. A problem of too much data can develop when information from these data are extracted by hand for a report. Obtaining p-va1ues produced by PROC T -TEST on many treatments and variables is an example of this problem Thus, it is necessary to find programming techniques that will extract this information. By combining PROC PRINTTO with SAS DATA STEP programming and PROC TABULATE, obtaining information such as p-va1ues will require less effort. By fi rst sendi ng the resu1 ts of T -Test to a dataset on di sk and then by readi ng in the data, using an INFILE statement, the programmer can search the data using various programming statements in order to find the desired values. Once these values have been found, output to the current dataset, and printed for verification, PROC TABULATE can then be used to create a tab1 e of p-va1 ues, thus saving hours of tedious hand work |
proc step, proc printto, proc tabulate, data step | source | 3 |
| 114 | On Migrating a SAS-based Data Base to a Production System | Thomas G. Kryzak | A non - EDP unit of the New York City Health and Hospital Corporation developed prototype SAS Prcx.::rrams to create a consolidated Clinical/Financial data base for 11 acute care hospitals. A review of external vendors' data base systems indicated that in-house developrent 'WOuld produce a systEm with greater functionality at less cost. It then \vas necessarv to convert the prototype progranh~ into~ a production system to be run on a regular basis by the EDP Department. This paper discusses the characteristics of the data, the reasons for creating a production system and the considerations involved in migrating to a full production environment: The reasons for creating a production system involved: 1) size of the data base 2) Canputer resource requirerrents 3) The local processing environment 4) System continuity 5) system integrity The implications of so doing were: 1) Documentation 2) Control Reports and Data Edits 3) Program Staging 4) Operator Instructions 5) Testing | raw data, batch job | source | 3 |
| 115 | Optimal Allocation for the Comparisons of Two Proportions | David C. Huang | sample, sample size, standard deviation | source | 6 | |
| 116 | Optimization Techniques for SAS Applications | Vernon E. Kirchman, Kirk Paul Lafler |
The versatility of SAS* as a programmer/end-user tool is generally well known and accepted by those acquainted with its capabilities. An important element that is often overlooked or may not even be considered when using SAS ;s the need of following a set of optimization guide- lines. It is often found that the simplest of requests can fall prey to violations of available resources, much of it due to the fact that little or no planning occurs prior to program execution. Examples of typical abuses as well as guidelines for their prevention will be presented to assist the programmer/end-user streamline applications | keep statement, length statement, proc append, proc step | source | 3 |
| 117 | Optimizing Storage of SAS Data Sets | Carol A. Bogardus, Martin J. Rosenberg |
By default, the SAS system stores numeric vari ab1es ina SAS data set in 8-byte representation. The LENGTH statement can be used to override the default and store numeric variables in a smaller space, thereby reducing storage costs. While this may be safely performed with integer valued variables, non-integer val ues wi 11 frequently be truncated leading to serious computational or logical errors. This paper presents an easy to use MACRO for identifying which variables in SAS data set are integer valued, allowing safe usage of the LENGTH statement | numeric variable, dataset, proc step, length statement, flag variable | source | 4 |
| 118 | Performance Analysis of Superheater Data from an Operating Nuclear Reactor | Kenny C. Gross, B.R. Seidel |
A detailed investigation was made of historical data recorded at EBR-II during operation with superheater SU-712. The objecti ve of thi 5 study was to analyze and characterize the performance of 72 duplex steam tubes that became unbonded during a long period of operation. The information processing system ANALYZE, using SAS as its foundat i on software, was de ve loped to perform the required numerical manipulations, statistical analyses, and spectral analyses with an inpdut data set containing some five million data values. The ANALYZE system was successfully employed (1) to characterize the performance of all the steam tubes in terms of frequency and relative severity of unbonding~ and (2) to establish a correlation between the observed anomalous behavior of the superheater and its operating parameters. Results from this in vestigation were used to guide materials examinations and physical tests that were performed after SU-712 was removed from operation | proc step, correlation, master data management, proc freq, regression analysis | source | 6 |
| 119 | Performance Tracking in a Complex Environment | Gordon R. Stauffer | Performance tracking techniques have evolved considerably over the past several years with the introduction of new technologies and methodologies. Improvements in data acquisition, integrated data base systems. powerful reporting procedures, and color graphics have provided significant advances in the productivity of performance analysis. The evolution of these tools has allowed analysts to redirect their focus from data collection/reduction to interpretation. New methods continue to be developed which further enhance analysis of performance variables and improve the surveillance of workload and service levels. An overview of the evolution of tracking methods, a current, successful approach (which was developed at a data center where the author managed performance services), and ideas for future consideration will be discussed | capacity planning, building block approach | source | 4 |
| 120 | Practical Capacity Planning at the Automobile Club of Southern California | Jason G. Shane | At the Automobile Club of Southern California, a fonnal Data Processing Planning Department was established in March, 1982 to address areas such as Capacity Planning and Performance Management. This paper discusses the approach, tools and techniques used to set up the CPE function at the Club. Emphasis is placed on a practical | capacity planning, regression analysis | source | 5 |
| 121 | Predicting Arkansas Tornadoes with SAS/GRAPH Software | Craig A. Brown, Daniel W. Puckett, Ken Schriner, Richard M. Smith |
proc freq, dataset | source | 4 | |
| 122 | PROC ELPRINT: a Print Procedure that Supports Footnotes and Outlier Detection | George F. Fraction, Jeffrey B. Shotts, Richard E. Skrzynecki, S. Stanley Young |
PRoe ELPRINT is similar to PROC PRINT. Individual values can be footnoted with user defined footnotes. Two data sets are read - a print data set and a footnote dictionary. The symbol of a marked variable value is matched with its definition and the footnote is printed on the bottom of the print page. If requested, the procedure will mark statistical outliers - row or column outliers or interaction outliers (Bradu-Hawkins, 1982). The outlier detection methods are robust to presence of multiple outliers |
outlier, proc step, dataset, proc print, standard deviation, numeric functions | source | 6 |
| 123 | PROC GMAPas a Productivity Aid in Semiconductor Device Manufacturing | Gregory J. Hendrix, Michael J. Maslack |
