As managers, we are often asked to ensure that our business processes are optimized. There are many software solutions for simply tracking work. But true process optimization can be attained only by carefully examining inputs and studying how varying those factors influences the outcome. Without specialized tools such as SAS^® Simulation Studio, obtaining accurate results is very challenging due to the complexity of modern business. In this paper, we use SAS Simulation Studio to conduct experiments of how changes in work volume and resource availability impact process efficiency and capacity of an essential business process. Modeling a process in a simulated environment allows for true A-B testing. The point-and-click interface is easy to use and empowers even novice users to translate real-world scenarios into reliable estimates. The desktop deployment makes it easy to install for businesses of any size, and integration with JMP^® enables quick analysis of results, which promotes experimentation and data-driven decision-making. Business process management is often called a science and an art; analytics helps ensure that the science is solid.

There are many languages that co-exist in the ecosystem of your SAS^® toolbox. This Hands-On Workshop teaches you how to use four SAS languages - Base SAS, PROC SQL, Perl language elements, and the SAS^® Macro Language - to help you manipulate and investigate your data. Learn to leverage these powerful languages to check your data with simple, yet elegant techniques such as Boolean logic in PROC SQL, operators such as the SOUNDS-LIKE operator in the DATA step and PROC step, functions such as the SCAN function in the DATA step, efficient checking of your data with Perl regular expressions, and last but not least, the amazing marriage between PROC SQL and the SAS Macro Language to hold data you just found in a variable that you can use over and over again. This workshop focuses on coding techniques for data investigation and manipulation using Base SAS.

If the goal is to improve student outcomes, intervention analytics is required! The last few decades have seen a tremendous growth in the depth and clarity of big data in education, including a greater emphasis on using analytics to improve student outcomes. Despite the genesis and transition from data-driven decision-making to actionable data to visual analytics, has there really been a systemic transformation in the effective use of analytics in education? This paper introduces the use of intervention analytics to improve the educational outcomes for all students. A key element of this process is reliance on the single-source SAS^® model that transforms data, integrates an innovative analytics model, identifies unique educational interventions, and transitions information to web-based platforms.

Learning SAS^® for the first time as a budding epidemiologist in graduate school, I remember being amazed at how powerful and customizable SAS was in making sense of huge amounts of data. I also remember being very intimidated by all the moving parts involved in learning such a program! My study and use of SAS has continued since that time, specifically now as a Doctor of Public Health student. I have often wondered how much farther along my research would be at this point had I learned SAS in high school. When one of my professors recently told me that resources are available for teaching SAS at that level, it prompted me to ask a local high school teacher what she thought about teaching SAS to her students. When she replied that she thought it was a great idea but did not know how she could do this, I realized this was a gap that needed to be filled. The aim of this presentation is to raise the awareness of educators about the importance of, and resources available for, teaching SAS to high school students.

The new tmCooccur action in SAS^® Viya^® 3.4 detects significant co-occurrences of terms in sentences. It can identify patterns and nuances such as the following: 1) the same word used with different meanings (for example, They met along the bank of a river versus They made a deposit in a bank); 2) verbs used as regular verbs versus verb-particle combinations (for example, shut the heck up versus shut the door please); and 3) less common uses of negation, such as statements like He liked the show very much versus He liked the show not at all. This paper describes how the results of the tmCooccur action can be used to generate word embeddings on the basis of local information (in a manner similar to GloVe and Word2vec), and also how it can be used to create virtual terms that represent the most significant term co-occurrences. These new composite terms can then be used to create better topic, concept, and category definitions. Doing so can generate significant gains in lift for predictive modeling on some real-world data. Finally, this paper describes how to use the tmCooccur action with other kinds of transactional data, including IOT (Internet of Things) and genomic data.

The last few years have been an exciting time for the map geeks at SAS. Weve added a lot of geospatial functionality throughout SAS^®, and we cant wait to show it off. This paper goes over all the additions, including GeoEnrichment, geocoding, the SGMAP procedure, contours, WebMap support, routing, location pins, clustering, and custom polygon support. We also show you the enhancements in progress and detail the roadmap for the future.

Its impossible to know all of SAS^® or all of statistics. There will always be some technique that you don't know. However, there are a few techniques that anyone in biostatistics should know. If you can calculate those with SAS, life is all the better. In this session, you learn how to compute and interpret a dozen of these techniques, including several statistics that are frequently confused. The following statistics are covered: prevalence, incidence, sensitivity, specificity, attributable fraction, population attributable fraction, risk difference, relative risk, odds ratio, Fishers exact test, number needed to treat, and McNemars test. The 13th, an extra bonus tool, is the ability to produce statistical graphics using the Output Delivery System (ODS) in conjunction with SAS/STAT^® software. With these tools in your tool chest, even non-statisticians or statisticians who are not specialists will be able to answer many common questions in biostatistics. The fact that each of these can be computed with a few statements in SAS makes the situation even better.

In this paper, we describe how to use SAS^® Customer Intelligence for campaign execution and post-campaign reporting, and how to avoid three most commonly found campaign execution hurdles. SAS Customer Intelligence is an integrated marketing management suite. It combines multichannel campaign management, marketing resource management, and marketing performance management with strong data management and a wide range of analytic capabilities. Using an example of a monthly summary campaign, this paper helps you understand SAS Customer Intelligence, what it is, what its possibilities are, and where it can be used. We present three challenges to address using SAS Customer Intelligence in the hospitality industry: 1) Generating one row per guest with different offers; 2) Creating a marketing history ID; and 3) Creating an output file with order information. We also describe the main benefits of using SAS Customer Intelligence in campaign execution and how it can be implemented.

The SGPLOT procedure is a powerful procedure for creating various graphics. Sometimes, just a graphic is not satisfying, and you might want to add some text to it. This presentation talks about three powerful and easy-to-learn ways to customize text in our figures by using special statements and options in PROC SGPLOT.

Would you like to know how the clinical trial you design today might perform in the future? Will your design be able to reach the target patient enrollment fast enough while staying within budget? Come and learn how SAS^® Clinical Trial Enrollment Simulator, built on SAS/OR^® software, provides pharmaceutical and clinical research companies the power to predict the future performance of their clinical trial enrollment designs.

