WHAT S NEW IN SAS INCLUDING BASE, STAT, SAS ENTERPRISE GUIDE

Transcription

1 WHAT S NEW IN SAS INCLUDING BASE, STAT, SAS ENTERPRISE GUIDE

2 AGENDA WHAT S NEW Base SAS/Stat SAS Enterprise Guide 2

3 SAS 9 WHAT S NEW THEME Productivity enhancements Improved graphics More powerful algorithms & flexible models

4 Blog with program to reproduce graph

5 KEEP IN MIND MOVING BETWEEN VERSIONS, OS, OR 32-BIT TO 64-BIT SAS Programs Should run without any issues Please run and compare a representative group SAS Output ODS may cause output differences such as page breaks and fewer or more pages Printer Destinations may need to be changed SAS Procedures Behavior may have changed between releases 31o1n.htm SAS Datasets Windows allows you to use datasets interchangeably without using CEDA Other OS will require CEDA (Cross-Environment Data Access) Compatibility Calculator

6 WHAT S NEW IN BASE SAS 9.2

7 SAS BASE ENHANCEMENTS BASE procedures Checkpoint/Restart & Code Analyzer ODS and reporting Graphs

8 SAS 9.2 BASE PROCEDURES Proc SORT Linguistic collation (NLS/ICU) language aware ordering SORTSEQ=<encoding> PRESORTED option Proc MIGRATE cross platform support PROC SUMMARY/MEANS MODE statistic New formats and functions UnPUT Track execution within a DO group

9 SAS SAS CODE ANALYZER SAS Code Analyzer PROC SCAPROC Generates metadata about SAS job Captures information about the job step, I/O information such as file dependencies and macros Automatic grid-enabled job creation (with SAS Grid Manager in 9.2 Phase 2)

10 SAS 9.2 CHECKPOINT/RESTART Checkpoint/Restart STEPCHKPT - runtime utility to record information about DATA and PROC steps in a checkpoint library Program execution resumes with the step that did not complete when the failure occurred Global statements (filename) and macros are re-executed

11 SAS GRAPH ENHANCEMENTS ODS Statistical Graphics SGPlot, SGScatter, SGPanel Graphics Template Language Statistical Graphics Editor (SGEditor) New default styles True color support Network diagrams Truetype fonts

12 SAS GRAPH ENHANCEMENTS SGEDITOR

13 SAS GRAPH ENHANCEMENTS Proc template; define statgraph myplot; layout overlay; entrytitle "Weight by Height"; Modelband "regcli"/ display=(outline); Modelband regclm ; Regressionplot y=weight x=height/ clm= regclm cli= regcli ; scatterplot y=weight x=height /primary=yes; endlayout; end; run; GRAPH TEMPLATE LANGUAGE

14 SAS IMPROVED GRAPHICS ods graphics on; ods html; proc reg data=sashelp.class; model Weight = Height; quit; ods html close; ods graphics off;

15 SAS NETWORK DIAGRAMS Hierarchical Circular Hexagonal Fixed Position Multi-Level Force

16 SAS PNG OUTPUT GIF transparency embedded in webpage PNG transparency embedded in webpage

17 SAS GRAPH ENHANCEMENTS TRUETYPE FONTS

18 SAS 9.2 : Enhanced Base Procedures PROC SUMMARY/MEANS MODE proc means data=tmp1.demo mean median mode var age; run; Also available in PROC TABULATE, PROC REPORT

19 SAS 9.2 : ENHANCED BASE PROCEDURES PROC FREQ TABLE statement options CROSSLIST proc freq data=demog; table gender*age/crosslist; run; NLEVELS proc freq data=demog nlevels; table gender*age; run; CONTENTS= ods html contents='c:\temp\whatnewc.html'; proc freq data=demog; table sex*age/contents='demography Data for Age'; table sex*weight/contents='demography Data for Weight'; run; ods html close;

20 SAS 9.2 : ENHANCED BASE PROCEDURES PROC FREQ WEIGHT statement options ZEROS New plots produced: Frequency Deviation Odds ratio kappa Supports ODS graphics ods graphics on; proc freq data=demog; table gender*age/plots=freqplot; run; ods graphics off;

21 SAS 9.2 : ENHANCED BASE PROCEDURES PROC UNIVARIATE CDFPLOT Observed Cumulative Distribution Function PPPLOT Probability Probability Plot Both Plots Support: Beta Exponential Gamma Lognormal Normal Weibull Supports ODS graphics ods graphics on; proc univariate data=aircraft; var deviation; ppplot /normal; run; ods graphics off;

22 WHAT S NEW IN BASE SAS 9.3

23 WHAT S NEW IN BASE SAS 9.3 Key Feature Review SAS Windowing Environment defaults to HTML output and ODS Graphics With the default destination in the SAS windowing environment now being HTML (rather than LISTING), graphs are automatically integrated with tables in the same HTML file ODS Graphics is enabled by default with the new HTMLBLUE style as a default setting New LISTING, HTML, PCL, and PRINTER destinations support Scalable Vector Graphics (SVG) User control to change back to the pre-9.3 behavior exists

24 WHAT S NEW IN BASE SAS 9.3 Key Feature Review PROC GROOVY enables SAS code to execute Groovy code on the Java Virtual Machine. Groovy statements that are written as part of your SAS code Run statements that are in files that you specify with PROC GROOVY commands Parse Groovy statements into Groovy Class objects, and run these objects or make them available to other PROC GROOVY statements or Java DATA Step Objects.