The art of semiconductor device and process char acterization has become an increasingly difficult task for many reasons. One of the most pervasive of these is the manipulation and interpretation of the massive quantity of data that can be generated with today t s highly automated and sophisticated testing equipment. Until now, the summarization of this data into logical subgroups (i.e., device sites or wafers) and subsequent statistical analy sis has been both difficult to do, and even more difficult, if not impossible, to portray to oth ers. This poster presentation is intended to il lustrate the use of the SAS/GRAPH GHAP procedure in the context of data analysis and portrayal. As such, it will include both SAS code listings for the various graphic outputs and analyses as well as examples of possible graphic outputs |
proc step, proc gmap | source | 6 |
| 124 | PROC PECAN - A SAS Procedure | Sivaram Kalyandrug, Jay H. Lubin |
The PECAN procedure fits a multiplicative relative risk regression model to l-R matched case-control data using an exact conditional likehood, where R is the number of controls and may vary for each matched set of individuals. The procedure outputs maximum likelihood parameter estimates, their standardized values p-values. the asymptotic covariance matrix and the maximized log-likelihood. Up to six different models can be s_pecified involving sets of covariates. The procedure also outputs a chi-square based on a score statistic to test the contribution of additional covariates in nested models. To test the influence of each set on the overall estimation, an option has been included in the procedure that estimates the parameters when each set is singly excluded from the data. The parameter estimates are obtained using an iterative NewtonwRaphson | estimate, by group processing, likelihood function, degrees of freedom, proc step, relative risk, dataset, confidence interval, null hypothesis, regression analysis, logistic distribution, chi-square distribution, class variable, out=, covariance, logistic regression | source | 8 |
| 125 | PROC TABULATE Applications | Andrew A. Norton | In some applications, however, PROe TABULATE cannot compute all of the statistics that we would like to present in a PROC TABULATE table. This paper will describe techniques for using PROC TABULATE to format and report statistical results obtained from other sources: other procedures, DATA steps, other statistical packages and programs, published data, and so on. In effect, the statistical computation phase of PROC TABULATE is bypassed. The reporting facilities of the procedure are thereby made available for the presentation of data from any desired source. This capability substantially enlarges the realm of situations in which the flexible and powerful table formatting capabilities of PROC TABULATE can be used. Potentia] Applications The techniques described in this paper can be used to incorporate statistics such as medians and other percentiles into PROC TABULATE tables. They can also be used when PROC TABULATE supports the statistic, but not the desired weighting method |
proc tabulate, proc step, class variable, proc univariate, proc sort, do over, proc means, macro definition | source | 7 |
| 126 | Program Planning Using the GMAP Procedure in SAS/GRAPH Software | Michael P. Farrell, T. J. Gross, Tommy R. Nelson, T. R. White |
A methodology that produces Program Planning Charts using SAS/GRAPH software has been developed as an aid to program management in the CO2 Research Program of the U. S. Department of Energy. The methodology uses programs from MAPPER, a front-end program (written by D. A. Dahl) for DISSPLA to produce Logic Flow diagrams. Each research question, or set of questions, which comprise one phase of a research program, is represented by a boxed area containing the title of the particular research question(s). The GMAP procedure of SAS/GRAPH software reads user-supplied information regarding a variable aspect of each research question and shades the boxes accordingly. The |
proc gmap | source | 6 |
| 127 | Progress of the Portable Systems Division-Panel Discussion | Richard A. Usanis | sample, dataset | source | 3 | |
| 128 | Projecting Population Using SAS Software | Larry E. Robinson | The Virginia Department of Planning and Budget 1s leglslatlvely mandated to develop short and long-range population projections for use by other state agencies and the legislature for programs that involve or necessitate population projections. However, other clients include local and regional governments, universities, and business as well as private citizens. Prlor to 1983, thls task was done prlmarlly by consultants with our agency providing the current data, making model assumption decisions. and adjusting the mechanical projections. In 1984. the decision was made to develop the model in-house and SAS* soft ware was chosen as the best computer language to use. Using SAS software. we developed a model for roughly what the consultant was charging for one update cycle using his model. The marglnal cost for runnlng the model 1s low, allowing us to test many additional assumptions that we wouldn't have otherwise because of prohlbltlve costs. Several indlrect beneflts occurred | proc step, dataset | source | 3 |
| 129 | Prototyping Large Systems Using the SAS System | Stephen C. Martin, Robert M. Moreau |
dataset, macro definition, proc step | source | 3 | |
| 130 | PROVAL-An Automated Method for Evaluating Trade and Consumer Promotions for Package Goods Industries | Ambar G. Rao, Ross Dennis, Douglas Grady |
regression analysis, moving average, outlier | source | 5 | |
| 131 | RAGS: An Interactive SAS, IDMS, and FORTRAN Tool for Regression Analysis | Brent L. Peterson | Using the best features of many different pieces of software, RAGS (Regression Analysis and Graphics System) forms a powerful, yet user friendly tool for use in freight tariff negotia tions. In this paper, the motivations for RAGS' development and design will be discussed, along with examples of how RAGS is used in tariff negotiations with carriers who trans port Air Products and Chemicals' products. The paper will be concluded with a discussion of how and why such diverse software technologies were integrated, and the savings to the company by using RAGS. Highlights: RAGS provides rapid access to SAST~ regression and graph ics software, and through the use of ISPF'~ menus, requires no knowledge of the SAST~ commands needed to execute the SASlM procedures successfully. RAGS utilizes an IDMST~ (A Cullinet Corp. product) database to efficiently store regression model data. Data is transferred between SAST~ and IDMS'" via FORTRAN and physical se quential files | regression analysis, flat file, raw data, proc reg, proc step | source | 5 |
| 132 | Reading Unbalanced, Unstructured Data with SAS Software | Ronald P. Cody | proc print, proc step | source | 3 | |
| 133 | Regression and Diagnostics | Gerald R. Hobbs | In this article we assume a familiarity with the SAS software package and with the rudiments of least squares linear regression analysis. In Section I we review some of the relevant literature. The default output from PROC REG is explained briefly in Section II and newly developed techniques of model criticism and case analysis are discussed in the context of options to the proceedure PROC REG in the following tt1ree sections. Both real and fabricated data are used in examples to illustrate the behavior of certain statistics and diagnostics in various circumstances. The focus) then) is on explaining the regression di8gnostics which are produced as options under PROC REG |