SAS^® Visual Investigator provides users with the flexibility to build applications that are customized to their business needs. This flexibility includes the ability to define how data is analyzed to generate alerts, how data is indexed to enable users to search, how data is presented to the users, what data users can enter into the system, and what workflows are in place with respect to data entry. As a result, SAS Visual Investigator can be used for applications in many different markets. This paper covers some of the best practices for building applications on top of SAS Visual Investigator to ensure maximum performance and utilization of available resources.

Starting a new SAS^® project can be very stressful in the beginning, but there are SAS functions that you can use in the initial stage of a project to explore the data and the business processes. Like many other consultants, I have developed a check-list that has become so vital to my personal onboarding. I routinely use it at the beginning of every project. In this paper, I share this check-list with you, along with functions I use myself, as well as a few examples to better demonstrate my process. I hope they serve as an inspiration for other SAS consultants and programmers to improve their effectiveness and success as they embark on new journeys.

Although not as frequent as merging, a data manipulation task that SAS^® programmers are required to perform is vertically combining SAS data sets. SAS provides multiple techniques for appending SAS data sets, which is otherwise known as concatenating, or stacking. There are pitfalls and adverse data quality consequences for using traditional approaches to appending data sets. There are also efficiency implications with using different methods to append SAS data files. In this paper, with practical examples, I examine the technical procedures that are necessary to follow to prepare data to be appended. I also compare different methods that are available in Base SAS^® to append SAS data sets, based on efficiency criteria.

The number of vendors providing solutions for predicting student success keeps increasing. Each vendor claims to be able to use your data and to predict the likelihood of retention or graduation for you. This all comes with a financial cost and required time and effort to prepare and update data to feed their systems. Is creating an in-house system a better solution? This paper walks through one way to make that determination. If your institution has gone through changes impacting the continuity of data or has made other significant structural changes, finding an off-the-shelf solution can be more challenging. Should you include additional variables that the commercial products do not consider? A side benefit of creating the model in-house is validating multiple models on past students as well as working with those who will be using the model output to generate buy-in. Rolling out a commercial solution is of no benefit if those using it do not trust the results. SAS^® has the tools to do the modeling whether you use Base SAS^®, SAS^® Enterprise Miner (TM), or SAS^® Visual Analytics. Why not leverage what you have to support your students success?

Are you a SAS^® infrastructure architect? Do you want to know what has worked for other SAS customers and what things should be avoided? Attend this presentation to learn from others experiences with SAS. We also encourage you to share your experiences with us.

At least three human vaccine quality scandals broke out in China in the past five years. The recent one that was reported in July 2018 triggered an outcry from Chinese parents on the internet. They were eager to find out whether the vaccines that their children got were at risk. However, there was no complete vaccine tracing database that was available to the public for easy searching. We gathered public vaccine sourcing information and consolidated the data to make a comprehensive database for general public to check their vaccine sources. We used Base SAS^® for web scraping, data cleansing, and data consolidation. Our database is used to provide an easy way to check the risk associated with vaccine incidents.

In binary logistic models for credit risk and marketing applications, as well as for many social science studies, there are ordinal predictors that are considered for use in the model. The modeler faces the question of how to use these ordinal predictors. Choices include conversion to dummy variables, recoding as weight of evidence, or assuming an interval scale for the ordinal levels and entering the effect as linear (or possibly as some transformation). The use of dummies or weight of evidence add parameters to the model (for weight of evidence the added parameters are implicit). This paper provides a procedure to aid in decision-making regarding the handling of ordinal predictors. First, a measure is computed of the monotonic tendency of the ordinal versus the target (the dependent variable). If the ordinal is nearly monotonic, then a simple numeric recoding of the ordinal is made. This imposes an interval scale. Is this recoding appropriate? To answer this question, a model comparison test is made between the saturated model (all dummies) and the model with the numeric (recoded ordinal) predictor. Acceptance of the null hypothesis of no difference allows the numeric (recoded ordinal) to be considered for inclusion in the model. SAS^® code that carries out the ordinal to numeric recoding as well as the model comparison test is included. This code can process many ordinal predictors using an efficient, data-driven approach that does not require hard coding.

One of the most common concerns for large organizations is the risk of performance degradation resulting from an increase in demand of SAS^® workloads. The most common pattern in a Linux platform is the delegation of authentication requests to the corporate Active Directory via Pluggable Authentication Modules (PAM) using the System Security Services Daemon (SSSD), an area that is not sufficiently understood nor explored as yet. A SAS administrator is expected to minimize the impact that an increase in user concurrency brings to the SAS login elapse time, which can cause performance degradation. Did you know that such performance problems can be addressed by enabling and configuring a brand new multithreaded SAS authentication daemon available on the latest version of SAS? Did you also know that the authentication process can be further improved by configuring SSSD caching options, which will avoid multiple calls to Active Directory by the same user session? Through a combination of both these daemons, we can remarkably improve the performance of your SAS^® Platform to support a large number of concurrent users. Its time to face your daemons.

When running SAS^® programs that use large amounts of data or have complicated algorithms, we often are frustrated by the amount of time it takes for the programs to run and by the large amount of space required for the program to run to completion. Even experienced SAS programmers sometimes run into this situation, perhaps through the need to produce results quickly, through a change in the data source, through inheriting someone else's programs, or for some other reason. This paper outlines twenty techniques that can reduce the time and space required for a program without requiring an extended period of time for the modifications.

Although organizations are increasingly leveraging SAS^® Viya^®, many have a substantial investment in SAS^® 9.4 applications and will for some time. This paper looks at a new feature introduced in SAS Viya 3.4 that provides single sign-on with SAS 9.4, enabling users to move seamlessly between web applications in both environments without having to log on again, and making it possible to use them collaboratively (for example, displaying a SAS^® Visual Analytics report from SAS Viya in the SAS^® Information Delivery Portal).