25 PROCEDURES PROC FORMAT LOCALE= option to create a format catalog for current locale User-defined format for a missing value now overrides MISSING option Max labels for MULTILABEL is now 255 Picture statement directives %n formats the number of days in a duration %s formats fractional seconds %z formats a UTC time-zone offset %Z formats a time-zone name Use SAS Explorer to view format in informat definitions

26 PROCEDURES PROC PRINT Completely integrated with ODS Separate table for each BY group Observation counts reset to zero for each BY group BYLINES up to 512 characters If PRINT procedure errors or terminates, output might be produced

27 SAS BASE STATISTICAL PROCEDURES PROC CORR CORR POLYSERIAL option added

28 SAS BASE STATISTICAL PROCEDURES PROC FREQ Produces agreement plots Provides exact unconditional confidences lines for relative risk and the risk difference Frequency plots and cumulative frequency plots are no longer produced by default when ODS Graphics is enabled. You can request these plots with the PLOTS=FREQPLOT and PLOTS=CUMFREQPLOT options in the TABLES statement.

29 PROC FREQ EXAMPLE

30 SAS BASE STATISTICAL PROCEDURES PROC UNIVARIATE Support 5 new fitted distributions Gumbel Inverse Gaussian Generalized Pareto Power Function Rayleigh Available in the CDFPLOT, HISTOGRAM, PROBPLOT, PPPLOT AND QQPLOT statements

31 PROC UNIVARIATE EXAMPLE

32 SAS BASE PROCEDURES PROC TABULATE NOCELLMERGE specifies that data cells are not merged with other data cells in the table Does not work with the traditional monospace output

33 NEW STATEMENTS LIBNAME JMP Associates a libref with a JMP data table for read/write RESETLINE Restarts the program line numbering in the SAS log to 1

34 SELECTED SYSTEM OPTIONS DLCREATEDIR Create a directory that is named in a LIBNAME statement ORIENTATION=LANDSCAPE PORTRAIT Specifies page orientation for ODS documents AUTOCORRECT NOAUTOCORRECT Turns on/off autocorrection of SAS names

35 NEW MACRO FUNCTIONS %SYSMACEXEC Indicates whether a macro is currently executing %SYSMACEXIST Indicates whether macro is defined in WORK.SASMACR %SYSMEXECDEPTH Returns the depth of macro nesting %SYSMEXECNAME Returns the name of the macro executing at a nesting level

36 NEW MACRO STATEMENTS %SYSMSTORECLEAR Closes stored compiled macros and clears SASMSTORE= %SYSMACDELETE Deletes a macro definition from WORK.SASMACR

37 BASE SAS 9.4

38 SAS 9.4 NEW BASE PROCEDURES PROC AUTHLIB PROC DELETE PROC DS2 PROC FEDSQL PROC JSON PROC PRESENV PROC STREAM SAS 9.4 Documentation

39 SAS 9.4 BASE PROCEDURES WITH ENHANCEMENTS PROC APPEND PROC CONTENTS PROC COPY PROC DATASETS PROC EXPORT PROC FCMP PROC IMPORT PROC MIGRATE PROC QDEVICE PROC XSL PROC PWENCODE PROC SORT PROC HADOOP SAS 9.4 Documentation

40 BASE SAS PROGRAMMING LANGUAGES Two new programming languages: DS2 FedSQL

41 BASE SAS NEW PROGRAMMING LANGUAGE DS2 is a new SAS proprietary programming language. It Enables DS2 language statements from Base SAS Includes additional data types, ANSI SQL types, programming structure elements, and user-defined methods and packages. Allows embedded FedSQL in some statements Runs anywhere Base, In-Database (via In-Database Code Accelerator), High Performance Analytics (via HPDS2)

42 DS2 WHY USE IT? DS2 is beneficial in applications that Need to use the new DS2 data types or programming techniques/structures. Can utilize parallel processing (for computationally intense processes). Require large data movement.

43 DS2 WHY USE IT? Data Source DS2 brings the code to the data, eliminating the need to move large amounts of data. SAS code

44 DS2 SAMPLE SYNTAX Similar to the Data step, but different.. proc ds2; data _null_; method init(); dcl varchar(16) str; str = 'Hello World!'; put str; end; enddata; run; quit;

45 PROC DS2; data _null_; method init(); dcl varchar(20) foo; foo = '**> Starting'; put foo; end; method run(); set ds2_sas.banks; put _all_; end; method term(); dcl char(11) bar; bar = '**> I quit!'; put bar; end; run; quit; Initial processing Final processing BASIC DS2 SYNTAX Execution loop

46 BASE SAS PROGRAMMING LANGUAGES FedSQL is a SAS proprietary implementation of the ANSI SQL:1999 core standard. It provides A scalable, threaded, high-performance way to access, manage, and share relational data in multiple data sources A common SQL syntax across all data sources Support for new data types

47 OUTPUT OUTPUT DELIVERY SYSTEM Enhancements to the Output Delivery System enable you to Use the ODS Report Writing Interface (RWI) to create and manipulate predefined ODS objects in a DATA step to create highly customized output. Create EPUB, HTML5, and Microsoft PowerPoint files. Arrange ODS output objects exactly where you want them on a page, or use dynamic placement of objects by using a grid structure. (ODS Layout) Animate multi-page GIF images and SVG files by setting system options.