regression analysis, proc reg, predictor variable, proc step, proc calis, correlation, cook's distance | source | 8 |
| 134 | Repeated Measures in Nonlinear Models | Ronald W. Helms, Keith E. Muller |
Inherently nonlinearizable models arise nat urally in phamacokinetics. econometrics and many other applications. If more than one measure ment of a variable is made on each subject. then some account must be taken of correlation be tween the repeated measures. Possible methods can be characterized by their assumption about the error covariance structure. Estimation presents few problems for any existing method. However, hypothesis testing. particularly in small to moderate sample sizes, is problematic for most methods. Suggestions are made to help the analyst select a method appropriate for the particular nature of the model and the covari ance structure | proc nlin, covariance, regression analysis, repeated measures, weighted least squares, likelihood function, hypothesis testing, correlation, sample, proc calis, response variable, proc reg | source | 5 |
| 135 | Report Writing: A Case Study | Tim Lehman, Ann Lehman |
proc step, proc tabulate | source | 4 | |
| 136 | Run Time Component Analysis Using SAS Software | H. Pat Artis | One of the most fundamental techniques employed in computer performance evaluation (CPE) studies is the decomposition of process execution times into their component parts. In this paper, we will examine how MVS/XA SMF and RMF data can be used to decompose the execution time of batch jobs. 1.0 Introduction One of the most common problems presented to the performance analyst is a process execution time that does not meet the expectation (i.e., service objective) of some user. When presented with such a problem, the most logical first step is to decompose the execution time into its component parts. Once this process has been completed, the analyst can identify specific service times or server delays that can be improved to achieve the desired service level. Prior to MVS/XA, SMF and RMF data did not provide sufficient information to effectively decompose the execution time of batch programs | elapsed time, format statement, cpu time | source | 3 |
| 137 | SAS Data Set Containing SAS File Directory via User-written Procedure | Mark T. Duffin | proc step, proc print | source | 2 | |
| 138 | SAS Graphic Macros for Producing 2-Dimensional and 3-Dimensional Step-graphs | Akos Felsovalyi | macro definition, data step | source | 5 | |
| 139 | SAS Macro Techniques for the Development of Integrated On-line Systems | Roger B. Glaser, Clark M. Roberts, James H. Sorenson, John S. Willard |
The "Information Age" is placing ever increas ing demands on information processing profes sionals to produce more and more automated systems. Management not only expects systems to be developed faster, but also requires them to be easy to use, ,flexible enough to adapt rapidly to change and to execute efficiently. Fortunately, there are more development alternatives available today to meet these challenges than in the past. Personal compu ters and 4th generation mainframe languages can be used directly by those requesting the system when requirements are not complicated. Very large and complex systems, or systems where the volume of transactions is large, still require the use of traditional data processing approaches. But what about the vast spectrum between these extremes? The majority of organizations categorize systems in this middle-ground as complex and apply slower traditional methods to develop them, increasing their already large backlogs. We offer a compromise | macro definition, macro facility, dataset, temporary dataset | source | 10 |
| 140 | SAS Macros for Bootstrapping and Cross Validating Regression Equations | Richard T. Carson | bootstrap, data cleaning, sample, regression analysis, proc calis, proc step | source | 6 | |
| 141 | SAS Macros for Designing Hierarchical Data Files | Robert Landon | SAS has the capability of efficiently handling a wide variety of complex system data files. When creating files for input or output in SAS, the programmer can employ macro coding techniques to design hierarchical file structures. These are especially useful when data is to be stored for subsequent use in a SAS library. When this method is used to input data from tape files stored in COBOL 1 ibrary formats (e.g., COBOL copylibs), it greatly enhances the ease of determining data positions, counting table occurrences, handling related data in macro ublocks", and preserving data integrity between files. To effectively create SAS macros for data operations, the programmer groups lower level data for input, selects a meaningful macro name for the group, and includes documentation of data elements within each macro group by using comments. A macro or a combination of macros is used for selecting data for program operations | sas library, dataset, macro definition | source | 4 |
| 142 | SAS Matrix Algorithms for the Recovery of Interblock Information in Lattice and Incomplete Block Designs | Paul L. Cornelius | covariance, macro definition | source | 3 | |
| 143 | SAS Menu-driven System using SAS 82.4 Macros under IBM MVS/TSO | Mark T. Duffin, Gustave F. Jonas |
dataset, batch job, proc step, macro definition, macro variable, sample, %if-%then-%else | source | 6 | |
| 144 | SAS Methods for Obtaining the Exact Hypotheses from Type 1, 11, III, and IV Estimable Functions in Terms of Cell Means Model | Richard Lustick, P. K. Tandon |
proc genmod, response variable, main effect, proc glm, macro definition | source | 8 | |
| 145 | SAS Procedure for Optimal Presentation of Cluster Profiles | Akos Felsovalyi | The presentation of cluster profiles is a graph where each cluster corresponds to a line (its profile). The variables of the cluster analysis form the x-axis, the cluster means form the y axis and the means of a cluster are joined by a straight line (Figure 6). The presentation is optimal if the order of the variables is so chosen that the number of intersections of the cluster profiles is minimized. The optimal presentation yields better inter pretation of the clusters I interrelationships. This paper presents the CLPF SAS procedure (CLuster ProFile) that produces the optimal presentation of the cluster profiles | cluster analysis, proc step, proc cluster | source | 6 |
| 146 | SAS QUOTE: SAS Software With Costing/Pricing Systems | Sam Morris | SASQUOTE is a set of programs and procedures developed at E-SYSTEMS, Eel Division, -using SAS to access the Division costing/pricing database and process this data to meet many internal and customer decision support and analytical requirements. It makes extensive use of the SAS MACRO Language, looping capabilities, PROC TRANSPOSE, and SAS/GRAPH. The potential for growth and additional applications is Vast |
do loop, proc step, macro definition, dataset, macro facility, proc transpose | source | 6 |