Hybrid vehicles let you use the best technology for a given scenario, blending fuel efficiency with long driving range, and low emissions with quick refueling. Its like two cars in one two motors magically complementing each other under the hood, invisible to the driver. And as the driver, you have just one control panel that chooses the best engine for the circumstances, adjusting to changes in real time. We marketers face a similar dilemma: we have different engines for different scenarios. Not only do we have to learn how to drive multiple systems to address the diversity of interactions we have with our customers, we must make them work together seamlessly to deliver delightful customer experiences. Sounds more like Marketing: Impossible! Are you thinking what Im thinking? Yes! What the world needs now is a revolutionary approach hybrid marketing! The good news is that SAS has been working hard on this and already has an answer for you. In this session, you learn what hybrid marketing is and why its important to your marketing success. You learn how this leverages your existing investment in SAS^®, and keeps your secret sauce analytics and your data safe, as well as keeps your business processes in compliance with privacy-related regulations. We share our unique approach to this need, showcase a few analytic tricks we've already built for marketers, and describe in both business and technical terms how hybrid marketing can work for you.

This paper shows how you can use SAS^® Visual Data Mining and Machine Learning and other SAS products to build and compare various predictive models. First, you use SAS Visual Data Mining and Machine Learning to create several models and you choose one of them as your champion model. You can publish all these models to different destination types such as Hadoop, Teradata, or SAS^® Cloud Analytic Services (CAS). You use SAS^® Embedded Process to score the data against these published models where the data reside. You can also register the models to SAS^® Model Manager and compare them against other models for final champion model selection. You can then test these models to validate them for scoring. If you notice a degradation in the model, you can retrain the model. Retraining the model triggers a run of all the pipelines in the associated Visual Data Mining and Machine Learning project, and the recalculated project champion is automatically registered back to Model Manager. In addition, you can score streaming data by using SAS^® Event Stream Processing on the models in Model Manager. SAS Visual Data Mining and Machine Learning also provides a scoring API that enables you to score models directly in Model Studio by using RESTful interfaces. This paper shows how you can unleash the full power of your models by taking advantage of the model processing capabilities in all these SAS products.

Are you a SAS^® programmer searching for insights within Salesforce data? Accessing and analyzing your information is now easier than ever before with the all-new SAS/ACCESS^® Interface to Salesforce. This paper provides you with an overview of how to explore Salesforce objects as if they were native SAS data sets. Learn how to directly execute the SOQL language used by Salesforce by using the SQL procedure, and take advantage of the advanced features of implicit SQL to let SAS do some of the work for you. With the help of SAS/ACCESS Interface to Salesforce, you'll begin to understand your data in no time!

An all-too-common practice in data analytics is the use of pre-defined business groupings to drive analyses and subsequent decision-making. Product categories drive assortment layouts in retail, but are they really indicative of what shoppers buy at the same time? Geographies define divisions for sales-oriented companies, but do customers in neighboring states really behave the same? In the Compliance Technology Analytics group at Wells Fargo, account types from an existing vendor solution were driving the peer-to-peer comparisons for the anti-money laundering (AML) transaction program that covers our securities business. Accounts of the same type, or peer group, are compared to one another to identify anomalies and out-of-bounds behavior for AML alerting. However, not all accounts in the same peer group are created equal, resulting in sub-optimal comparisons and less valuable alerts for our investigative team. Using SAS^® Enterprise Miner (TM), we built more meaningful account peer groups with k-means clustering. With the new clusters, we have an apples-to-apples comparison of accounts when performing transaction monitoring and anomaly detection.

The advantage of an indexed Base SAS^® engines data set is that when it is queried with a WHERE clause, the process might be optimized and the index subsetting works faster than a sequential read. Unfortunately, the Base SAS engine is unable to use more than one index at a time to optimize a WHERE clause. Especially when the WHERE clause contains an OR condition between two different (indexed) variables, the Base SAS engine does not optimize it and executes a sequential read. This presentation offers a simple, hash table-based solution to the following question: How do you handle a situation in which you have a WHERE clause with the OR condition on two different indexed variables and you want to use the advantage of indexing to subset the data faster?

Did you ever wish you could use the power of SAS^® to take control of Microsoft Excel and make Excel do what you wanted when you wanted? Well, one letter is the key to doing just that, the letter X, as in the SAS X command that opens the door to all operating system commands from SAS. The Microsoft Windows operating system comes with a facility to write a series of commands called scripts. These scripts have the ability to open and reach into the internals of Excel. Scripts can load, execute, and remove Visual Basic for Applications (VBA) macro code and control Excel. This level of control enables you to make Excel do what you want, without leaving any traces of a macro behind. This is Power.

The United States Census Bureau (USCB) Demographic Systems Division (DSD, collectively USCB-DSD) utilizes Dataflux^® Data Management Server and DataFlux^® Data Management Studio for data management and cleansing. USCB-DSD accesses DataFlux Data Management Studio via a Microsoft Windows virtual desktop (vDESK) where users can perform data management tasks unique to their individual program area. Due to USCB security restrictions, users do not have access to the local Windows program files where the DataFlux Data Management Studio configuration is stored by default. Users also do not have the capability to establish a System Data Source Name (DSN) as required to make the connection to the repository. To solve these problems, USCB-DSD implemented a shared repository consisting of an Oracle database management system (DBMS) with the DataFlux data connection configuration saved on a shared file system and user credentials saved in the SAS metadata. This paper discusses how the USCB uses DataFlux Data Management Studio, the architecture of the USCB environment, the administrative problems encountered due to security restrictions, and a solution that minimizes the administrative problems.

This case study provides a real-world example of how Base SAS^® was used to read in more than 185 Excel workbooks to check the structure of more than 10,000 worksheets, and then to repeat the process quarterly. It illustrates how components such as the SASIOXLX (XLSX) Engine, PROC SQL (to create macro variables), SAS^® dictionary tables, and SAS macros were used together to create exception reports exported to Excel workbooks. The structure of the worksheets, such as worksheet names and variable names, were checked against pre-loaded templates. Values in the worksheets were checked for missing and invalid data and percent differences of numeric data. And, it was verified that specific observations were included in specific worksheets. This case study describes the process from its inception to its ongoing enhancements and modifications. Follow along and see how each challenge of the process was undertaken and how SAS Global Forum papers and presentations contributed to this quality check process.