48 SAS ODS POWERPOINT

49 SAS ODS POWERPOINT TwoContent layout

50 SAS ODS EPUB Output SAS Reports to ebook format to be viewed on ipad & iphone

51 EPUB DESTINATION EXAMPLE ods epub file="glm.epub" title="my First ODS EPUB E-book" options(creator="sas Programmer" description="my First ODS EPUB Book" subject="proc GLM" type="ods EPUB book"); ods graphics on; proc glm data=drugtest; class Drug; model PostTreatment = Drug PreTreatment; run; quit; ods epub close; SAS Global Forum Paper of epub Destination

52 EPUB DESTINATION SAMPLE OUTPUT

53 AXISTABLE STATEMENT AXIS ALIGNED TEXT VALUES

54 ENHANCED GRAPHS

55 OTHER NEW FEATURES The Work library data sets and catalogs, and the values of global statements, macro variables, and system options can be preserved between SAS sessions. The PRESENV Procedure The PRESENV procedure preserves all global statements and macro variables in your SAS code from one SAS session to another.

56 OTHER NEW FEATURES The SAS language now supports time zones based on Universal Coordinate Time (UTC). Data sets and catalog time stamps can specify the time based on a specific time zone. SAS can also determine the time for an area, taking into account Daylight Savings Time. There are some new functions to support UTC: TZONEID-returns the current time zone ID. TZONENAME-returns the current standard or daylight savings time and the time zone name. TZONEOFF-returns the user time zone offset. TZONES2U-converts a SAS datetime value to a UTC datetime value. TZONEU2S-converts a UTC datetime value to a SAS datetime value. There is also a TIMEZONE= SAS system option.

57 WHAT S NEW IN SAS /STAT FOR SAS 9.2

58 KEY BENEFITS FOR THE STATISTICIAN New and improved analytical methods Statistical Modeling Bayesian Analysis Survey Data Analysis Easy-to-use production quality graphics Many enhancements to existing procedures

59 ODS STATISTICAL GRAPHICS NOW PRODUCTION! Automates creation of commonly used displays in statistical procedures Experimental in SAS 9.1 Production in SAS 9.2 for over 50 procedures Functionality is being used to provide modern statistical graphics in a variety of use cases convenient specialized graphics in statistical procedure output standalone graphics for preliminary analysis of data customized graphics (using the underlying Graph Template Language) NOTE: A SAS/GRAPH license will be required in order to use the ODS graphics functionality in SAS/STAT for 9.2.

60 SAS/STAT ANOVA BOXPLOT CALIS CLUSTER CORRESP FACTOR GAM GENMOD GLM GLIMMIX GLMSELECT PROCEDURES SUPPORTING ODS GRAPHICS KDE LIFEREG LIFETEST LOESS LOGISTIC MDS MI MIXED MULTTEST NPAR1WAY PHREG PLS PRINCOMP PRINQUAL QUANTREG REG ROBUSTREG RSREG TRANSREG TTEST

61 BASICS OF ODS GRAPHICS Invoke with ODS GRAPHICS ON; Procedure options determine which plot I get Destination determines where I see my plots ODS HTML STYLE=JOURNAL; Style determines what all my plots look like

62 SIMPLE LINEAR REGRESSION IS REALLY SIMPLE ods graphics on; proc reg; model Weight = Height; quit;

63 THE GLIMMIX PROCEDURE The new GLIMMIX procedure fits generalized linear mixed models, based on linearization. For data with correlations or nonconstant variability Response is not necessarily normally distributed Assumes normal (Gaussian) random effects Conditional on these random effects, data can have any distribution in the exponential family Default estimation is restricted pseudo-likelihood for models with random effects

64 THE QUANTREG PROCEDURE Quantile regression extends the regression model to conditional quantiles of the response variable, such as the median or the 90th percentile. Makes no distributional assumption or link assumption Robust to outliers Robust to heterogeneity Used as an exploratory tool / extreme value analysis Equivariant to monotone transformations

65 MODEL SELECTION WITH PROC GLMSELECT Provides great versatility in model selection: Multiple selection methods Cross-validation Graphical aids Crossed and nested effects Classification effects Selection from thousands of effects

66 NEW MCMC PROCEDURE PROC MCMC Markov Chain Monte Carlo simulations Flexible simulation-based procedure that is suitable for fitting a wide range of Bayesian models Uses a random-walk Metropolis algorithm Model specification similar to PROC NLIN, and syntax similar to PROC NLMIXED.

67 POSTFITTING BONANZA New shared architecture means more postfitting inference capabilities for many linear modeling procedures. EFFECTPLOT ESTIMATE LSMEANS LSMESTIMATE SLICE TEST GENMOD, LOGISTIC, ORTHOREG LOGISTIC, ORTHOREG, PHREG, SURVEYLOGISTIC LOGISTIC, ORTHOREG, PHREG, SURVEYLOGISTIC, SUREVYREG GENMOD, LOGISTIC, MIXED, ORTHOREG, PHREG, SURVEYLOGISTIC, SURVEYREG GENMOD, GLIMMIX, LOGISTIC, MIXED, ORTHOREG, SURVEYLOGISTIC, SURVEYREG ORTHOREG, SURVEYREG

68 NEW PLM PROCEDURE The new PLM procedure performs postfitting inference on model fit information stored by ten modeling procedures: STORE statement saves model info as a SAS item store PROC PLM provides the full complement of postfitting inference statements plus EFFECTPLOT. Saves computation time and is a tool for confidentiality governance See Introducing PROC PLM and Postfitting Analyses in Very General Linear Models by Randy Tobias and Weijie Cai in the 2010 SGF proceedings.

69 OTHER NEW PROCEDURES Proc SURVEYPHREG fits the Cox model for proportional hazards to sample survey data. Proc TCALIS Same analysis as CALIS plus multiplegroup analysis, enhanced mean structure analysis, pathlike model specification, support of LISREL-type models, and more. Proc HPMIXED - Fits linear mixed models using sparse matrix techniques

70 HIGHLIGHTS: TTEST ODS Graphics univariate and comparative versions of histograms, box plots, and densities SIDES= option for one-sided versions of tests and confidence intervals Equivalence tests Analysis of simple crossover designs

71 Add cumulative probabilities to SCORE output Inverse link added to CONTRAST statement ROC statement for ROC comparisons Odds ratio plots and other plot enhancements Fuzzy comparisons in exact computations HIGHLIGHTS: LOGISTIC

72 HIGHLIGHTS: LIFETEST Proc LIFETEST SURVIVAL statement enables the creation of confidence bands for the survivor function S(t) Number of subjects at risk can be displayed for Kaplan Meier survival curves Smoother hazard function using the kernel method can be specified

73 HIGHLIGHTS: ANALYSES COVERAGE Thus, you can now easily obtain Bayesian results for the following analyses in SAS: Regression Poisson regression Logistic regression Loglinear models Accelerated failure time models Proportional hazards models

74 WHAT S NEW IN SAS /STAT FOR SAS 9.3

75 BUT FIRST... With the 12.1 release of SAS/STAT as part of SAS 9.3, we turn to a new releasenumbering scheme NEW NUMBERING FOR ANALYTICAL PRODUCTS SAS/STAT released independently of Base SAS All analytical products follow new numbering scheme SAS/STAT released every months

76 SAS/STAT 9.3 KEY FEATURES Key Features New FMM procedure (experimental) for finite mixture modeling Shared frailty models in PHREG procedure Expanded capabilities for Bayesian analysis Production SURVEYPHREG procedure New model diagnostics in NLIN procedure

77 FINITE MIXTURE MODELING PROC FMM fits statistical models to data where the distribution of the response is a finite mixture of univariate distributions. Performs maximum likelihood estimation for all models Provides Bayesian analysis for several models. Useful for applications such as estimating multimodal or heavytailed densities modeling overdispersed data.

78 Shared Frailty Models in PHREG When experimental units are clustered, the failure times of those units within a cluster tend to be correlated. You can account for within-cluster correlation by using a shared frailty model. The effects of clustering are incorporated in the model as normally distributed random variables. The PROC PHREG procedure uses a penalized partial likelihood approach to fit the shared frailty model.

79 MCMC PROCEDURE PROC MCMC is a general purpose Markov chain Monte Carlo simulation procedure that fits Bayesian models. The new RANDOM statement facilitate fitting Bayesian models with random effects: Simplifies specification of hierarchical random-effects models Significantly reduces simulation time while improving convergence Allows linear or nonlinear random effects that can be drawn from univariate or multivariate distributions.

80 NLIN PROCEDURE PROC NLIN fits nonlinear regression models New experimental features diagnose the model fit: Plots of the fitted model, diagnostics, tangential and Jacobian leverage, and local influence Global measures of nonlinearity Box s bias statistics

81 MI PROCEDURE PROC MI creates multiply-imputed data sets for incomplete multivariate data. The experimental FCS statement specifies a multivariate imputation by fully conditional specification (FCS) methods. For data with an arbitrary missing data pattern, the FCS methods enable you to impute missing values for all variables, assuming that a joint distribution for these variables exists. The FCS method requires fewer iterations than the MCMC method.

82 Production SURVEYPHREG Procedure PROC SURVEYPHREG provides survival analysis in the form of Cox proportional hazards regression for sample survey data. Production with SAS/STAT 9.3 Handles time-dependent covariates Provides EFFECT statement

83 KEY FEATURES OF SAS/STAT 12.1 Standardized rates and proportions Quantile regression: model selection Quantile selection: censored data Multivariate adaptive regression splines Updates to Bayesian software Additional highlights

84 STDRATE PROCEDURE The STDRATE procedure computes direct and indirect standardized rates and risks for study populations. Compares directly standardized rates or risks for two study populations with the same reference population Computes Mantel-Haenszel effect estimates, such as the rate difference, for two study populations without a reference population

85 QUANTSELECT PROCEDURE The QUANTSELECT procedure performs model selection for quantile regression. Forward, backward, stepwise, and LASSO selection methods Variable selection criteria: AIC, SBC, AICC, and so on Variable selection for both quantiles and the quantile process EFFECT statement for constructed model effects (splines) Multithreaded Experimental in SAS/STAT 12.1

86 MULTIVARIATE ADAPTIVE REGRESSION SPLINES The multivariate adaptive regression splines method is a nonparametric approach for modeling high-dimensional data. It combines both regression splines and model selection methods: Parametric forms not required Constructs regression spline terms without pre-specified knots Original publication by Friedman (1991)

87 BAYESIAN UPDATES BAYESIAN CAPABILITIES ARE AVAILABLE THROUGH TWO CHANNELS THE BAYES STATEMENT IN SELECTED PROCEDURES AND A GENERAL MODELING PARADIGM WITH THE MCMC PROCEDURE. IN SAS/STAT 12.1: PHREG procedure supports Bayesian frailty models Gamerman algorithm is the default sampling mechanism in the GENMOD procedure.

88 ADDITIONAL HIGHLIGHTS PROC FMM becomes production with SAS/STAT 12.1 and adds the truncated normal and truncated negative binomial distributions. Partial R-square for relative importance of parameters in PROC LOGISTIC Miettinen-Nurminen confidence limits for the difference of proportions in PROC FREQ Poisson sampling in PROC SURVEYSELECT Post-stratification estimation in the SURVEYMEANS procedure REF= option added to the CLASS statement for GLM, MIXED, GLIMMIX, and ORTHOREG procedures Heat map and mosaic plots

89 MOSAIC PLOT In PROC FREQ plots=mosaic plot

90 HEAT MAP PROC REG Heat maps automatically created when number of points exceed the default

91 WHAT S NEW IN SAS /STAT FOR SAS 9.4

92 SAS 9.4 SAS HIGH PERFORMANCE PROCEDURES High Performance Statistics High Performance Data Mining High Performance Econometrics High Performance Forecasting High Performance Optimization High Performance Text Mining HPLOGISTIC HPGENSELECT HPREG HPLMIXED HPNLMOD HPSPLIT HPFMM HPCANDISC HPPRINCOMP HPPLS HPQUANTSELECT HPREDUCE HPNEURAL HPFOREST HP4SCORE HPDECIDE HPCLUS HPSVM HPBNET HPTSDR HPCOUNTREG HPSEVERITY HPQLIM HPPANEL HPCOPULA HPCDM HPFORECAST OPTLSO Select features in OPTMILP OPTLP OPTMODEL OPTGRAPH HPCDM HPTMINE HPTMSCORE Common Set: HPDS2, HPDMDB, HPSAMPLE, HPSUMMARY, HPIMPUTE, HPBIN, HPCORR SAS/STAT 13.2 Documentation SAS/Stat High-Performance Procedure Documentation

93 HP PROCEDURES 12.3

94 HP PROCEDURES 12.3

95 EXAMPLE HPBIN PROCEDURE proc hpbin data=ex12 numbin=5; input x1/numbin=4; input x2; ods output Mapping=Mapping; run; proc hpbin data=ex12 WOE BINS_META=Mapping; target y/level=nominal order=desc; run; HPBIN Procedure Documentation

96 EXAMPLE HPIMPUTE PROCEDURE proc hpimpute data=sampsio.hmeq out=out1; input mortdue value clage debtinc; impute mortdue / value = 70000; impute value / method = mean; impute clage / method = random; impute debtinc / method = pmedian; run; HPIMPUTE Procedure Documentation

97 HP PROCEDURES 12.3

98 HP PROCEDURES 12.3

99 EXAMPLE HPGENSELECT PROCEDURE proc hpgenselect data=getstarted; class C1-C5; model Total = C1-C5 / Distribution=Tweedie Link=Log; code File='ScoringParameters.txt'; run; HPGENSELECT Procedure Documentation

100 EXAMPLE HPSPLIT PROCEDURE proc hpsplit data=sashelp.hmeq maxdepth=7 maxbranch=2; target BAD; input DELINQ DEROG JOB NINQ REASON / level=nom; input CLAGE CLNO DEBTINC LOAN MORTDUE VALUE YOJ / level=int; criterion entropy; prune misc / N <= 6; partition fraction(validate=0.2); rules file='hpsplhme2-rules.txt'; score out=scored2; run; HPSPLIT Procedure Documentation

101 SAS ANALYTICS 13.1 LATEST RELEASE OF SAS ANALYTIC SOFTWARE SAS ANALYTICS 13.1 was released in December Available on all platforms Requires Base SAS 9.4M1 Builds on the 12.1 and 12.3 releases See support.sas.com/stat for further information.

102 SAS/STAT 13.1 OVERVIEW Survival analysis with interval-censored data Sensitivity analysis in multiple imputation for missing data Item response theory (IRT) models Fine and Gray model for competing risks data Bayesian choice models Path diagrams for structural equation models High Performance Statistics: principal components, model selection for quantile regression, finite mixture models, canonical discriminant analysis For more information about the analytical product enhancements, see support.sas.com.

103 SAS ANALYTICS 13.1 BAYESIAN CHOICE MODELS Discrete choice models are used in marketing research to model decision makers choices among alternative products and services. The experimental BCHOICE procedure performs Bayesian analysis for discrete choice models. Default noninformative priors or prior specification Posterior samples and diagnostics produced Nested logit, probit, and random effects models included SAS/STAT User s Guide: The BCHOICE Procedure

104 SAS ANALYTICS 13.1 ITEM RESPONSE THEORY MODELS Item response theory (IRT) models are used in education and other fields to calibrate and evaluate test items (SATs, GREs, GMAT). The experimental IRT procedure fits IRT models in SAS/STAT, including: Rasch model One-, two-, three-, and four-parameter models Graded response model with logistic or probit link Different items that have different response models Exploratory and confirmatory analysis Factor score estimation SAS/STAT User s Guide: The IRT Procedure

105 SAS ANALYTICS 13.1 COMPETING-RISK MODELS Competing risks occur with time-to-event data when some different event impedes the evaluation of the event of interest. The PHREG procedure now provides the Fine and Gray (1999) competing risk model, which focuses on the cumulative incidence function. SAS/STAT User s Guide: The PHREG Procedure

106 SAS ANALYTICS 13.1 SURVIVAL ANALYSIS FOR INTERVAL-CENSORED DATA The ICLIFETEST procedure computes nonparametric estimates of the survival function for data that are interval-censored. Supports syntax similar to PROC LIFETEST Tests equality of survivor functions Supports multiple comparison SAS/STAT User s Guide: The ICLIFETEST Procedure

107 SAS ANALYTICS 13.1 PATH DIAGRAMS IN THE CALIS PROCEDURE Path diagrams provide visually informative representations of the interrelationships among variables in structural equation models. Diagrams are provided for the initial model and the standardized and unstandardized solutions. SAS/STAT User s Guide: The CALIS Procedure

108 SAS ANALYTICS 13.1 NEW PRODUCTION PROCEDURES Three procedures become production with the 13.1 release. These include: ADAPTIVEREG procedure for fitting multivariate adaptive regression spline models QUANTLIFE procedure for quantile regression for right-censored data QUANTSELECT procedure for model selection for quantile regression models

109 SAS ANALYTICS 13.1 HIGHLIGHTS OF OTHER ENHANCEMENTS The Tweedie distribution is now supported by the GENMOD procedure (and HPGENSELECT). The elastic net method is provided in the GLMSELECT procedure. PROC MCMC has been multithreaded, providing performance boosts for models with complex likelihoods and/or many random effects. Power for PROC GLM-type MANOVA and repeated measurements is now provided in PROC GLMPOWER. PROC SURVEYMEANS produces domain quantile estimates.

110 SAS ANALYTICS 13.2 THIRD RELEASE OF SAS ANALYTIC SOFTWARE FOR SAS 9.4 SAS ANALYTICS 13.2 was released in August Available on all platforms Requires Base SAS 9.4M2 Builds on the 12.3 and 13.1 releases See support.sas.com/stat for further information.

111 SAS ANALYTICS 13.2 NEW PROCEDURES Fit proportional hazards regression models to interval-censored data. (ICPHREG) Perform analysis for spatial point patterns in two dimensions. (SPP) Provide weighted generalized estimating equations. (GEE) experimental Fits models using several linear predictive methods including PLS (HPPLS) Performs High performance quantile regression analysis (HPQUANTSELECT) SAS/Stat Documentation for 13.2 Documentation Tab

112 SAS ANALYTICS 13.2 HIGHLIGHTED ENHANCEMENTS TO PROCEDURES The FACTOR procedure generates path diagrams. The FMM procedure fits multinomial models. The IRT procedure generates polychoric correlation matrices, item characteristic curves, and test information curve plots. The MCMC procedure supports a categorical distribution in the MODEL, RANDOM, and PRIOR statements. The NLMIXED procedure enables you to specify more than one RANDOM statement in order to fit hierarchical nonlinear mixed models. The SEQDESIGN procedure enables you to create a ceiling-adjusted design that corresponds to integervalued sample sizes at the stages for nonsurvival data. The LOGISTIC procedure enables you to add or relax constraints on parameters in nominal response and partial proportional odds models. The FREQ procedure now provides score confidence limits for the odds ratio and the relative risk. The GLMSELECT procedure enables you to apply safe screening and sure independence screening methods to reduce a large number of regressors to a smaller subset from which model selection is performed.

113 WHAT S NEW IN SAS ENTERPRISE GUIDE 4.3, 5.1 & 6.1

114 ENTERPRISE GUIDE 4.3 Key Features New program editor with auto complete and integrated syntax help, Analyze SAS Program - code analyzer Open WRS reports Explicit SQL pass-through option Stored Process wizard UI enhancements Tasks NEW: Map Chart, Scatter Plot Matrix, Create Format from a dataset, ENHANCED: Filter and Sort, Random Sample, Tile Chart, Principal Components In-database optimized syntax generation sort on database if DBMS engine Autoexec Process Flow Send data files directly to JMP for further analysis Rapid Predictive Modeling (RPM) Enterprise Miner integration Customized task for model development - handles outliers, missing values, rare target events, skewed data, colinearity, variable selection, and model selection

115 ENHANCEMENTS FOR SAS PROGRAMMERS SAS ENTERPRISE GUIDE 4.3

116 ENHANCED PROGRAM EDITOR Mouse over Tooltip with details for SAS PROCS, SAS Functions

117 ENHANCED PROGRAM EDITOR - AUTOCOMPLETE Connects to data library listing accessible data tables Provides list of SAS PROCS, SAS Functions and Macro variables, assisting user while coding

118 ENHANCED PROGRAM EDITOR Converts code into Indented Format

119 ENHANCED PROGRAM EDITOR ANALYZE PROGRAM Coverts lines of Code into Process Flow for visual representation and ease understanding of the program flow

120 ENHANCED PROGRAM EDITOR OPTIONS General Options for the Editor Helps set the options for Code Appearance including Font Size and Font Colour for various key elements of the code

121 ENHANCED PROGRAM EDITOR OPTIONS Options to set the Autocomplete keywords/characters for coding Options to set the rules for code indentation

122 AD-HOC ANALYSIS ENHANCEMENTS SAS ENTERPRISE GUIDE 4.3

123 CREATE FORMAT Two Tasks 1. Create Format 2. Create Format from Data Set

124 MAP CHART

125 SCATTERPLOT MATRIX

126 FILTER AND SORT Customize the Task Name Specify where to save the output table

127 INTEGRATION WITH RAPID PREDICTIVE MODELER Integration with Rapid Predictive Modeler

128 INTEGRATION WITH RAPID PREDICTIVE MODELER Self service analysis and report generation using wizard driven RPM task

129 EXPLICIT SQL PASS-THROUGH PROC SQL; CONNECT TO ORACLE as con1 (PATH=ora10g2 authdomain=oracleauth); CREATE TABLE WORK.QUERY_FOR_CANDY_CUSTOMERS AS SELECT * FROM CONNECTION TO con1 ( SELECT t1.name, t1.region, t2.orderdate, t2.units, t2.discount FROM CANDY_CUSTOMERS t1 INNER JOIN CANDY_SALES_HISTORY t2 ON (t1.custid = t2.customer) WHERE t2.units < 2500); DISCONNECT FROM con1; QUIT; Query options to set explicit SQL Passthru for increased query performance utilizing database engine capabilities

130 SAS ENTERPRISE GUIDE 5.1 KEY NEW FEATURES 32 and 64-bit versions Enhanced program editor with auto complete for dataset variables Improved tooltip support for SAS keywords with help documents Reusable queries (Query Templates and Sub Queries) Data Explorer tool to view a SAS data set without adding the data set to project Create a custom view of your data by selecting columns to include in the view, creating a filter, and sorting data. Quickly calculate basic charts and statistics for each column in a SAS data set Add a data exploration view to project Show ODS Statistical Graph Enhanced Tile Chart Task Enriched interface for creating Stored Process

131 SAS ENTERPRISE GUIDE 5.1 NEW DEFAULT STYLE FOR SAS REPORT AND HTML RESULTS HtmlBlue is the new default style for SAS Report and HTML results

132 PROGRAMMERS PRODUCTIVITY FEATURES IN SAS ENTERPRISE GUIDE 5.1

133 ENHANCED EDITOR Auto completion for Variables in Data sets Provides the ability to view the list of variables from the selected data set.

134 ENHANCED EDITOR Added Syntax Auto-Completion support for HPA PROCs HPLMIXED Mixed Linear Models HPLOGISTIC Logistic Regression HPNLIN Non-Linear Regression HPREG - Regression HPCORR - Correlation

135 ENHANCED EDITOR Link to on-line help and documentation Enhanced tool tip support with web link to Product Documentation and Samples

136 FLEXIBLE EDITOR OPTIONS Ability to control many Editor options and Internet Search Engine

137 SAS ENTERPRISE GUIDE 5.1 ENHANCED PROGRAM EDITOR Ability to split and stack the programming window side by side or tiled Work with different parts of same program in single view

138 BUSINESS ANALYSTS PRODUCTIVITY FEATURES IN SAS ENTERPRISE GUIDE 5.1

139 INTERACTIVE DATA EXPLORATION Quick Stats: Obtain descriptive statistics quickly Filter and Sort Data on focused columns Generate data views for Process Flow

140 INTERACTIVE DATA EXPLORATION Insights, with Quick stats Business Filters

141 CHOICE OF DESCRIPTIVE STATISTICS

142 QUERY BUILDER ENHANCEMENTS Re-usable query templates Saved query settings, including data references Can be run on same or different data sources Ones that return only a single column can be used to create subqueries Point-and-click Subqueries Nested queries (output from one as input to another) Supported on WHERE clause, HAVING clause, and SELECT statement Subquery must either return a scalar value (for single value comparisons) or one column of data of the same type (for In a list comparisons)

143 REUSABLE QUERIES Define queries and save them as Templates

144 REUSABLE QUERIES

145 CODE EXECUTION OPTIONS SAS Programs and Task Properties Options for code submission

146 PARALLEL EXECUTION

147 INTEGRATION WITH ODS GRAPHICS DESIGNER 9.2 ODS Graphics Designer was introduced 9.3/Enterprise Guide 5.1 Task to start the ODS Graphics Designer interface 9.3/Enterprise Guide 5.1 New Task to Open the output file of a ODS Graphics Designer session *.SGD

148 ODS GRAPHICS DESIGNER

149 ODS GRAPHICS DESIGNER

150 SHOW ODS STATISTICAL GRAPH

151 SHOW ODS STATISTICAL GRAPH

152 SHOW ODS STATISTICAL GRAPH SELECT A DESIGN

153 SHOW ODS STATISTICAL GRAPH SELECT A DATA SOURCE

154 SHOW ODS STATISTICAL GRAPH OUTPUT

155 SAS ENTERPRISE GUIDE 6.1 WHAT S NEW The ability to analyze a SAS program to determine whether there are any possible internationalization issues. When you analyze a program for internationalization, SAS Enterprise Guide lists the lines of code that might be affected and suggests substitutions when possible Administration enhancements, such as the new stand-alone installer and application streaming support. The new installer is much smaller, thereby making it easier to install over a distributed deployment, especially using provisioning tools such as System Center Configuration Manager (SCCM).

156 SAS ENTERPRISE GUIDE 6.1 WHAT S NEW Improved programmer productivity with the new Log Summary window, which lists all the errors, warnings, and notes that were generated when the program ran, as well as related line numbers and a sample of the affected code The ability to use notes to add information to a process flow or to specific objects in the process flow Task Gallery Integration with SAS high-performance tools with the addition of the High-Performance Logistic and High- Performance Linear Regression tasks Improved Scheduler Windows Scheduler integration

157 SAS ENTERPRISE GUIDE 6.1 FOR THE PROGRAMMER/DEVELOPER Code or Log tab

158 SAS ENTERPRISE GUIDE 6.1 NOTES Note

159 SAS ENTERPRISE GUIDE 6.1 NOTES Note linked to program (or Task)

160 SAS ENTERPRISE GUIDE 6.1 TASK GALLERY

161 SAS ENTERPRISE GUIDE 6.1 TASK GALLERY ZOOM

162 SAS ENTERPRISE GUIDE 6.1 NEW RFM TASK

163 SAS ENTERPRISE GUIDE 6.1 HIGH PERFORMANCE PROCEDURES High performance

164 SAS ENTERPRISE GUIDE 6.1 SCHEDULING New Interface

165 SNEAK PEAK AT WHAT S COMING SOON!

166 RELEASED 2014 SAS STUDIO

167 RELEASED 2014 SAS IN-MEMORY ANALYTICS FOR HADOOP

168 RELEASED 2014 SAS VISUAL STATISTICS

169 WHAT S COMING SAS ENTERPRISE MINER

170 RESOURCES

171 RESOURCES WHAT S NEW What s New in SAS What's New in SAS Enterprise Guide

172 ADDITIONAL SUPPORT ENTERPRISE GUIDE TUTORIALS View Video Free Tutorials SAS Enterprise Guide Documentation Getting Started with SAS Enterprise Guide 6.1 SAS Enterprise Guide Tutorial for Statistics

173 RESOURCES SAS ANALYTICS SAS/Stat Newsletter /index.html Stat, IML, OR, ETS Papers Discussion Forum Stat Videos 101AA6

174 SAS/STAT : On Deck: SAS/STAT : Making Use of Incomplete Observations in the Analysis of Structural Equation Models: The CALIS Procedure?s Full Information Maximum Likelihood Method in SAS/STAT : The RANDOM Statement and SAS Global Forum Papers More: Moving On with PROC MCMC : Look Out: After SAS/STAT 9.3 Comes SAS/STAT 12.1! Online Help What is new in SAS/STAT 9.3 What's Changed since SAS/Stat 9.22 support.sas.com/statistics

175 RESOURCES SAS 9.4 Online documentation: SAS/STAT High Performance Documentation Leveraging Big Data Using SAS High-Performance Analytics Server High Performance Statistical Modeling SAS Meets Big Iron: High Performance Computing in SAS Analytic Procedures Blog Post: Documentation for high-performance analytics (definitions too)

176 RESOURCES SAS 9.4 Online documentation: ODS EPUB Destination ODS PowerPoint Destination

177 RESOURCES LIVE WEBINARS FROM CUSTOMER LOYALTY From A to Z: A SAS Enterprise Guide Case Study - October 21 Getting Started With SAS Enterprise Guide November 18 Hidden Gems in SAS Enterprise Guide - November 19 Getting Started with SAS Enterprise Miner November 3 Getting Started With SAS Forecast Server November 10 Getting Started With SAS Text Miner November 17 Data Mining Tasks With SAS Enterprise Guide 6.1 November 24 Introduction to SAS Add-in for Microsoft office November 6 Introduction to SAS Enterprise Business Intelligence Server November 13 Top 10 SAS Support Resources December 9

178 RESOURCES LIVE WEBINARS FROM CUSTOMER LOYALTY What s New in 9.4 for Foundation (programmers) November 4 Top 10 Ways to Optimize Your SAS code - October 29 Introduction to SAS Administration: The Basics December 1 Overview of Administration Task in SAS Visual Analytics November 25 Advanced SAS Administration Tasks in SAS 9.4: Backup Strategies and SAS Server logging October 15 SAS 9.4 Metadata Security Overview October 20 Metadata Server Clustering October 28 Considerations for Transitioning from PC to Server November 5

179 Connect with me: LinkedIn: QUESTIONS? Thank you for your time and attention! sas.com