| 147 | SAS Software and the Chi-square: Collapsing Cells with Less Than Five Expected Observations | Elizabeth Zobre | When testing the goodness-of-fit of actual data against a hypothesized distribution fWlction, the chi-square test statistic requires each expected cell frequency to be greater than or equal to five observations (2). Often this requirement creates a problem when there are few observations and/or nany values of the BY variables. This paper illustrates a SI\S algorithm designed to satisfy this requirement. Intrrxlngtion Often in industry, a need ar ises for testing actual data against a hypothetical distribution function. The exanples of the paper are related to queueing models which assume the arrival rate follows a POisson distribution and the service time follows an exponential distribution. In order to test these underlying assunptions, the nurrber of units arriving in the * system" I111lSt be counted and the length of time each unit stays in the system IRlst be IOOasured | proc freq, p-value | source | 5 |
| 148 | SAS Software for Log-linear Models | Peter B. Imrey | response variable, sample | source | 14 | |
| 149 | SAS Software for Stratified Analysis | William M. Stanish | · This paper illustrates the usage of the FREQ procedure for stratified categorical data analysis. The procedure gives one the capability of exploring the relationship between two variables after controlling for any number of stratification variables. Statistics are printed for each stratum, as well as for the overall summary. For general r by c contingency tables, three types of Cochran-Mantel-Haenszel statistics are computed to, test for partial association. Also computed for 2 by 2 tables are Woolf-Haldane and Mantel Haenszel estimates of the common odds ratio and risk ratio, together with 95% confidence limits and the Breslow-Day test for homogeneity of the odds ratio. Examples are given |
relative risk, alternative hypothesis, correlation, sample size, sample, proc freq, confidence interval, odds ratio, null hypothesis, response variable, anova | source | 6 |
| 150 | SAS Software Introduction from a Menu-driven System | Thomas F. Harwood | bar chart, vertical bar chart | source | 6 | |
| 151 | SAS Software Made User Friendly | Russell G. Hackett Jr. | data dictionary, proc step | source | 6 | |
| 152 | SAS Techniques for Reading Variable Length Record Formats | Andrew V. Bowden Jr. | length statement, trailing @, input statement | source | 6 | |
| 153 | SAS-based Data Management System for Clinical Drug Trials | Jean A. Bitney | lookup, macro definition | source | 5 | |
| 154 | SAS-based Information Systems for Start-Up and Operation of Nuclear Power Plants | Jacob C. Epperson | Nuclear energy is considered the most efficient source for producing electricity. Once in service, a nuclear power plant can operate at a much lower cost than other types of power plants because of lower fuel costs. This "fuel differential" translates to millions of dollars in savings to consumers. To realize these savings, however, consumers may have to wait as long as 10 years from the time develop ment is initiated. Construction. start-up and operation of a nuclear facility are closely monitored by the utility and regulatory agencies to assure that quality is constantly maintained. The information requirements generated can become unmanageable without automation. Due to the dynamic nature of this environment, the development of maj or information systems with traditional data processing methods has been slow. leaving a noticeable gap in fulfilling the overall requirements | quality assurance, audit trail, data wrangling, past due, macro definition, macro facility | source | 10 |
| 155 | Screen Creation - Automatic Generation of SAS Data Entry Screens Without a SAS Data Set as Input | Berton Shankman | This is a ftmction which generates SAS data entry screens autanatical1y. 'These screens may be used singly or in Dll~tiples to establish screen based data processIng applications. '!he only requirerrent for its use is knowledge of the attributes of the data fields on each observation. It does not require prior existence of a data dataset or screen dataset to generate a completely new screen. Input to this function is facilitated through a series of screens. The function generates the SAS code required. 'Ihis code is executed in on-line IIDde. The output is a SAS screen dataset, which will be saved by SAS if this has been specified in tile function. An interaction of IBM's Dialog Manager (ISPF) and SAS/FSP is used to . accomplish this task. No knowledge of e~ther is needed. However. in its current envirornnent at the Federal Reserve Board this function is looked upon as a prograrrner / analyst productivity tool and has never been used by others |
dataset, data wrangling | source | 5 |
| 156 | Selecting Media for Self-paced Instruction | Julie M. Brown, Trudy D. Bruce |
When selecting courseware for self-paced instruction it is essential for the reviewer to have a standard for evaluation. This paper presents a criteria for evaluating courseware which is derived from the principles of instruction used for development. These principles are based on the critical conditions and events that influence learning. The principles include prerequisities for learning, modes of presentation, and feedback and reinforcement. Examples from both video and computer-based training will be used to illustrate the concepts discussed. I nvalid assumptions regarding learning will be identified and "put to rest." This presentation will better equip those persons who make the decisions about end-user training by providing them with the developer's viewpoint | problem solving, likelihood function | source | 8 |
| 157 | Selecting Training for End Users | Bill Marriott, Jean Ussery |
No longer the private domain of data processing professionals, the computer is being used with increasing frequency by other professionals in their day-to-day jobs. These end users require training that is targeted toward their special needs and applications. Too often, the end user receives training designed for the DP professional. After all, if the course effectively trained our programmers, it should train the users, right? Rising demand for computer services and a steady increase in the creation of information centers encourages end users to do more of their own data processing work. These changes within organizations have made end-user training an area of concern for many training coordinators. This paper defines and explores the special training needs of the end user. Suggestions are made to help training coordinators evaluate and select courses for end-user training. You might also incorporate these ideas into training you design | data wrangling, input statement | source | 5 |
| 158 | Selection of Variables for Identification of Transfer Function Models on SAS Software | James A. Deddens, Wang Ping |
correlation, time series analysis, covariance, sample | source | 6 | |
| 159 | Sharing Application Development Between I/Sand the SAS User | Al C. Ketchum, Diane P. Moshier, Gordon W. Slater |
Sharing Application Development Between liS and the SAS User A method is proposed for sharing Application Develop ment responsibility between Information Systems personnel and the End User community, via SAS. This approach is currently being undertaken in Owego, NY, with development of an Equipment/Property Control System for the IBM Federal Systems Division (FSD). The programming software base is CICSjVS (Customer Informa tion Control System/Virtual Storage) under OSjMVS. SAS is the major report processing interface. The system is being installed in two releases. Release 1 was installed June 8, 1984; release 2 is scheduled for June 28,1985 (see Figure 1). Sharing Application Design Information Systems Personnel End User Community Where IBM Owego, N.Y. (Federal Systems Division) Who! EquIpment Property ContT-ol System Software | proc freq, data integrity | source | 4 |