In the present fast-moving times, when real-time decisions are needed, data quality is highly appreciated as high-quality predictive models are created based on it. Nowadays, analytics is turning into a mature, value-adding technology. A special benefit is considering future costs and revenues. In my final thesis work, which was finalized and submitted in March 2019, I pointed out the issue around increasingly evolving digital marketing to optimize the example online campaign. The aim of my project is to describe the procedure of the application of a strategy that is sensitive to profit/loss, and to compare the models of standard logistic regression and weighted logistic regression with the application of such a strategy. When perceiving the goal of a selected digital campaign from the advertisers point of view, achieving the best click-through rate (CTR), the media agency must also consider the potential costs of cost per click (CPC) or the most effective profit. This balance (win-win strategy) needs to be understood and needs to be reached very sensitively. The goals I tackle in my model focus on setting bids optimizing CPC (cost per click). I have already taken cost-sensitive classification techniques into account while learning the model. They optimize the predictions of the target variable by specifying the costs.

Are formats associated with your data? Formats that you created in SAS^® using the FORMAT procedure? If you want to use SAS^® Visual Analytics on SAS^® Viya^® with your data, you need to load your formats so that they can be found. In SAS^®9, this meant adding the catalog that contained your formats to the format search path. On SAS Viya, there is an analogous task but the process is different. First, SAS Viya cannot read a SAS catalog, so you need to convert your catalog to a structure that SAS Viya understands. Next, you need to place the formats in a location where SAS Viya can find them. And finally, you need to tell SAS Viya to load the formats whenever a new session is started. This paper explores how you, as a user of SAS Visual Analytics on SAS Viya, can accomplish these steps and make your formats available. You can do this by using SAS^® Environment Manager, SAS^® Studio, or shell scripts. Each of these methods is described in detail, including sample code, and the benefits and limitations of each are presented.

Cars in Denmark are heavily taxed, and the level of taxation is changed from time to time. In recent years, first an extra tax was introduced in order to reduce fuel usage, and two years later a new tax was introduced in order to make more secure cars affordable. The first intervention was in favor of small cars, while the second was in favor of large cars. Whether such changes in the taxation rate were introduced immediately or were announced some months before makes a difference in their impact. Also, changes in taxation are expected by everyone interested in the negotiations in parliament. Statistics Denmark publishes the number of cars sold in many segments. In this paper, the car segments small, medium, and large are used in order to illustrate the effect of these changes in taxation. The models applied are estimated by the ARIMA procedure with intervention components, because exogenous, deterministic effects like changes in taxation should be modeled separately by deterministic components and not as a part of the stochastic model. In this paper, models for exponentially decreasing impacts of an intervention are of interest. However, situations where impacts are increasing up to the actual date of an announced future change in taxation also exist in this data set. Such situations are easily modeled by reverting the direction of time.

Do you have a complex report involving multiple tables, text items, and graphics that could best be displayed in a multi-tabbed spreadsheet format? The Output Delivery System (ODS) destination for Excel, introduced in SAS^® 9.4, enables you to create Microsoft Excel workbooks that easily integrate graphics, text, and tables, including column labels, filters, and formatted data values. In this paper, we examine the syntax used to generate a multi-tabbed Excel report that incorporates output from the REPORT, PRINT, SGPLOT, and SGPANEL procedures.

Prostate cancer (PrCA) incidence and mortality rates are higher among African-American men than any other racial group. Informed decision-making about prostate cancer screening could result in early detection and potentially reduce cancer health disparities. Currently, there are some, but few, computer-based decision aids to facilitate PrCA decisions of African-American men, but no scale has been validated to assess the extent to which African-American men accept and use a computer-based PrCA screening decision aid. Using parallel analysis, this study determined the dimensionality of the Computer-Based Prostate Cancer Screening Decision Aid and Acceptance Scale using data from a purposive sample of 352 African-American men aged 40 years and older who resided in South Carolina. Exploratory factor analysis was conducted using maximum likelihood, squared multiple correlations, and Promax rotation. Internal consistency reliability was assessed using Cronbachs alpha. Parallel analysis was used to determine the dimensionality of the scale using SAS^® macro language. Results showed the optimal factor structure of the Computer-Based Prostate Cancer Screening Decision Aid among African-American men was a 24-item, 3-factor model. Factor loadings ranged from 0.32 to 0.94 with 11 items loading on Factor 1, 8 items on Factor 2, and 5 items on Factor 3. Parallel analysis is a valuable method for determining the dimensionality of the Computer-Based Prostate Cancer Screening Decision Aid.

The classic Traveling Salesman Problem (TSP) establishes a list of cities to visit and costs associated with travel to each location. The goal is to produce a cycle of minimum cost that visits each city and returns the salesperson to the home location. What happens if an imposed time limit on the journey makes visiting each location impossible? Assuming each location is of equal value, then the goal transforms into visiting as many locations as possible within the imposed time limit. This variation is known as the Prize-Collecting Traveling Salesman Problem (PCTSP). In this paper, we use the OPTGRAPH procedure to implement the TSP, develop an approach to implement the PCTSP, and utilize SAS^® Viya^® to map results. Set in a suburb of metropolitan Atlanta, our motivation stems from the need to acquire as many virtual Pokemon as possible for Alice, the nine-year old daughter of one of the authors. Analysis of results from executing routes generated by SAS^® is included.

Shift tables display the change in the frequency of subjects across specified categories from baseline to post-baseline time points. They are commonly used in clinical data to display the shift in the values of laboratory parameters, ECG interpretations, or other ordinal variables of interest across visits. The INTO clause in the SQL procedure can be used to create macro variables for the denominators used in these tables. These macro variables can be accessed throughout the program, allowing for easy computation of percentages and the ability to call the same macro variable to display the subject count value in the header. This paper outlines the steps for creating a shift table using an example with dummy data. It describes the process of creating macro variables in PROC SQL using the INTO clause, creating shift table shells using the DATA step, conducting frequency tabulations using PROC FREQ, calling the macro variables to calculate and present the incidence and percent, and using the macro variables for the subject count value in the headers. It then discusses the efficiency of the use of PROC SQL to create macro variable denominators over other methods of calculating denominators, such as in PROC FREQ. Code examples are provided to compare shift table generation techniques.

This article goes over the SAS^® procedures that can fit the Rasch model, such as the LOGISTIC, GENMOD, NLIN, and MCMC procedures. It describes how to use them to fit the standard Rasch model and the Rasch model with all person or item parameters fixed.

Your website tracking generates a mountain of data loaded with gems waiting to be discovered. Analyzing this data uncovers untold stories and insights when, why, and how do people visit your website? But that's only part of the story. By combining front-end data with back-end data, beautiful gold nuggets of knowledge emerge. Uniting the unique visitor identification capabilities of SAS^® Customer Intelligence 360 with customer relationship management (CRM) data empowers you to know not only when, why, and how people use your website, but also who is using it. You can see how different people and personas behave. In turn, CRM data can be fed back into SAS Customer Intelligence 360 to enable you to target each person or persona segment with more relevant content and promotions. This paper delves into how the marketing technologists behind sas.com integrate the abilities of SAS Customer Intelligence 360, SAS^® Enterprise Guide^®, and SAS^® Visual Analytics to generate golden insights.

Insight into and understanding of factors associated with student success is of great significance among universities and the prospective student population, including their parents. Prospective students are interested in learning about completion rates and their financial well-being after graduation in the form of expected earnings across universities. Various SAS^® procedures in SAS^® Enterprise Guide^® were used to import, merge, clean, and analyze a big data set of universities and colleges collected over a period of 10 years. The study uses publicly available College Scorecard data to study the factors associated with student success and financial well-being. SAS^® Visual Analytics was used to visualize and present the findings of the study. In this study, we found a few institutional characteristics to be strongly related to both student success and financial well-being.

Effective communication of foundational statistical concepts such as null and alternative distributions, Type I and Type II error, and statistical power is crucial for the teaching and learning process. Unfortunately, demonstrating such abstract concepts is often limited to lecture and simple static visualizations. Creating aesthetically appealing, yet clear and understandable, dynamic visualizations can make teaching and learning abstract concepts accessible to everyone. This paper explores the application of SAS^® Graph Template Language, data manipulation, and macro programming with dynamic parameters for creating intuitive and user-friendly visualizations to convey the relations among variables that influence statistical power. This paper provides a step-by-step review and application of the techniques used, highlighting the creation and use of macros and macro variables (for example, by using the SQL procedure), PDF functions, TEMPLATE procedure concepts such as series plots, band plots, and draw statements, precedence of statements, transparency, rich text support, droplines, dynamic anchor points, and manipulation of graph space.

Obtaining run-time statistics and performance information from SAS^® Data Integration Studio batch processes has historically been a challenge. And with growing source populations, tight batch windows, and the demand for SAS^® analytics and predictive model results, being aware of negative performance trends in the essential data management layer is as important as ever. With SAS^® Job Monitor, the ETL (extract, transform, and load) administrator can take advantage of an integrated set of components that provide a more complete view of job load status, historical run-time trends, and other audit-related information.

This paper describes a method to apply the National Center for Health Statistics (NCHS) data presentation standards for proportions across NCHS reports and data products. This method was developed by the Data Suppression Workgroup at NCHS. The new standards replace the use of the relative standard error (RSE) for proportions, which can perform poorly for very small and very large proportions. Standards are important for publications with many estimates from multiple data sources and limited space for measures of precision such as confidence intervals or standard errors. A SAS^® macro, with code for direct computation based on mathematical formulae, is used to apply the multistep data presentation standards based on minimum denominator sample size and on the absolute and relative widths of a confidence Interval calculated using the Clopper-Pearson method (or the Korn-Graubard method for complex surveys). The macro provides a clear and efficient method to implement the SAS code, and is therefore useful to programmers who wish to replicate, match, or benchmark against published NCHS estimates. This macro can also be useful to programmers and analysts with their own data systems. Data from the National Health Interview Survey (NHIS) are used to demonstrate the use of the macro.

Like right censoring, left truncation is commonly observed in lifetime data, which requires proper technique to analyze in order to achieve unbiased results. This paper introduces how to analyze such left-truncated lifetime data using the native PHREG procedure, including generating survival estimates, testing homogeneity across samples, as well as generating survival plots. Alternatively, the paper demonstrates two new macro functions, namely %LT_LIFETEST and %LT_ LOGRANKTEST, to conduct survival analysis with more flexibility for data with left truncation, such as enabling users to choose specific estimation time points and to perform weighted log-rank tests. The paper also provides updates on the previously presented %LIFETEST and %LIFETESTEXPORT macro functions for generating formatted survival analysis reports with added support for left-truncated data.

Administrative health-care data including insurance claims data, electronic medical records (EMR) data, and hospitalization data contains standardized diagnosis codes to identify diseases and other medical conditions. These codes use the short-form name of ICD, which stands for International Classification of Diseases. Much of the currently available health-care data contains the ninth version of these codes, referred to as ICD-9. Although, the more recent 10th version, ICD-10, is becoming more common in health-care data. These diagnosis codes are typically saved as character variables, are often stored in arrays of multiple codes representing primary and secondary diagnoses, and can be associated with either outpatient medical visits or inpatient hospitalizations. SAS^® text processing functions, array processing, and the SAS colon modifier can be used to analyze the text of these codes and to identify similar codes or ranges of ICD codes. In epidemiologic analyses, groups of multiple ICD diagnosis codes are typically used to define more general comorbidities or medical outcomes. These disease definitions based on multiple ICD diagnosis codes, also known as coding algorithms, can either be hardcoded in a SAS program or defined externally from the programming. When coding algorithm definitions based on ICD codes are stored externally, the definitions can be read into SAS, transformed to SAS format, and dynamically converted into SAS programming statements.

Model validation is an important step in establishing a prediction model. The model validation process quantifies how well the model predicts future outcomes. However, there are very few SAS^® programming examples showing the validation process. We previously developed a generalized mixed effect model that predicts peri-operative blood transfusion from patients characteristics. In this paper, we demonstrate the SAS techniques that we used to validate such a model. This prediction model was developed using the GLIMMIX procedure. The validation methods include calibration using the SGPLOT procedure, discrimination using the ROC option in the LOGISTIC procedure, and sensitivity analysis with bootstrapping method using SAS macro language.