| 160 | STARS and FACES - Procedures for Constructing Graphical Profiles of Multivariate Data | Harold W. Gugel | Humans are exceptionally proficient at processing visual information. With the widespread avail ability of computer generated graphics, much interest in recent years has focused on graphical methocis to supplement analytical techniques for studying complex multivariable data. This paper describes the implementation of several of these methocis as two new $AS procedures named STARS and FACES. These procedures create graphical profiles of multivariate observations in the form of star like polygons and face-like images to collectively picturize the information in the data. Color, combined with several analytical techniques, is added to current graphical methods to extend their utility at representing high-dimensional data. Ease-of-use in executing the procedures and flexi bility is stressed to provide comprehensive and effective tools for exploring and summarizing mul tivariable data | proc step, correlation, outlier, sample | source | 6 |
| 161 | State Space Modeling | John C. Brocklebank, David A. Dickey |
source | 11 | ||
| 162 | Statistical Graphics Made Possible by the SAS/GRAPH ANNOTATE Facility under VMS | Patricia M.D. Benoit | proc sgplot, box plot, by group processing, sample, data science | source | 7 | |
| 163 | Subset Selection, Not Multiple Comparisons | Paul N. Somerville | When the object of an experiment ;s to select the best of a number of treatments, then the use of multiple comparison procedures is ineffi cient. A subset selection procedure which selects the smallest subset which contains the best treatment with pre-specified confidence ;s preferable. Examples are the procedure of Gupta (1956, 1965) and the recently proposed multiple range subset selection of Somerville (1984). This paper includes preliminary tables and algorithms for the case where the underlying common variance is unknown, but for which an estimate with v degrees of freedom is available. An example is included | degrees of freedom, sample | source | 3 |
| 164 | Suitability of Microcomputers for Statistical Analysis | Phil Spector | This paper will examine some of the features of microcomputers which make them an appealing tool for statistical analysis, as well as some of the problems which may be experienced when trying to develop programs for use on the microcomputer. Illustrations of these concepts will be provided by examining some features of PROC GLM in the SAS System under PC-DOS |
data science, proc step | source | 4 |
| 165 | SURVCOMP-SAS MACRO for Descriptive Survival Analysis | Jeannette Y. Lee, J. Michael White |
The goal of many clinical trials is to evaluate the effectiveness of treatments in prolonging patient survival. In these studies, interest is not only on what proportion of patients died but on how long after treatment the deaths occurred. Typically, survival data are illustrated by a plot of the cumulative proportion surviving against time, and summarized by the median survival time. A commonly used approach to estimating the cumulative proportion surviving and median survival time in SAS is to access the program BMDPlL through the SAS/BMDP interface. As the output of BMD1'1L cannot be saved in a SAS dataset, plots of the cumulative proportion survlvlng cannot be generated using PROe GPLOT. Another drawback of BMDP1L is that it overestimates that median survival time. This paper presents %SURVCOMP, a SAS %MACRO which calculates the cumulative proportion surviving; plots it against time using PROe GPLOT; and estimates that median survival and its 95% confidence limits |
median survival, confidence interval, macro definition, by group processing, censoring, clinical trial, proc step, format statement, kaplan-meier plot | source | 8 |
| 166 | Survey Design from Start to Finish | Timothy A. Grunwald, Samuel M. Slowinski |
Prior to approval of a benchmark survey for Federal Reserve System consumer credit components, questions arise as to how much error can be expected in the estimates. The amount of error depends on the sample size, stratification, and method of estimation. SAS Macros were written to use Neyman alloca tion over a number of possible sample sizes and produce standard errors using ratio esti mation | sample, sample size, macro variable, proc calis, proc step | source | 6 |
| 167 | Taking Advantage of the Mapping and Three-dimensional Display Capabilities of SAS/GRAPH Software | Bradford S. Simmons | How people use energy is a major research issue throughout the world. San Diego Gas & Electric embarked on an ambitious program to study energy demand and how it could be reduced in homes with central air conditioners and electric water heaters. An experiment was performed using radio controlled devices designed to turn off air condi tioners and electric water heaters during the utility's peak period of operation. A variety of very large data bases were used in this study. The data in cluded information about the struc tural characteristics of houses family demographic data, outdoor temperature recorded every fifteen minutes and energy information recorded every five minutes in the home. The energy re corder information alone required 400,000,000 bytes of storage space on an annual basis. The challenge was to organize and reduce the data to a manageable form. SAS and SASGRAPH were used to perform this function. There were several levels of analysis performed on this data | proc tabulate, data science | source | 4 |
| 168 | Taking MDS Beyond Similarity Data | J.O. Ramsay | likelihood function, sample, maximum likelihood estimation, degrees of freedom, sample size, proc step, regression analysis | source | 7 | |
| 169 | Techniques for Using the SAS System in an On-line Production Environment | Barbara Kennedy | source | 4 | ||
| 170 | TESS-The Equipment and Software System | Richard David, Mark Freeman |
A major concern of large data processing organizations Is the management of com puter hardware, software and communica tion lines. TESS - The Equipment and Software System was developed to assist in the management and control of these items. Functional capabilities include: Inventory Control Vendor Invoice Approval Budget/Expense Forecasting Bill To/Charge Back Configuration Management Features Accounting Development objectives included that the system be on-line, easy to use and easy to maintain. These objectives were achieved by using SAS software. SAS software was chosen because of the various procedures and facilities which are included in the package and because of its ease of maintenance. Among the facilities available which are utilized in TESS are the macro language, %INCLUDE, stored formats and SAS/FSP software. By combining these facilities an on-line system was created which enables users to interactively editor browse data files |
batch job, macro facility, proc step | source | 6 |
| 171 | The Current and Potential Role of the SAS System in Industrial Applications | James F. Sattler, Joshua Sharlin |
quality control, data governance | source | 12 | |
| 172 | The Development of User Support Manuals for SAS Software: Some Guidelines | David L. Masamitsu | data wrangling, sas log | source | 6 | |
| 173 | The Emerging SAS User and the Traditional DP Establishment | Jim C. Knoop | What is it about the existing structure of many large corporate data processing organizations that can transform the proficient SAS user, rich in technical skills but often data-poor. from a vague threat to the status quo into a confirmed nemesis? Why do many frustrated SAS users perceive established procedures and standards enforced by DP areas as obstacles to he sUr mounted rather than as necessary safeguards? Furthermore, given that many SAS users are technically self-sufficient, why are their dealings with DP departments so often characterized by the same adversary relationship traditionally associated with non-technical end-users? By focusing on the technically self-sufficient SAS user as a case in point, this paper address es a somewhat unusual aspect of the ongoing conflict between end-users and the DP !festablishment lt * Both the dynamics and the roots of this conflict are examined, and practical sug gestions are presented which should help SAS users: 1) to anticipate what to | data wrangling, audit trail | source | 8 |
| 174 | The Evolution - of a Decision Support System in Conjunction with a Clinical/Financial Data Base | Sara O. Briller | calendar year, bar chart | source | 6 | |
| 175 | The HP 7470 Plotter as an Offline SAS/GRAPH Device | Thomas A. Miron | proc step, macro facility | source | 5 | |
| 176 | The Introduction of SAS Software and the Strategy Used to Spread Its Use | Akira Yagi | data science, sas function | source | 6 | |
| 177 | The Jonckheere-Terpstra Non parametric Test for Trend Utilizing SAS Macros | Robert S. Hoop, Iraj Mohebalian |
When the apriori researeh interest is testing for trend in a completely randomized design, the Jonchkheere-Terpstra non parametric test is usefuL It is used -to test the hypothesis that three or more treatments are in a definite order, e.g., the second is better than the first, the third better then the sec:ond. A SAS MACRO has been written to produce 1. Hypothesis tests for small or large sample sizes 2. F.s timates of treatment differenc.es 3. Confidence lntervals of estimates of treatment differenc.es 4. Estimates of treament differences weighted by sample size 5. A graph of the medians for the treatment groups | sample, sample size, anova | source | 5 |
| 178 | The Obsolescence of the SAS System | Grant Blank | The SAS System has established itself as one of the standards against which quality software is measured. It is so good that, in the area where it began, statistical analysis, it has virtually no direct competition. Yet, the environment for the software that SAS Institute provides is changing. Users are becoming more knowledgeable and are demanding more sophisticated but easier-to-use information processing and data analysis capabilities. Furthermore, microcomputers provide a much more user-friendly environment than mainframes and their speed and memory capacity are increasing rapidly while their costs are dropping. The software appropriate for this environment is different than that designed for other environments. Its hallmark is a user interface that is highly interactive, easy-to-Iearn and easy-to-use. The SAS System is not well-adapted to this environment | data science, proc step | source | 6 |
| 179 | The PC Version Of the SAS System | Claire Cates, Jeffrey A. Polzin, John P. Sall |
source | 7 | ||
| 180 | The Practical Value of Logistic Regression | Frank E. Harrell Jr., Kerry L. Lee |
Logistic multiple regression using the | logistic distribution, regression analysis, logistic regression, class variable, predictor variable, proc step, response variable, discriminant analysis, multiple regression, likelihood function, macro facility, proc freq | source | 6 |
| 181 | The Role of Invertibility and Symmetry of Polychotomous Metameters for GSK and Other Related Analyses | James E. Dunn | logistic distribution, sample, likelihood function, proc step, proc glm, proc genmod | source | 10 | |
| 182 | The SAS Biostatistical Workbook: Bridging the Gaps Between Statistical Methods, Statistical Applications, and Statistical Computing | David A. Ludwig | anova, proc freq, regression analysis, multiple regression | source | 6 | |
| 183 | The SAS Display Manager System for the OS/MVS Environment | R. Michael Oatsvall | source | |||
| 184 | The SAS System as a Statistical Simulation Language | Timothy J. Breen, Robert M. Hamer |
This paper discusses use of the SAS System as a tool for simulation: as a simulation language. First, we will define some terms and lay groundwork. Second, we discuss characteristics of the type of simulation about which we will restrict ourselves, the statistical simulation or monte carlo study. Third, we will enumerate characteristics of a good simulation language, and fourth, we will examine the SAS System in light of these characteristics. Finally, we include a brief discussion of methods for performing simulations quickly and efficiently using the SAS System. Rubinstein (1981) cites Naylor et al (1966), who define simulation as " | proc step, error message, data science, monte carlo, proc printto, sample | source | 8 |
| 185 | The SAS/GRAPH ANNOTATE Facility | Anthony L. Friebel | The annotate facility permits you to customize SAS/GRAPH procedure output. The facility is capable of defining polygons, controlling pen movement, placing text on the page, and so on. It adds to existing procedures the capability of producing custom graphic output from a DATA STEP |
source | 3 | |
| 186 | The SUMMSTAT Macro: Getting More Mileage Out of PROC SUMMARY | Michael D. Rhoads | proc step, proc means, macro variable, proc print, macro conditional, macro definition, proc sort, proc transpose | source | 6 | |
| 187 | The Tortoise and the Hare-A SAS Fable in VAX/VMS | Jane M. Booker, Myrle M. Johnson |
The intf".rest and on€stions SUGr I 84 re~ardin~ nsa~e of that emerged in SAS on Vax 11/700 series machines oromoted the followin\! investi\!ation. Three areaf; of concern are examined: hmo7 does SAS comnare to BMDP in r/o Rnd CPU time, what condit inns canse machine degradRtion when multinle users run SAS jobs, Rnd how eff icient is the l1se of SAS data sets. Since the Vax is very slow in r/o comnared to IBM machinef; and SAS is very he.qvilY based in r/o oner.qtions, we Rnticinated th.::tt SAS wOllld be the tortnise and BMDP the hare in both real time and CPU nerfnrmance. However. we Rll know how that fable ended. The stndies were comnarisons of statistical nrocednres nsin\! SAS ven;inn 4.07 on a Vax 11/700 series with VMS. The viewuoint is that nf a statistician and not 3 cOMpnter (or even a SAS) expert. I. Introduction B.<I.sed nn SAS/VMS I1ser comnl;:dnts ahout slnw ,iob execntion, the followin\! investig.qtinn comnared SAS to BMDP, monitored SAS 11se for machine de!!rA.dation |
cpu time, proc step | source | 4 |