Human trafficking networks commonly move from city to city in an effort to both avoid the police and to take advantage of certain events that might increase their business. In this study, unsupervised and unstructured textual data scraped from advertisements on Backpage.com is analyzed using SAS^® Viya^® textual machine learning analytics techniques to ultimately identify the types of appeals that human traffickers use in recruiting ads. We plan to use the results of this analysis to assist law enforcement in properly using its resources when identifying advertisements in which recruiting ads and their contents are properly categorized. The idea behind our analysis is to find out more about those recruiting ads by separating them from other advertisements, based on several unique features. These ads appear all over the country, but they originate in only a small number of places. Using these places and the unique features from each ad enables us to map each potential recruitment ring. Custom Concepts in SAS Viya enables us to classify indicators of the ad being the reason for the advertisements appearance, such as certain words or phrases that might indicate whether the trafficking network moved to the area in order to benefit from the recruitments increased demand. Assigning term densities to the Custom Concepts enables us to group similar posts based on their duration and attractor, and we can therefore identify direct links to traffickers.

This session is designed for data scientists looking to develop custom analytics solutions backed by SAS^® analytic capabilities. First, we describe an analytics environment based on the SAS^® Viya^® and SAS^® Cloud Analytic Services (CAS) architectural foundation. Next, using cybersecurity as an example, we demonstrate how SAS analytics can be implemented in Python in this environment. Finally, we include a JupyterHub installation, preconfigured with Secure Sockets Layer (SSL) encryption and Lightweight Directory Access Protocol (LDAP) authentication.

The SASEFRED interface engine is the best-kept secret in SAS/ETS^® software. It dramatically reduces the amount of time and effort required to include economic indicator variables in your time series analysis. Using the SASEFRED engine in SAS^® Enterprise Guide^®, I can directly query the economic database of the Federal Reserve Bank of St. Louis. This public database contains over 529,000 economic time series aggregated from 86 sources. In this paper, I forecast wine demand and enrich my predictions via the inclusion of economic variables such as Retail Sales: Beer, Wine, and Liquor Stores and Producer Price Index by Industry: Beer, Wine, and Liquor Stores. Although I use a retail example, this technology is relevant to all industries. The diversity of economic variables provided in this database ensures that it is useful to virtually every time series analysis and industry. This specific example leverages SAS Enterprise Guide and SAS^® Forecast Server as interfaces. However, this functionality works on SAS^® 9.4 as well as on SAS^® Viya^® technology.

We are frequently asked by the customer if we can use the local Indian currency symbol in their SAS^® Visual Analytics dashboards. Customers want the Locale Indian currency symbol (Rupee ₹) displayed behind their data so that they can locate where the exact values are. Unfortunately, by default SAS^® has limitations on displaying currency symbols based on locales that are currently available in SAS. This paper discusses the problems and limitations of the current features and provides an example of how this new currency symbol can be added in SAS Visual Analytics.

SAS has partnered with expert Game Developer HEC Montreal (https://erpsim.hec.ca/) to create an Analytics Simulation Game. Much like an airline flight simulator, the objective of this game is to provide an active learning opportunity for aspiring data scientists to learn and apply analytics. Using SAS^® Enterprise Miner (TM), students will work with a special simulated data set, derived from US Census Data and other sources, that features a million members of a major charity. Students are placed in the role of a data scientist for this charity, and they must unlock the value in members data to target which individuals should be called to increase the funds raised for the charity. Students first predict the amount each individual might give, but in a second round, they predict the uplift from contacting them instead. Each student submits their decisions to a special, live web-based scoreboard, where they can see how much they raised with their models and how they compare with other classmates. Students can use SAS Enterprise Miner via an Amazon Web Services (AWS) cloud instance, on their own PC, or in laboratories. This SAS^® Global Forum marks the official release of the game, in time for professors and educators to use it in their Spring or Fall 2019 curriculum.

SAS^® has exceptional analytics capabilities, but to process data we often need to extract, transform, validate, and correct the data that we get from various sources to make best use of its capabilities. Suppose in an application, we process customer data where we get information from data entry monthly, with multiple records having data entry errors. Periodically, we need to identify and correct those entries in the final SAS data set as part of the data validation and correction process. It is time-consuming to manually update each record monthly. Therefore, the need for an automated process arises to produce a final corrected data set. This paper demonstrates how we can update only the incorrect values in a SAS data set by using the external file, which provides only the corrected values (finder file). This process does not make any data merges or SQL joins for the data correction. The process uses the FORMAT procedure and creates the customized formats using CNTLIN for the finder file. PROC FORMAT creates the variable to be corrected and a unique master key having several variables concatenated to avoid errors in the correction process. Using this format, code corrects the invalid values in the variable and all remaining variables remain the same. This paper is intended for intermediate-level SAS developers who want to build data validation and data correction programs using SAS.

JMP^® has a robust set of functions that facilitate data cleaning and transformation. This discussion focuses on using the Match function in a formula to create new columns. Branched survey question responses from the MyLymeData patient registry are used to demonstrate the power of the Match function to quickly and accurately create a Likert scale describing patient reported health status following antibiotic treatment and to further create subgroups based on their responses.\

Starting in SAS^® 9.3, the R interface enables SAS users on Windows and Linux who license SAS/IML^® software to call R functions and transfer data between SAS and R from within SAS. Potential users include SAS/IML users and other SAS users who can use PROC IML just as a wrapper to transfer data between SAS and R and call R functions. This paper provides a basic introduction and some simple examples. The focus is on SAS users who are not PROC IML users, but who want to take advantage of the R interface.

Have you ever wondered why the log file is one of the first items that SAS Technical Support requests? The reason is that the log file contains a wealth of information about what actually occurs in the application or platform. So, what exactly is Technical Support looking for? How can the SAS^® Object Spawner log help pinpoint the underlying issue? By reading this paper, SAS administrators will gain insight into how to use the object-spawner log to identify issues more quickly and avoid downtime. The focus of this paper is on the object-spawner log, because the object spawner is the gateway to other application servers. The paper provides an overview of the object-spawner log, describes useful loggers that you should be aware of, details how to identify common issues by looking in the logs, and offers solutions for those issues. With this knowledge, you can be proactive in solving problems that arise.

Python is one of the most popular programming languages in recent years. It is now getting used in machine learning and AI. The big advantage of Python is that it is considered a one-stop shop that performs several functions with one programming language. Although SAS^® is the most powerful analytical tool in a clinical study, Python expands the reporting activity of SAS and can potentially advance a SAS programmers career. SASPy is the new ticket - it is Python for SAS programmers. It is also known as the Python library in SAS and enables you to start a SAS session in Python. This means Python can support activities from SAS data handling to reporting. This presentation describes the basic usage of SASPy and introduces tips for handling SAS data sets. It discusses possible reporting activities using Python in a clinical study.

Hospital readmission is an adverse but often preventable event that has been shown to be related to a hospitals quality of care (Frankl et al., 1991). To reduce unplanned 30-day readmission rates among Medicare beneficiaries, the Centers for Medicare Medicaid Services introduced measures of risk-standardized readmission rates (RSRRs) in both its quality reporting and pay for performance programs. RSRRs are calculated for five medical conditions, two procedures, and one hospital-wide measure for participating hospitals. These results are publicly reported on Hospital Compare, an online tool that patients can use to guide their decisions about where to seek care. Radar plots provide a tool that can illuminate differences and similarities of rates within and across hospitals. This can be useful for patients seeking the best possible care, or for hospital quality departments trying to understand their hospitals performance compared to their peers. This technique can be implemented using the POLYGON statement in the SGPANEL procedure (Hebbar, 2013). The code can be modified to add grids, tick marks, and labels, which provide more information on the estimates. Plotting the hospitals RSRRs against the observed national rate enables one to quickly see which hospitals perform worse or better than the national average and which perform similarly in certain measures. In summary, radar plots are an effective way to display multiple hospital RSRRs at once and make quick comparisons.

Health-care professionals who successfully turn data assets into data insights can quickly realize benefits such as reduced costs, healthier patients, and higher satisfaction rates. Some insights are discovered by comparing your data to national benchmarks, and others are found in your own data. Using SAS^® Visual Analytics with various publicly available health-care data resources, we demonstrate effective techniques for trending health plan quality performance metrics, highlighting geographical variations in care, and assisting the users in making inferences about their data. Visualizing data enables decision makers and data analysts to uncover new associations, engage users, and communicate their messages successfully.

In a perfect world of reporting, all code is automated and successfully invoked without manual intervention or ever touching the SAS^® production code. Unfortunately, this is not the case for all reports across a company. This paper demonstrates how to efficiently add and remove recipients on a SAS report distribution list without manipulating the SAS production code. To do this, my coworker and I developed an email distribution list macro along with a macro to send the report. Together, these two macros provide a solution to efficient reporting and can be used to create a report inventory. Using this macro enables users to get away from altering SAS production code to simply add or remove users who receive a report, which is a best practice for most companies. We demonstrate how to develop and use these macros efficiently.

An internal graphics challenge between SAS^® and R programmers at the Mayo Clinic led to the creation of a circos plot with SAS^® 9.4M3 features. A circos plot is a circular visualization of the relationship between objects, positions, and other time-point-related data. The circos plot features a curved rectangle for each node or object around the perimeter of the circle and a Bezier curve for each path between nodes that is pulled toward the center of the circle. Each rectangles size and curves size is determined by its proportion of the total population. The primary challenge of creating this plot with the current SAS^® Statistical Graphics procedures is that these procedures do not include polar axes to simplify plotting circular or curved objects. The macro %CIRCOS is an example of overcoming these limitations creatively using trigonometry to prove that these types of graphs are still possible without polar axes.

CASL This is an introductory paper to help you understand why one would want to adopt the programming language CASL. SAS^® Cloud Analytic Services language (CASL) is a language specification used by both SAS^® clients and other clients to interact with SAS^® Cloud Analytic Services (CAS). CASL is a statement-based scripting language that supports the entire analytical lifecycle (data management, analytics, and scoring), as well as the monitoring of the CAS server. The strength of this language can be seen in how easy it is to initialize parameters to CAS actions and to manipulate the results of these actions. The results can then be used to prepare the parameters for the execution of another action. The goal is to provide an environment that enables a user to string together multiple actions to accomplish a specific result. CASL provides access to all of your favorite features in SAS, including Base SAS^® in-database procedures, the DATA step, DS2, FedSQL, formats, macros, and the Output Delivery System (ODS). CASL is ideal for splicing together access to the CAS server between the use of other SAS procedures from a SAS client or from another client, or simply to create your own custom application.

Report localization was first introduced in SAS^® Visual Analytics Designer 7.1. This feature made it easy for a user or a language specialist to translate a single report into multiple languages for report users. Many SAS^® sample reports and customer reports have been translated using the 7.x releases of SAS Visual Analytics Designer. However, a user had to have a report-designing role to translate report content, which meant that the design of a report could be altered during localization. SAS Visual Analytics 8.3 is the first release based on SAS^® Viya^® that lets you localize reports. Starting in the 8.3 release, report content can be translated without the possibility of altering your report design. There is a new set of command-line interfaces (CLIs) in SAS Viya. The SAS reports CLI enables you to localize report content without requiring the report to be open for editing in SAS Visual Analytics. In this presentation, we introduce the SAS report CLI features and demonstrate how you can use them to localize your reports.

Join us as we explore new features and functionality in the SAS^® Function Compiler (FCMP). Integration with Python, support for running analytic scoring containers (ASTORE), and a new FCMP action set are the main topics we cover. Learn how to leverage your existing investment in Python by calling Python functions from an FCMP function. Get the most from your ASTORE by porting it from SAS^® Viya^® to SAS^® 9.4 TS1M6, and then run it from an FCMP function called from the SAS^® DATA step. Learn how to port your favorite user-written functions and subroutines to SAS^® Cloud Analytic Services, and then use them within a computed column or another action. This paper shows you the tips and tricks you need to integrate your existing FCMP code with these new SAS technologies. Included are several examples to quickly get you started.

There are times when it is convenient to be able to compare two file system areas that are supposed to be equivalent. This could be during a migration activity involving changing servers or SAS^® versions. This paper covers techniques that can help facilitate the comparison of programs, logs, Microsoft Excel files, and SAS data sets, as well as techniques for digesting the results.

The latest releases of SAS^® Data Management provide a comprehensive and integrated set of capabilities for collecting, transforming, managing, and governing your data. The latest features in the product suite include capabilities for working with data from a wide variety of environments and types including Apache Hadoop, cloud data sources, relational database management system (RDBMS) files, unstructured data, and streaming. You also have the ability to perform extract, transform, load (ETL) and extract, load, transform (ELT) processes in diverse run-time environments such as SAS^®, Hadoop, Apache Spark, SAS^® Analytics, cloud, and data virtualization environments. There are also new artificial intelligence features that can provide insight into your data and help simplify the steps needed to prepare data for analytics. This paper provides an overview of the latest features of the SAS Data Management product suite and includes use cases and examples for leveraging product capabilities.

This paper demonstrates the challenges you might face and their solutions when you use SAS^® to process large data sets. First, we demonstrate how to use SAS system options to program query efficiency. Next, we tune SAS programs to improve big data performance, modify SQL queries to maximize implicit pass-through, and re-architect processes to improve performance. Along the way, we identify situations in which data precision might be degraded, and then leverage FedSQL and DS2 to conduct full-precision calculations on a wide variety of ANSI data types. Finally, we look at boosting speed by using parallel processing techniques. We begin by using DS2 in Base SAS^® to speed up CPU bound processes, and then kick it up a notch with the SAS^® In-Database Code Accelerator for in-database processing on massively parallel processing (MPP) platforms like Teradata and Apache Hadoop. And for the ultimate speed boost, we convert our programs to use distributed processing and process memory-resident data in SAS^® Viya^® and SAS^® Cloud Analytic Services (CAS).

In 2018, Ernestynne delivered one of the top-rated presentations at the SAS^® Users New Zealand conference. The presentation talked about a future of working with both open-source and proprietary software. It covered the benefits and provided some examples of combining both. The future is now here. This 90-minute Hands-On Workshop drills down into examples that were briefly touched on during that presentation. Attendees get a taste of working with R in SAS^® Enterprise MinerTM and some general best practice tips. There is an introduction to the Scripting Wrapper for Analytics Transfer (SWAT) package, which enables you to interface with SAS^® Viya^® in open-source languages like R. This presentation briefly covers avoiding some traps that you can fall into when working with the SWAT package. The techniques learned here can be extended to enable you to work with Python in SAS^®. The emphasis of this Hands-On Workshop is how to combine SAS with open-source tools, rather than teaching attendees how to code in R.

SAS^® procedures can convey an enormous amount of information sometimes more information than is needed. Most SAS procedures generate ODS objects behind the scenes. SAS uses these objects with style templates that have custom buckets for certain types of output to produce the output that we see in all destinations (including the SAS listing). By tracing output objects and ODS templates using ODS TRACE (DOM) and by manipulating procedural output and ODS OUTPUT objects, we can pick and choose just the information that we want to see. We can then harness the power of SAS data management and reporting procedures to coalesce the information collected and present the information accurately and attractively.

Have you ever found a table of data on a web page and wanted to scrape that data into a SAS^® data set? This paper shows you how to do it in seconds using SAS^® Graphics Accelerator for Google Chrome, which is a free extension for Google Chrome. First, you open the web page in Chrome. Then, you extract data from the table into your Laboratory within SAS Graphics Accelerator for Google Chrome. SAS Graphics Accelerator for Google Chrome automatically derives variable labels from column headings and variable lengths from observations. It also derives variable types if all observations in a column match patterns for a single type. If they don't, you can choose the type of a variable manually. Finally, you generate a SAS program that creates a data set in your WORK library with correct variable labels, lengths, and formats. With SAS Graphics Accelerator for Google Chrome, you'll be the fastest data wrangler on the web!

SAS^® Visual Forecasting, the new-generation forecasting product from SAS, includes a web-based user interface for creating and running projects that generate forecasts from historical data. It is designed to use the highly parallel and distributed architecture of SAS^® Viya^®, a cloud-enabled, in-memory analytics engine that is powered by SAS^® Cloud Analytic Services (CAS), to effectively model and forecast time series on a large scale. SAS Visual Forecasting includes several built-in modeling strategies, which serve as ready-to-use models for generating forecasts. It also supports custom modeling nodes, where you can write and import your own code-based modeling strategies. Similar to the ready-to-use models, these custom models can also be shared with other projects and forecasters. Forecasters can use SAS Visual Forecasting to create projects by using visual flow diagrams (called pipelines), running multiple built-in or code-based models on the same data, and choosing a champion model based on the results. This paper uses a gradient boosting tree model as an example to demonstrate how you can use a custom modeling node in SAS Visual Forecasting to develop and implement your own modeling strategy.

This paper discusses three different types of storage services available from Amazon Web Services (AWS)Amazon S3, Amazon Aurora, and Amazon Redshift and how SAS^® can access data from each type of storage service for analytical purposes. Amazon S3 stores data as objects, such as files. You can access these files by using the S3 procedure in SAS. Amazon Aurora is a relational database that is part of Amazon Relational Database Service. The engine for Aurora is compatible with both MySQL or Postgres. Depending on whether the Aurora database is based on MySQL or Postgres, you can use SAS/ACCESS^® Interface to MySQL or SAS/ACCESS^® Interface to Postgres to access the data in Aurora. Amazon Redshift is a fully managed, scalable data warehouse in the cloud. You can use SAS/ACCESS^® Interface to Amazon Redshift to access data stored in Amazon Redshift.

SAS/STAT^® and SAS/ETS^® software has several procedures for working with count data based on the Poisson distribution or the quadratic negative binomial distribution (NB-2). Count data might have either an excess number of zeros (inflation) or the situation where zero is not a valid outcome (truncation). The zero-inflated Poisson and negative binomial models are available with the GENMOD and FMM procedures. The FMM procedure also provides the zero-truncated Poisson and negative binomial distributions. Other types of count data models such as the restricted and unrestricted generalized Poisson, linear negative binomial (NB-1), and Poisson-Inverse Gaussian (PIG) distributions can also serve as choices to model count data and likewise are subject to zero-inflation or truncation. Programming statements entered into the NLMIXED procedure in SAS/STAT can model both zero-inflated and zero-truncated count data with other types of distributions, and as a result improve model fit.