| 188 | The Use of SAS Software as a Data Base for Compiling Safety Data on New Pharmaceuticals | David E. Tappe | The problem of having to pool and summarize safety data from many clinical trials to satisfy governmental requirements for approving an investigational new drug for marketing was simplified using a SAS' data base. Safety data from each trial was collected on an ongoing basis and saved in a SAS data library of 5 files, each containing related or similarly structured data. In this way, the data were stored in a uniform manner in one identifiable place and ambiguities or inconsistencies in the data were identified and resolved at the time the study was analyzed, when such resolutions could easily be obtained. The system was facilitated by incorporating the creation of files of safety information for individual studies into selected programs in a standardized library of SAS software used in analyzing each study | libname statement, clinical trial, clinical data | source | 5 |
| 189 | The Use of SAS/GRAPH Software to Project Census Metropolitan Areas onto a Map of the United States | Susan Ohm | A set of procedures have been written which create a IlBp of the United States showing all the metropolitan ststistical areas (MSA) as defined by the U.S. Census Bureau. This is done using a Zip Code to County Conversion Data base. In addition each county is assigned to the appropriate MSA, if it belongs in one. The same zip code data base was then used to take our internal data files which only had zip code geographic infonmation and create files with county and MSA information on them. These converted data files were then used to map customer demographics and company positioning by metropolitan areas. Examples of this use show market penetration, membership concentration, and cities which are gOing to be brought on line in different phases of a program. IN'rnODUCTION USAA is a financial services company, concentrating mainly on insu~ce and investment products for Commissioned Officers and their fonner dependents | dataset, proc step | source | 6 |
| 190 | The User-written SAS PROC: An Evolutionary Tool in the Information Center | Ronald G. Fleming | with the progression of end-user computing and the availability of "simple to use I! languages such as SAS, there is a new direction for SAS users and programmers alike. The 1980's are proving to be the era of the Information Center, the necessary link between end-user computing and the data processing environment. User-written FRoes CQuld evolve into a complete Information Center tool. The scenario is typical, an end-user would like to use data in his/her custom system and SASe The application programmer does not know how to create a SAS data set from the customized system or read a SAS data set by the system. Creating external files would become a nightmare for both the end-user and the application programmer. The Information Center links the user community to the data processing department. Similarly, the user-written SAS PROC links the SAS user I s application and the programmer's customized system |
proc step, data wrangling | source | 4 |
| 191 | Transportation Planning Support Information System-Using CMS EXEC2 to Execute SAS Macros in a Panel Driven Environment | George Ramsey | A system (DRAMS), under development by the Issue Analysis and Information SeIVices Branch Is de scribed. The functions of the system are: data storage and maintenance; provide transportation related data to planners and analysts: produce publication quality tables and graphs: and provide SAS users with an easy to use method of extracting data for more specialized analyses. This menu driven system utilizes panels formatted under Display Management System and controlled by EXEC2 programs. Users select data and data processing routines by moving the terminal's cur sor to the option desired and pressing a function key. DeciSions made In this way cause EXEC2 to write CMS GETSAS and SAS macro language statements to a temporary SAS source file. When the decision process Is complete, EXEC2 initiates execution of the SAS source file | macro definition, data wrangling, macro variable | source | 4 |
| 192 | Two-way Communication Between SAS Software and the SCA Statistical System: The LINKSCA and SCALINK Procedures | Lon-Mu Liu, Houston H. Stokes |
PROC SCALINK provides SAS users a link to the SeA Statistical System. PROC LINKSCA will read an SeA formatted file into SAS. An SeA formatted file can be creatad by the SeA System or by SAS (via a new opt i on in PROt SeAL INK) * The advantage of two way 1 i nkage between SAS/ETS software and the SeA statistical system is illustrated. An SAS dataset is created and passed to SeA which in turn is used to estimate a vector autoregressive (VAR) model. The VAR coefficients are saved by SeA ina f i 1 e wh i eh is then read back into SAS for processing by the SAS/ETS PROt MATRIX procedure MRATIO which is used to transform the estimated VAR model to vector moving-average (VMA) form. Documentation for the PROC LINK5CA is provided |
proc step, dataset, time series analysis | source | 6 |
| 193 | Understanding the ANNOTATE Facility Reference Systems | Anthony L. Friebel | source | 4 | ||
| 194 | Unique Uses of SAS Software for Determining Medi-Cal (Medicaid) Hospital In-patient Reimbursement | Robert O. Bernstein | The Hospital Reimbursement Section (HRS) of the California Department of Health Services (DRS) has been using SAS to help hold down the high and rising cost of inpatient hospital care in California since 1980. SAS is used to calculate reimbursement rates for the 549 hospitals which provide services to Medi-Cal (Medicaid) patients in California. Three regulations, each with its own set of formulas, are used by HRS to limit hospital inpatient reimbursement. SAS is used for all of the programming needs of HRS. Many unique and useful tools have been developed over the years to deal with these regulations. The SAS system used by HRS is now a complete pack age that is used to determine approximately $1.5 billion in Medi-Cal reimbursement each year | paid loss, dataset | source | 6 |
| 195 | Usage of Base SAS and SAS/GRAPH Software for Distributor Sales Performance | Rosanna Hodlik, Mike Marvin |
sas base, data wrangling | source | 6 | |
| 196 | Use of the SAS Macro Language in Developing Control Chart Limits | Thomas P. Walker Jr. | Control charts are an important tool in applied statistics, especially in qualjty control. This report describes the generation of control chart limits for process data in which the limits are developed by first grouping the raw process readings into many relatively small groups and then, from this grouped data, obtaining subsets, which are stable with respect to the standard deviation and the mean. An important feature of this report is the ~ demonstration of the effective use of the SA~ Macro Language in facilitating the calculations required to develop the stable subsets and the censor limits. Without the macro language capabilities, the computations would be long, tedious, and error-prone. An example is given, and the SAS Macros, which make up the program, are listed | standard deviation, sample size, sample, proc step, macro facility, macro definition | source | 4 |
| 197 | User-written Device Drivers | Howard Houston | bufsize, char. | source | 4 | |
| 198 | User-written I/O Drivers and the SAS System for Minicomputers under VAX/VMS | Tom Cole | sample, infile statement | source | 4 | |
| 199 | Using EPIC as a Device Driver for SAS/GRAPH Software Printer Graphics on the Xerox 9700 Printing System | Robert B. Holland, Louis Roberts, Thomas G. Steger |
source | 5 | ||
| 200 | Using Macro-driven Menus to Selectively Process SAS Procedures | Michael Gutting | When writing a SAS program to be executed by an individual with little or no SAS experience, the interface should be as transparent as possible. SAS is an ideal language to use in the creation of many different kinds of "production" programs. The difficulty with using SAS is that in order for someone to take advantage of all of the interactive capabilities of SAS, the user needs to have a working knowledge of the language. This sometimes limits the usefulness of an otherwise very desirable program. A more desirable situation is one in whi ch the user of the program has the ,.bil i ty to enter SAS interactively via a TSO CLIST, selectively process a wide variety of data, view the interim output, choose to print or delete the output upon termination of the program and have the entire process be transparent to the user. The method discus sed in th i s paper allows a non-technical user to selectively execute SAS code based upon responses to menus created using SAS/FSP |
proc step, macro variable, macro definition | source | 4 |
| 201 | Using PROC COPYLIB to Create a SAS-based Data Dictionary | Wayne Beekman | data dictionary, proc step, out= | source | 3 | |
| 202 | Using PROC LP to Optimize a Common Stock Portfolio | Raymond A. Gibley | linear programming, proc step | source | 2 | |
| 203 | Using SAS Software as the Report Writer in a Data Base Package | Jack Jimieson, Thomas M. Koenig |
proc step, proc print, sample, title statement, if-then-else | source | 6 | |
| 204 | Using SAS Software to Estimate and Plot a Regression Line and Confidence Band in Multiple Treatment Linear Regression | Howard Hait | In an experiment where samples from various treatments are drawn and observed over some prespecified length of time, regression meth ODS are often used to fit a line and confidence band representing treatment behavior over the time frame. The best fitting regression line is one that takes into account treatment effect and the results of tests of homogeneity of treatment slopes. Unfortunately, a line created by the SASjGRAPH R-series interpolation command RLCLH would be from a model where these condi tions have been ignored. Therefore, a macro is written which uses the output from PROC GLM to construct a line and confidence band that best represent the data after evaluating these conditions |
regression analysis, proc reg, proc step, proc genmod | source | 6 |
| 205 | Using SAS Software to Generate Unsupervised LANDSAT Classifications | Virginia R. Hetrick, Kenneth M. Portier |
This paper briefly describes a series of Statistical Analysis System (SAS) procedures and a single FORTRAN program which may be used in the analysis of digital image data from the LANDSAT series of satellites. These procedures permit the user to complete a digital image classification without the need for the highly specialized image processing systems generally used for this type of analysis | proc princomp, proc fastclus, proc step, histogram | source | 6 |
| 206 | Using SAS/FSP and SAS/GRAPH Software with Canadian Census Tapes | William R. Fehlner, Diane Mitchell |
The University of Toronto Computing Services has a large collection of tapes with data from three Canadian censuses. A project was recently undertaken t'o make it easier for users to locate the tape with the desired information, and to access the data. All necessary details about the tapes themselves were put in a SAS data set, and documentation written to facilitate its use through PROC FSBROWSE or PROC PRINT. Several sample SAS pro9rams were written to illustrate varIOUS techniques for reading the census information. Since census data is frequently presented on maps, UTeS is working with the Department of Geography to expand its collection of SAS map data sets. As a |
proc step, dataset, sample | source | 6 |
| 207 | Using the SAS Macro Facility to Tabulate Counts and Percents for Individual Variable Values | Thomas J. Higgins, Rich Potter |
macro variable, proc step, proc freq, macro facility, macro definition, variable, proc sort, dataset | source | 6 | |
| 208 | Using the SAS System for Course Authoring | John C. Boling | source | 6 | ||
| 209 | Using the SAS System to Process 1980 Census Summary Tapes | John G. Blodgett | The series of Summary Tape Files produced by the U.S. Census Bureau represent the primary machine readable products of the 1980 decennial census. Because of their massive size and com plexity these files present a challenge to any shop that needs to use the information in a timely and efficient manner. The staff of the UMSL Urban Information Center has created a SAS based software package that we feel can signif icantly reduce the costs and programmer effort required to use these files effectively and reliably | dataset, sample, libname statement | source | 5 |
| 210 | Writing Efficient SAS DATA Steps | Stephen M. Beatrous, Darylene C. Colbert, Virginia M. Dineley, Bruce M. Tindall |
Often a user has a choice of two or more programming styles that will accomplish a task equally well. Normally, the style chosen depends on the user-'s previous exper'jence with some other programming language. However, the chosen style may not necessarily be the one that pI"oduces the most efficient SAS code internally. This paper" presents an overview of how a user's SAS progt'am is optimized in the DATA STEP. Examples of vat-jous SAS programming techniques are presented, along with the style that gives the user the optimal SAS code |
put statement, dataset, input statement | source | 4 |
| 211 | Writing SAS Procedures | Terri L. Betancourt | Writing a SAS procedure requires a certain level of programming skill. The procedure writer needs to have a clear' understanding of data structures, pointer manipulation, and dynamic storage allocation, as well as a waf-king knowledge of the SAS System. Although oriented toward SAS users in the minicomputer environment, most of the ideas presented here will apply to other users as well. The intent of this paper is not to explain the technical information and detail requi red to write a SAS procedure -- that information is available in the SAS Programmer's Guide. Instead, the attempt is to construct a conceptual framework upon which the programmer can build | proc step, dataset, variable list, options statement, proc standard, sas base, batch job, data science, proc print, by group processing | source | 7 |
| 212 | ZOOMMAP and Other Special Techniques for SAS Mapping Applications | Warren Repole Jr. | proc step, proc gmap | source | 6 |
This site serves as a convenient hosting location and does not work as a publisher. All copyrights belong to their respective owners.
The information on this site is provided for general reference purposes only. Whilst every effort is made to ensure that the site is up to date and accurate, no warrant, nor any responsibility or liability for, the accuracy or completeness of the content or for any loss which may arise from reliance on information contained in this site is provided. Links to other sites are provided for your convenience but no responsibility or liability for the content of those sites or of any external site which links to this site is accepted. References to any specific commercial product, process, service, or corporation are also for information only, and do not constitute endorsement, recommendation or favoring.
SAS is a registered trademark of SAS Institute Inc. SAS, Statistical Analysis System, and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration.