SAS Analytics: Highlights of 15.2 Releases
New releases of SAS analytics software now available!
SAS/ETS 15.2 Highlights
This is a maintenance release that addresses reported issues in SAS/IML software. The documentation includes minor updates and editorial corrections.
SAS/IML 15.2 Highlights
This is a maintenance release that addresses reported issues in SAS/IML software. The documentation includes minor updates and editorial corrections.
SAS/IML Studio 15.2
The SAS/IML Studio product is no longer being developed or released.
SAS/OR 15.2 Highlights
This is a maintenance release that addresses reported issues in SAS/OR software. The documentation includes minor updates and editorial corrections.
SAS/QC 15.2 Highlights
This is a maintenance release that addresses reported issues in SAS/QC software. The documentation includes minor updates and editorial corrections.
SAS/STAT 15.2 Highlights
This is a maintenance release that addresses reported issues in SAS/STAT software. The documentation includes minor updates and editorial corrections.
SAS Simulation Studio 15.2 Highlights
The SAS Simulation Studio product is no longer being developed or released.
SAS Enterprise Miner 15.2 Highlights
Java Web Start is not supported for SAS Enterprise Miner 15.2. A new client download page can be used in place of Java Web Start.
SAS Text Miner 15.2 Highlights
SAS Text Miner 15.2 offers performance enhancements to previously existing features.
SAS Analytics: Highlights of 15.1 Releases
New releases of SAS analytics software now available!
SAS/STAT 15.1 Highlights
SAS/STAT 15.1 includes three new procedures in addition to many enhancements to existing functionality. Highlights of the new release include the following:
- NEW! PROC BGLIMM provides full Bayesian inference for generalized linear mixed models (GLMMs).
- NEW! PROC CAUSALGRAPH examines the structure of graphical causal models.
- NEW! PROC RMSTREG analyzes time-to-event data by using regression with respect to the restricted mean survival time (RMST).
- PROC ICPHREG fits the semiparametric proportional hazards model to interval-censored data.
- PROC PHREG performs Bayesian analysis of the proportional hazards spline model.
- PROC QUANTREG enables you to perform observation-wise conditional distribution analysis.
For More Information
|
|
SAS/QC 15.1 Highlights
SAS/QC 15.1 includes enhancements to several procedures.
- PROC CAPABILITY has the following enhancements:
- The BARFILL= option in the HISTOGRAM statement enables you to fill the histogram bars in different cells of a comparative histogram with different colors.
- The PCTLAXIS option in the QQPLOT statement has a new PCTLORDER= suboption, which specifies the tick mark values to be labeled on the percentile axis.
- In the INTERVALS statement, you can specify METHOD=7 to compute nonparametric tolerance intervals.
- PROC RAREEVENTS enables you to apply tests for special causes to rare events charts. These tests can improve the sensitivity of rare events charts by detecting patterns of measurements
that might indicate unusual variation.
- The CCHART, NPCHART, PCHART, and UCHART statements in PROC SHEWHART support the new LANEY option, which computes control limits as recommended by Laney.
- The EWMACHART and MACHART statements in PROC MACONTROL support the new LABASYMPTOTIC option, which labels varying control limits with asymptotic limit values.
- PROC SHEWHART, PROC ANOM, PROC CUSUM, and PROC MACONTROL recognize new macro variables that you can use to control details of ODS Graphics
output more conveniently.
For More Information
|
|
SAS/IML 15.1 Highlights
SAS/IML 15.1 supports several new statements and functions:
- The CONTINUE statement stops the processing of
the current iteration of a DO loop and resumes processing at the next iteration of the DO loop.
- The new digital filtering functions enable you to design digital filters and apply those filters to signals. Many of the digital filtering functions start with
the "DF" prefix. The new function and subroutines are CCEPSTRUM, DFCONV, DFDESIGN, DFFILT, DFFREQZ, DFFREQZZPK, DFMEDFILT, DFORDER, DFSOSFILT, DFSOSFREQZ, DFSOSFREQZZPK, ICCEPSTRUM, and RCEPSTRUM.
- The experimental KPCATRAIN subroutine computes a
kernel principal component (kPCA) analysis from training data. The experimental KCPASCORE function uses the kPCA model to score new data.
- The FEVAL function enables you to evaluate a function
indirectly by specifying the name of the function and its arguments.
- The LEAVE statement exits the current DO loop and
resumes processing at the statement that follows the DO loop.
- The MODULESTACK function returns the names of all
modules in the module call stack.
- The SPECTROGRAM subroutine displays a spectrogram
of a short-time Fourier transform of a time series signal.
- The TABLESORT subroutine sorts a table by one or
more columns.
For More Information
|
|
SAS/IML Studio 15.1 Highlights
SAS/IML Studio 15.1 supports the following new features:
- Formats and informats that have been added since SAS 9.2.
- New statements for DO loops. The LEAVE and CONTINUE statements enable you to control when a DO loop terminates or proceeds to the next iteration, respectively.
- New syntax for the FROM keyword of the CREATE and APPEND statements. You can now specify multiple matrices on those statements.
For More Information
|
|
SAS Enterprise Miner 15.1 Highlights
SAS Enterprise Miner 15.1 provides the following improvements and enhancements:
- For the Gradient Boosting node, the default value for the Leaf Fraction property has been changed from 0.1 to 0.001.
- For the HP Text Miner node:
- Support was added for 18 additional languages: Arabic, Croatian, Czech, Danish, Farsi, Greek, Hindi, Hebrew, Hungarian, Indonesian, Norwegian, Polish, Romanian, Slovak, Slovene, Swedish,
Thai, and Vietnamese.
- All languages, except Hebrew and Norwegian, now support the Find Entities property.
- For UTF-8 installations, default stop lists were added for Czech, Danish, Dutch, Greek, Spanish, Finnish, Hebrew, Hindi, Croatian, Hungarian, Indonesian, Italian, Norwegian, Polish,
Portuguese, Romanian, Russian, Slovak, Slovene, Swedish, and Turkish. Previously, default stop lists were available only for English, French, and German.
For More Information
SAS Text Miner 15.1 Highlights
What's New in SAS Text Miner 15.1
SAS Text Miner 15.1 offers performance enhancements when running text mining nodes and viewing results. In addition, new stop lists are available on WLATIN1 and UTF-8 installations from
the SASHELP library. These stop lists can be specified using the Stop List property of the Text Parsing node.
For More Information
SAS/OR 15.1 Highlights
SAS/OR 15.1 includes performance improvements to its optimization procedures and solvers, including the following enhancements:
- The LP, MILP, network, QP, NLP, and CLP solvers and the decomposition algorithm (for LP and MILP) add performance and stability improvements.
- The LP and MILP solvers significantly reduce their memory consumption.
- The LP solver replaces its presolver.
- The MILP solver and the OPTMILP procedure add the ability to report multiple solutions.
- The network solver adds a path enumeration algorithm to find all paths between specified nodes.
- The connected components algorithm in the OPTNET procedure adds support for the thin internal graph format.
- The DECOMP algorithm adds the COMMUNITY value to the METHOD= option.
- The CLP solver adds the CUMULATIVE predicate, and the CLP procedure adds the CUMULATIVE statement.
- The NLP solver adds a new interior point solution algorithm on an experimental basis.
- The LSO (local search optimization) solver is added and is called by the OPTMODEL procedure.
For More Information
|
|
SAS Simulation Studio 15.1 Highlights
SAS Simulation Studio 15.1 adds several enhancements, including the following:
- Improvements in the layout of the graphical user interface (GUI), including collapsible block templates and tabbed project panels
- Easier access to recently opened projects
- The ability to save the size and position of Model and Experiment windows in the last project opened
- Ease-of-use enhancements for models and experiments
- Improved performance in model execution, especially when reading or writing data multiple times
- Authentication during batch mode model runs without the need to open the GUI
For More Information
|
|
SAS/ETS 15.1 Highlights
SAS/ETS 15.1 includes a new interface engine and enhancements to several procedures:
- SASEOECD interface engine: This new engine enables you to retrieve time series data from the Organisation for Economic Co-operation and Development (OECD) website.
- PROC AUTOREG: The CLASS statement is production, and you can output standard errors of prediction intervals by specifying the STDERR= and STDERRM= options
in the OUTPUT statement.
- PROC COPULA: You can use a SIMULATE statement with a BY statement, provided that you specify a FIT statement before the SIMULATE statement. You can also
use multiple FIT and SIMULATE statements with a BY statement.
- PROC COUNTREG: This procedure includes two new options for variable selection. The NOSPLITEFFECTS option requests that effects that involve class variables
not be split into individual effects that correspond to class levels, and the RETAINEFFECT option requests that the effects named within parentheses be retained during the variable selection process.
- PROC SPATIALREG: Taylor and Chebyshev approximation techniques for spatial error models and spatial Durbin error models have been added.
- PROC SSM: You can now request that parameter estimation be based on a new type of likelihood called marginal likelihood.
- PROC TMODEL: Most features are now production. Only the RANDOM statement and the QUADHESS=ANALYTIC option remain experimental. The LUSOLVER= option has been
added to provide better control over the solution of matrix equations that occur in FIT and SOLVE tasks.
- PROC UCM: You can now add a transfer-function component to your model by using the new TF statement. You can also request that parameter estimation be
based on a new type of likelihood called marginal likelihood.
- PROC VARMAX: Conditional forecasts and scenario analysis are supported for the vector autoregressive (VAR) model, Bayesian VAR model, vector error correction
model (VECM), and Bayesian VECM, with or without exogenous variables. You can now plot dynamic conditional covariances for DCC GARCH models by specifying PLOTS=CONDCORR. The OUTSTAT statement
now includes the value of the log-likelihood function that is calculated at the parameter estimates.
For More Information
|
|
SAS Analytics: Highlights of 14.3 Releases
New release of SAS analytics software now available!
SAS/STAT 14.3 Highlights
SAS/STAT 14.3 includes one new procedure in addition to many enhancements to existing functionality.
Highlights of the new release include the following:
- The new CAUSALMED procedure estimates causal mediation effects from observational data.
- The GAMPL procedure now supports the Tweedie distribution.
- In PROC FREQ, the COMMONRISKDIFF option in the TABLES statement provides estimates, confidence limits,
and tests for the overall risk (proportion) difference for multiway tables.
- The IRT procedure now supports the nominal response model, which enables you to do item analysis of
nominal responses.
- The NLMIXED and MCMC procedures add a CMPTMODEL statement that fits compartment models in pharmacokinetic
analysis.
- The PHREG procedure provides the cause-specific proportional hazards analysis for competing-risks data.
- The QUANTREG and QUANTSELECT procedures provide fast quantile process regression.
- The VARMETHOD=BOOTSTRAP option provides variance estimation by the bootstrap method for the survey
data analysis procedures.
- The TTEST procedure provides bootstrap standard error, bias estimates, and confidence limits.
For More Information
|
|
SAS/QC 14.3 Highlights
SAS/QC 14.3 has a new feature for the RAREEVENTS procedure.
In SAS/QC 14.3, the RAREEVENTS procedure can produce rare events charts that have distinct sets of probability limits for different phases of observations. The phases are defined by the values of the
character variable _PHASE_ in the data sets that are specified in the DATA=, LIMITS=, and TABLE= options.
The CHART statement supports the following new options related to phases:
- LIMITPHASES= specifies the phases for which probability limits are read from the LIMITS= data set.
- NOPHASEREF suppresses the vertical reference lines that separate phases.
- NOPHASEREFFILL suppresses the filling of graph walls for phases.
- PHASELEGEND displays phase labels in a legend across the top of the chart.
- PHASELIMITS labels probability limits and center lines with their values within each phase.
- READPHASES= selects phases from the DATA= or TABLE= data set for processing.
In addition, you can use the new HAXISLABEL= option in the CHART statement to specify a horizontal axis label for a rare events chart and in the COMPARE statement to specify a horizontal axis label
for a comparison chart.
For More Information
|
|
SAS/IML 14.3 Highlights
- The SAS/IML language supports new syntax for defining and manipulating lists. You can use square brackets to define a list, subscript notation to extract a sublist,
a dollar sign ($) to extract an item from a list, and the concatenation operator (||) to concatenate lists. For examples and a description of the syntax, see
Lists and Data Structures.
- You can transfer data between SAS/IML tables and R data frames by using the
ExportTableToR subroutine and the
ImportTableFromR function.
- You can analyze time series data by using the new functions for time-frequency analysis such as the short-time Fourier transform analysis.
- The TABLECREATE function supports creating a
table from multiple matrices.
- The SAS/IML wavelet functions now support ODS graphics.
For More Information
|
|
SAS Enterprise Miner 14.3 Highlights
SAS Enterprise Miner 14.3 provides the following improvement and enhancement:
- The SAS Viya Code node no longer requires SAS/CONNECT to access the SAS Viya procedures. The node enables streamlined access to the
latest machine learning techniques from within SAS Enterprise Miner.
For More Information
SAS Text Miner 14.3 Highlights
What's New in SAS Text Miner 14.3
SAS Text Miner 14.3 offers performance enhancements when running text mining nodes and viewing results. It also offers improved parsing performance for the following languages:
Arabic, Czech, Danish, Greek, Hebrew, Hungarian, Indonesian, Norwegian, Polish, Romanian, Slovak, Swedish, Thai, and Vietnamese.
These languages now use the HPTMINE procedure, which allows for multi-threaded parsing. The following languages continue to use the HPTMINE procedure: Chinese, Dutch, English, Finnish, French,
German, Italian, Japanese, Korean, Portuguese, Russian, Spanish, and Turkish.
For More Information
SAS/OR 14.3 Highlights
SAS/OR 14.3 includes performance improvements in the LP, MILP, QP, and NLP solvers. If you also license SAS Optimization
8.1 (or a later release), you can invoke it and SAS/OR 14.3 from the same SAS session.
SAS Simulation Studio 14.3, a component of SAS/OR 14.3 for Windows environments, adds Linux support (on an experimental basis) and also
adds an alternative user interface design.
For More Information
|
|
SAS/ETS 14.3 Highlights
SAS/ETS 14.3 provides the following improvements and enhancements:
- New SASEWBGO interface engine enables you to retrieve time series data from the World Bank Group Open (WBGO) data website.
- New TMODEL procedure incorporates high-performance computational techniques and offers new features that enhance the functionality of PROC MODEL.
- The PANEL procedure was enhanced by adding new options for dynamic panel estimation including DYNDIFF and DYNSYS options to estimate models with differenced and level equations.
- The QLIM procedure adds the no-U-turn sampler (NUTS) of the Hamiltonian algorithm that samples from the posterior distribution by using information from the gradient.
- The SASEFAME interface engine adds features to support remote access to MarketMap's (FAME) master and MCADBS servers.
- The SASEFRED interface engine adds support for USER= and DEBUG= options to store data sets and logs.
- The SASEQUAN interface engine was enhanced to support Version 3 of the Quandl API.
- The SASERAIN interface engine now supports the premium weather API by default while the free weather API has been discontinued by World Weather Online.
- The SSM procedure was enhanced by adding model-based temporal aggregation and temporal distribution for time series that have response variables of flow type.
- The UCM procedure now allows you to specify higher-order stochastic cycle components in your models.
- The VARMAX procedure was enhanced to include support for confidence intervals and standard errors for impulse response functions of exogenous variables and for all impulse
response functions of VARFIMA models now appear in corresponding plots and tables.
For More Information
|
|
SAS Analytics: Highlights of 14.2 Releases
SAS/STAT 14.2 Highlights
SAS/STAT 14.2 includes two new procedures in addition to many enhancements to existing functionality. Highlights of the new release include the following:
- The new PSMATCH procedure provides propensity score analysis.
- The new CAUSALTRT procedure performs estimation of causal treatment effects.
- The PHREG procedure now provides time-dependent ROC curves for Cox regression.
- The NLIN procedure now provides ESTIMATE and CONTRAST statements.
- The SURVEYIMPUTE procedure provides two-stage fully efficient fractional imputation and fractional hot-deck imputation.
- The FREQ and SURVEYFREQ procedures provide additional agreement statistics.
For More Information
|
|
SAS/QC 14.2 Highlights
SAS/QC 14.2 includes the following enhancements:
- The RAREEVENTS procedure is now production. You can compute probability limits based on a Weibull distribution, and the COMPARE statement provides different ways to
display the process data and reference distribution.
- The ANOM, CUSUM, MACONTROL, and SHEWHART procedures have new options that provide more flexibility with subgroup markers and labels.
For More Information
|
|
SAS/IML 14.2 Highlights
- Support for new nonmatrix data types: tables and lists.
- The RANDGEN subroutine supports the
following new distributions: Conway-Maxwell-Poisson, extreme value, generalized Poisson, Gompertz, Gumbel, integer, and shifted Gompertz.
- The SUBMIT statement
has been enhanced. Macro variables that are created in a SUBMIT block are available outside the SUBMIT block. Global SAS statements that are executed in a
SUBMIT block affect the environment outside the SUBMIT block.
For More Information
|
|
SAS Enterprise Miner 14.2 Highlights
- The new SAS Viya Code node enables you to execute code in your SAS Viya environment within a process flow diagram.
- The Score node and Score Code Export nodes create the analytic store files containing the score code for an entire flow that includes an HP Forest, an HP SVM node, and/or a SAS Viya Code node.
These files can be used by SAS Scoring Accelerator to deploy the score code in a database.
For More Information
SAS Factory Miner 14.2
- Batch code can be created to enable you to retrain your models with new data.
For More Information
SAS Text Miner 14.2 Highlights
- SAS Text Miner 14.2 offers performance enhancements when running text mining nodes and viewing results. It also offers improved parsing results.
For More Information
SAS/OR 14.2 Highlights
- Performance improvements in the LP, MILP, and NLP optimization solvers and in the decomposition algorithm for LP and MILP
- Better solutions, shorter solution times, and more benchmark problems solved within time limits
For More Information
|
|
SAS/ETS 14.2 Highlights
- The new SASENOAA interface engine enables you to retrieve severe weather data from the National Oceanic and Atmospheric
Administration (NOAA) Severe Weather Data Inventory web service.
- The new SASERAIN interface engine enables you to retrieve weather data from the World Weather Online website,
including data on temperature, precipitation (rainfall), weather description, weather icon, and wind speed.
- The new SPATIALREG procedure analyzes spatial econometric models for cross-sectional data that contain spatially
referenced or georeferenced observations.
- The HPCDM procedure is now production, with three new options: the MAXCOUNTDRAW= option specifies an upper limit on the
number of loss events (count), the SEVERITYSTORE= option specifies the item store that contains the context and estimates of the severity model, and the
COUNTSTORE= option supports internal count simulations with BY groups.
- The HPSEVERITY and SEVERITY procedures were enhanced by adding plots for comparative and per-distribution conditional
probability density functions. A new OUTPUT statement is available to write distribution function and quantile estimates.
- The QLIM procedure adds the RANDOM statement, which enables you to estimate the random-parameters models in addition to the
random-effects models.
- The SASEFAME interface engine was upgraded to support Fame CHLI 11.4.
- The SASEFRED, SASEQUAN, and SASEXFSD interface engines add the CONNECT= option to provide a secure connection via a
proxy server, the PROXY= option to specify which proxy server to use, and the DEBUG=ON option to provide diagnostic logging in the SAS log window.
- The SSM procedure adds two options: the new BREAKPEAKS option prints an alternate form of the break summary tables, and
the new ZSPARSE option enables the exploitation of the sparsity of the Zt matrices in the observation equation.
- The TIMESERIES procedure adds the GROUPS= option to allow an automatic grouping based on the w-correlations after a maximal
number of groups is specified.
- The VARMAX procedure adds support for vector autoregressive fractionally integrated moving average (VARFIMA) models,
which capture both short and long memory in the data and also allow for exogenous variables (VARFIMAX), constants, seasonal dummies, and time trends.
For More Information
|
|
SAS Forecast Server 14.2 Highlights
- Count Series Analysis and Forecasting
- PROC TIMESERIES - COUNT statement
- PROC TIMEDATA - TSA package
- Automatic Singular Spectrum Analysis and Forecasting
- PROC TIMESERIES - SSA statement
- Time-Frequency Analysis (TFA package)-PROC TIMEDATA
- Time Series Analysis (TSA package)- PROC TIMEDATA
- State-Space Models (SSM) Enhancements
- PROC SSM
- DEPLAG statement is production now.
- Two new experimental options on the PROC statement:
- SPARSEZ option enables exploitation of sparseness in the Z vectors of an SSM.
- BREAKPEAKS option results in an alternate form of the break summary tables to be printed when the CHECKBREAK option is used in the STATE or TREND statements, or the MAXSHOCK option
is used in the OUTPUT statement.
For More Information
SAS High-Performance Analytics
SAS High-Performance Econometrics
- The HPCDM procedure is now production, with three new options: the MAXCOUNTDRAW= option specifies an upper limit on the number of loss events (count), the SEVERITYSTORE= option specifies the item store
that contains the context and estimates of the severity model, and the COUNTSTORE= option supports internal count simulations with BY groups.
- The HPSEVERITY procedure was enhanced by adding plots for comparative and per-distribution conditional probability density functions. A new OUTPUT statement is available to write distribution
function and quantile estimates.
For More Information
SAS Analytics: Highlights of 14.1 Releases
SAS/STAT 14.1 Highlights
SAS/STAT 14.1 includes two new procedures in addition to many enhancements to existing functionality. Highlights of the new release include the following:
- The new SURVEYIMPUTE procedure provides missing value imputation for survey data.
- The new GAMPL procedure fits generalized additive models via penalized likelihood estimation.
- PROC HPSPLIT provides classification and regression trees via a standard modeling syntax.
- The MCMC procedure adds the Hamiltonian Monte Carlo (HMC) and No-U-Turn Sampler (NUTS) sampling algorithms for continuous parameters, support for leading and lagging variables, an
ordinary differential equation solver, and a general integration function.
- The IRT procedure fits generalized partial credit models.
- The LIFETEST procedure performs nonparametric analysis of competing-risks data.
- The POWER procedure supports Cox proportional hazards regression models.
For More Information
|
|
SAS/QC 14.1 Highlights
- The new, experimental RAREEVENTS procedure produces control charts for rare events. A rare events chart is better suited than traditional control charts to detecting changes in the
frequency of low-probability events. The data that are plotted in a rare events chart represent the times between successive events. Rare events charts have gained acceptance in health
care quality improvement applications because of their ease of use and suitability to processes that have low defect rates.
- The ANOM, CUSUM, MACONTROL, and SHEWHART procedures are now capable of producing graphs that you can edit by using the ODS Graphics Editor.
For More Information
|
|
SAS/IML 14.1 Highlights
- Support for large matrices on the Windows operating system (up to 231 elements, or more than 2 billion elements).
- The PACKAGE statement supports installing and using packages, which are ZIP files that contain source code, data sets, documentation, and sample programs. You can share and download
packages from the SAS/IML File Exchange.
- The eigenvalue computations in SAS/IML use vendor-supplied math libraries, if available.
For More Information
SAS Enterprise Miner 14.1 Highlights
- The PMML score code generated by SAS Enterprise Miner 14.1 is now PMML version 4.2.
- The Incremental Response node features a new Prescreening Variable property. This property enables you to specify whether node variable selection is performed using Net
Information Value scores or Adjusted Net Information Value scores.
For More Information
SAS Factory Miner 14.1
SAS Factory Miner is new in the 14.1 release and runs as an add-on to SAS Enterprise Miner.
- Builds and retrains multiple predictive models, including Bayesian networks, decision trees, generalized linear models, neural networks, random forests, regression, and support
vector machines across segments, such as state, region, or product.
- Provides model templates that can be customized to include data preparation capabilities, variable selection, and various options for the predictive model used. These templates can then be
shared across projects or users.
- Identifies a champion model based on one of several statistical criteria for each segment that can easily be deployed in production environments, and generates complete scoring code for each model.
For More Information
SAS Text Miner 14.1 Highlights
- The new HPBOOLRULE procedure replaces macros in the Text Rule Builder node.
- The Text Parsing node now does multithreaded parsing (using PROC HPTMINE) for most languages.
- You can now specify where scoring data sets are written by using the EM_TERM_LOC macro variable, which facilitates both deploying your text mining models and building stored processes.
- An _item_ variable with term | role information has been added to the transaction output that is exported from the Text Topic node and the Text Filter node. If these are
followed by an Association node, the user will be able to see the actual terms rather than the term IDs in the association plots.
For More Information
SAS/OR 14.1 Highlights
- Several optimization solvers improve their performance.
- The concurrent FOR loop (the COFOR loop) in PROC OPTMODEL can run in distributed mode (requires SAS High-Performance Optimization).
- PROC OPTMODEL adds a profiler that tracks the amount of time spent in problem generation, presolving, and various stages of the solution process.
- PROC OPTNET enables parallel computing, provides faster graph data input, and adds enhancements to three of its algorithms.
- The quadratic and nonlinear solvers add irreducible infeasible set (IIS) diagnostics.
- The decomposition algorithm expands the range of constraint matrix structures that it can detect automatically.
- The CLP procedure adds more variable and value selection strategies.
- SAS Simulation Studio adds controls on the order of execution for ports and for blocks, controls on the allocation of resource, availability adjustments, and automatic launching of the local SAS server.
For More Information
SAS/ETS 14.1 Highlights
- The new X13 procedure incorporates the X12 procedure in response to the US Census Bureau's inclusion of the X-12-ARIMA methodology in the X-13ARIMA-SEATS program.
PROC X13 also adds numerous options, displays additional tables, and changes the default value of the MAXITER= option to 1,500.
- The COUNTREG procedure adds the TEST statement, three statements that enable you to include spatial effects in a model, and more Bayesian analysis features.
- The HPCOUNTREG procedure adds the TEST statement and support for the Conway-Maxwell distribution.
- The HPPANEL procedure adds support for the between-groups estimator, between-time-periods estimator, and pooled OLS regression.
- The MODEL procedure adds the %EQAR and %EQMA macros.
- The QLIM procedure adds the RANDOM statement, which enables you to estimate the random-intercept models, and more Bayesian analysis features.
- The SSM procedure adds the DEPLAG statement, which simplifies the specification of models that have lagged values of response variables in the observation equation.
For More Information
|
|
SAS Forecast Server 14.1 Highlights
- SAS Forecasting now includes a new web-based Forecast Server Client. It combines the functions of two existing clients, while adding new functionality: demand classification segmentation for
targeted modeling, multi-stage modeling, exception rules to track series that meet certain criteria, forecast accuracy tracking across iterations, and parallel execution of modeling tasks.
The interface is now extensible via additional user-defined segmentation and modeling strategies.
For More Information
SAS High-Performance Analytics
SAS High-Performance Data Mining
- The new HP Bayesian Network Classifier node fits a Bayesian network for class targets.
- The HP Variable Selection node adds a tree-based selection method and a Boolean screening property for use with the LAR/LASSO algorithm.
- The HP Cluster node enables automatic selection of the number of clusters by using the ABC criterion.
- The HP SVM and HP Forest nodes can create an analytic store, which is a portable format of the model that can be used to score observations within a database.
- The HP Forest node has several other enhancements:
- A new variable importance method for variable selection, called Random Branch Assignments.
- New default values for Leaf Size (1) and Leaf Fraction (0.00001) to improve model accuracy.
- Improved data distribution when the HPFOREST procedure is run in distributed mode.
- The HP Regression node provides a fast backward selection method for logistic regression.
- The HP Tree node provides the following new features:
- Options for the Nominal Target Criterion property are expanded to include Information Gain Ratio and CHAID.
- Options for the Interval Target Criterion property are also expanded to include CHAID.
- A new Surrogate Rules property adds an option for handling missing values in the data.
- A new Use Input Once property controls whether an input can be used multiple times or at most once in a branch.
- Several new subtree methods are included, such as Cost-Complexity Pruning, which uses cross validation when no validation data are present, and pruning and subtree selection, based on
minimizing an assessment measure such as misclassification rate or average square error.
For More Information
SAS High-Performance Econometrics
- The HPCOUNTREG procedure adds a TEST statement and support for the Conway-Maxwell Poisson distribution.
- The HPPANEL procedure supports the between-groups estimator, the between-time-periods estimator, and pooled OLS regression.
For More Information
SAS High-Performance Statistics
- The new GAMPL procedure fits generalized additive models by penalized likelihood estimation.
- The HPGENSELECT procedure supports the LASSO method, BY-group processing, and the RESTRICT statement.
- The HPPRINCOMP procedure adds the METHOD= option in the PROC HPPRINCOMP statement, which specifies which principal component extraction method to use. The available methods are eigenvalue
decomposition, nonlinear iterative partial least squares, and an iterative method based on Gram-Schmidt orthogonalization.
- The HPSPLIT procedure supports MODEL and CLASS statements and cost-complexity pruning with k-fold cross validation as the default method of selecting the penalty parameter.
For More Information
SAS High-Performance Text Mining
- Enhancements to the HPTMINE procedure enable you to select or ignore parts of speech, attributes, and entities, as well as to build a search index.
- The HP Text Miner node now uses PROC HPTMINE to perform topic rotation and to create the topic table.
- Eleven parsing languages have been added to the Language property in the HP Text Miner node.
For More Information
SAS Analytics: Highlights of 13.2 Releases
SAS/STAT 13.2 Highlights
SAS/STAT 13.2 includes three new procedures in addition to many enhancements to existing functionality.
Highlights of the new release include the following:
- Weighted generalized estimating equations
- Proportional hazards regression models for interval-censored data
- Spatial point pattern analysis
- Finite mixture models for multinomial data
- Safe screening and sure independence methods for variable selection
- Polychoric correlation matrices, item characteristic curves, and test information curves for IRT models
- Score confidence limits for the odds ratio and the relative risk
- Parameter constraints for nominal response and partial proportional odds models
For More Information
|
|
SAS/QC 13.2 Highlights
- Geometric means of analysis variables are now computed.
- Confidence bands are available for stress-lifetime plots.
- ODS Graphics output is produced by using templates that are written in the Graph Template Language.
For More Information
|
|
SAS/IML 13.2 Highlights
- New EXECUTEFILE subroutine
- SAS/IML File Exchange
For More Information
SAS Enterprise Miner 13.2 Highlights
- The Model Registration Node enables you to specify the type of model to register.
- The Model Registration Node enables you to register a model on a remote metadata server by specifying macro variable values in your project start-up code.
- The HP Regression node produces a new variance inflation factor (VIF) table that can be used for multicollinearity detection.
- The HPForest node adds support for partitioned validation data.
- You can specify global metadata changes within SAS Code.
- The High-Performance Data Mining node adds support for SAP HANA and SPDE.
For More Information
SAS/OR 13.2 Highlights
- Improved performance for optimization solvers
- Experimental constraint logic programming solver in PROC OPTMODEL
- Covariance matrix output for nonlinear problems
- Additional block detection features for the decomposition algorithm
- Improved performance for SAS Simulation Studio, including parallel execution of design points and replications and new controls on saved data and numerical precision
For More Information
SAS/ETS 13.2 Highlights
- Convenient access to more than 8 million data sets from the Quandl website
- More Bayesian estimation features in several procedures
- First-differenced methods and new hypothesis tests for panel data
- PROC SEVERITY now supports a CLASS statement and enables you to specify a wide variety of model effects
- Multistep forecasts for multivariate GARCH models and improved cointegration tests in PROC VARMAX
- New options in PROC X12 for forecast confidence limits and outlier detection
For More Information
|
|
SAS Text Miner 13.2 Highlights
- German language support
- Support of documents that contain more than 32K characters
- Support of custom LITI files for custom entity and noun group extraction
- In addition, the HPTMINE procedure supports SVD-only mode, discovery of topics in the text corpus, and storage of the term-by-document matrix in the Base64-encoded sparse rows (BESR) format
For More Information
High-Performance Analytics Updates
- Statistics: partial least squares and quantile regression
- Econometrics: compound distribution models and regression models for count data
- Text Mining: parsing and custom entity and noun group extraction
- Data Mining: time series, cluster analysis, neural networks, support vector machines, and random forest modeling
SAS Analytics: Highlights of 13.1 Releases
Rich new releases of many SAS analytical products are available. The 13.1 releases of these products are included with the first maintenance release of Base SAS? 9.4. Software for statistics,
econometrics, forecasting, quality improvement, optimization, and high-performance analytics will be provided. The following are highlights.
SAS/STAT 13.1 Highlights
- Survival analysis for 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
For More Information
|
|
SAS/QC 13.1 Highlights
- Westgard rules for laboratory quality control
- Recurrent event plots and additional probability distributions for reliability analysis
- New macros for measurement system analysis
- Improved parameter estimation for the Johnson SU distribution
For More Information
SAS/IML 13.1 Highlights
- Enhanced linear programming and mixed integer linear programming optimization
- Enhanced data visualization through ODS Graphics
- Additional functions for data analysis and simulation
- New SAS programming window for SAS/IML? Studio
- Enhanced editing features in the SAS/IML Studio programming environment
For More Information
|
|
SAS Enterprise Miner 13.1 Highlights
- Time series data mining nodes
- Bayesian network classification
- Support vector machines
- Generalized linear models node
- Support for clustered mid-tier deployments
- Model registration
For More Information
|
|
SAS/OR 13.1 Highlights
- Support for user-defined functions by using PROC FCMP for optimization modeling in PROC OPTMODEL
- New concurrent processing loop in PROC OPTMODEL for threaded execution of solver invocations
- Support for multiple objectives in local search optimization
- Concurrent solver option for linear and nonlinear optimization
- GUI enhancements for SAS? Simulation Studio
For More Information
SAS/ETS 13.1 Highlights
- Models for limited dependent variables with endogenous regressors
- New multivariate time series techniques that provide performance gains
- Compound distribution modeling for loss simulation
- Multivariate Bayesian limited dependent models
- Convenient access to Federal Reserve Economic Data (FRED)
- Conway-Maxwell-Poisson regression
For More Information
|
|
SAS Forecast Server 13.1 Highlights
- Diagnostic engine multithreaded for model generation and for variable and event selection
- Forecasting engine multithreaded for model selection
- Support for SAS Grid Manager workspace servers for load balancing of multiple projects
- Interactive exploration and analysis of large volumes of data in SAS Time Series Studio
For More Information
|
|
SAS Text Miner 13.1 Highlights
- Text Profile node profiles changes over time
- Capability to filter language in text parsing node
- Addition of matrix graph in the Text Topic node for terms comparison
For More Information
High-Performance Analytics Updates
- Statistics: principal component analysis, finite mixture model analysis, canonical discriminant analysis
- Econometrics: panel data modeling, compound distribution models for loss simulation, multivariate copula simulation
- Text Mining: topic rotation
- Data Mining: cluster analysis, support vector machines, Bayesian network classification, generalized linear models
For More Information
New Statistical Graphs from SAS/STAT
SAS Analytics: Highlights of 12.1 Releases
The following are highlights.
SAS/STAT 12.1 Highlights
- model selection for quantile regression with the new QUANTSELECT procedure
- quantile regression for censored data with the new QUANTLIFE procedure
- multivariate adaptive regression splines with the new ADAPTIVEREG procedure
- direct and indirect standardized rates and risks for study populations with the new STDRATE procedure
- FMM procedure becomes production and adds support for additional distributions
- Bayesian frailty models in PROC PHREG
- modeling missing covariates in the MCMC procedure
- Mosaic plots and heat maps
- Partial proportional odds model
- Poisson sampling in PROC SURVEYSELECT
- Poststratification estimation in PROC SURVEYMEANS
- Fleming-Harrington estimates in PROC PHREG
- ESTIMATE statement in PROC QUANTREG
- Additional postprocessing inference for the LIFEREG and PROBIT procedures
For More Information
|
|
SAS/QC 12.1 Highlights
- The new MVPDIAGNOSE procedure produces principal component score plots and process variable contribution plots that are used to investigate the causes of unusual variation in a process.
- The MVPMODEL and MVPMONITOR procedures are now production.
- The CAPABILITY procedure has a number of new options for the CDFPLOT, COMPHISTOGRAM, HISTOGRAM, PPPLOT, PROBPLOT, and QQPLOT statements.
- The PARETO procedure now supports several new options in the HBAR and VBAR statements. These options enable you to create variations on the traditional Pareto chart and enable you to specify titles and
footnotes in graphs that are produced by using ODS Graphics.
- The RELIABILITY procedure now supports EFFECTPLOT, ESTIMATE, LSMEANS, LSMESTIMATE, SLICE, STORE, and TEST statements. These statements are common to many SAS/STAT? procedures and provide further
analysis of regression models that are fit by using a MODEL statement.
- PROC RELIABILITY also provides the following new features:
- confidence bands for mean and intensity functions for recurrent events
- two-sample log-rank-type test for recurrent events
- trend tests for recurrent events parametric models
- Duane plots
- profile likelihood confidence intervals for survival function
For More Information
SAS/IML 12.1 Highlights
- a new syntax for defining default parameter values in user-defined modules
- a new syntax for reading and writing SAS data sets whose names are not known until run time
- new statistical functions, subroutines, and modules
- support for additional distributions and parameters for generating random samples by using the RANDGEN subroutine
In addition, SAS/IML Studio is now available in both a 64-bit edition and a 32-bit edition and supports the following enhancements:
- support for the ODS HTML destination. SAS/IML Studio 12.1 uses the ODS HTML destination by default, which means that you can now view ODS statistical graphics.
- an improved ability to run multiple IMLPlus programs in parallel on computers with multiple processors or with multicore processors
- new method in the R class that enables you to copy plots that are created in R into the ODS HTML or LISTING destination
- support for user-defined formats in data tables and graphs
For More Information
|
|
SAS Enterprise Miner 12.1 Highlights
- SAS Enterprise Miner Core
- Several important and common SAS Enterprise Miner project properties have been promoted from macro variables that were deployed by using SAS code or project start code to property
selections that can be defined by using the Enterprise Miner GUI properties panel.
- The SAS Enterprise Miner client can now be opened directly into a specific project or diagram, or from the most recent project and diagram.
- The %AA_MODEL_REGISTER macro has been added to enable users to directly register models that were developed in SAS code into the SAS Metadata Server. Once in SAS Metadata Server, the common model data
can be accessed by SAS products such as SAS Model Manager, SAS Enterprise Guide, and SAS Data Integration Studio.
- The %AA_MODEL_EVAL macro has been added in order to compute lift and Receiver Operating Characteristic-type measures on any data set that contains probabilities and events. The %AA_MODEL_EVAL
macro uses a two-stage approximation algorithm to calculate the model performance measures.
- Batch code from the SAS Enterprise Miner client can be used more easily with input tables of different names and locations. The SAS Enterprise Miner 12.1 batch code now integrates
project start code that you can use to define libraries and options.
- The SAS Enterprise Miner SAS Code node for version 12.1 contains enhanced support for score code that contains SAS Procedure steps.
- PMML scoring for the Decision Tree, Regression, Neural Network, and Clustering nodes has been promoted to production status. New experimental functions have been introduced
for general regression and scorecards in the SAS Enterprise Miner 12.1 release.
- SAS Enterprise Miner Credit Scoring
- The Interactive Grouping node user interface has been redesigned to provide improved usability, performance, and computational scalability.
- The Interactive Grouping node has added a new Calculated variable role.
- The Scorecard node features an output variable that counts the number of adverse characteristics, and users can select named input variables for adverse characteristic reporting.
- The Scorecard node now processes indeterminate outcome values.
- SAS Enterprise Miner Applications
- The Gradient Boosting node now provides users with the capability to disable the H statistic calculation, resulting in improved run-time performance.
- The Incremental Response node has been promoted from experimental to production status.
- The Time Series Data Mining nodes have been promoted from experimental to production status. The production Time Series Data Mining nodes have been redesigned to provide greater ease of use
as well as improved performance and scalability. It is no longer necessary to precede Time Series Data Mining nodes with a Time Series Data Preparation node. Time series data can now be processed
with or without a numeric TimeID variable. Time Series Data Mining now supports sequence data. Season and Trend information is now extracted and included in Time Series Data Mining results.
- The Decision Tree output displays have been enhanced to display variable precision values in the split branches and nodes. This change improves the usability of the decision tree tool when mining with
extremely large and extremely small values.
For More Information
|
|
SAS/OR 12.1 Highlights
- SAS/OR enables multithreading, employing multiple computational cores on the same computing platform. This has a great impact in several areas of optimization, including problem creation in
OPTMODEL, nonlinear multistart optimization, and the new decomposition-based algorithm for linear and mixed integer optimization.
- The OPTMODEL procedure adds a SUBMIT block, which enables you to run other SAS code (including calling other procedures) within PROC OPTMODEL. You can perform analyses, display results, or
perform other tasks as integral parts of the process of creating and solving optimization models with PROC OPTMODEL.
- The new OPTNET procedure introduces a versatile set of algorithms for analysis and optimization of network structures (nodes and connecting arcs) that often occur as prominent
components of optimization problems.
- SAS Simulation Studio dramatically enhances your ability to work with large models and experiments, introducing new navigational tools, a new search capability, and an alternate Experiment
window layout that accommodates large numbers of factors and responses.
For More Information
|
|
SAS/ETS 12.1 Highlights
- PROC QLIM now provides users with Bayesian estimation methods for most univariate models supported by the procedure. The QLIM procedure allows users to estimate models of limited
dependent variables such as logit/probit, ordered logit/probit, Stochastic frontier, Truncated regression and Censored regression. The main features of the additions include:
- choosing prior distributions
- tools to initialize and tune the MCMC algorithm
- multithreaded Metropolis sampling
- convergence diagnostic tools
- The PANEL and AUTOREG procedures now offer convenient options to produce variance-covariance matrices for
estimates in the presence of heteroscedasticity and/or autocorrelation of unknown form. HAC estimates of standard
errors enable OLS estimates to be used for inference despite the presence of heteroscedasticity or autocorrelation.
- A variety of new model specification tests have been added to the PANEL and AUTOREG procedure. These tests
are used to check the statistical assumptions of the models, concerning stationarity, cointegration, and structural
change. These tests include p-values generated by high performance simulation methods. Many software packages report
only selected critical values for these tests.
- Many new ODS plots have been included with the SAS/ETS 12.1 release, including:
- Predictive and diagnostic plots for PROC COUNTREG
- Predictive and diagnostic plots for PROC QLIM including Bayesian plots
- Q-Q Plot for PROC SEVERITY
- Residual and structural break analysis in PROC SSM
- Dependency plots in PROC MODEL
For More Information
|
|
SAS Text Miner 12.1 Highlights
- The new Text Rule Builder node enables you to do predictive modeling directly from the term-by-document matrix, thereby allowing user-assisted or ?active? learning. You can use the Text Rule
Builder node to create rules that can be exported to SAS Content Categorization Studio.
- Improvements to previously existing text mining nodes include enhancements to the Text Filter node and viewer, the Text Topic node and viewer, and the Text Cluster node.
- The AFINN_SENTIMENT data set in the SAMPSIO library is new in this release. It contains two topics, ?Positive Tone? and ?Negative Tone?, that can be used as User Topics in the Text Topic node.
For More Information
High-Performance Analytics Updates
- Statistics: principal component analysis, finite mixture model analysis, canonical discriminant analysis
- Econometrics: panel data modeling, compound distribution models for loss simulation, multivariate copula simulation
- Text Mining: topic rotation
- Data Mining: cluster analysis, support vector machines, Bayesian network classification, generalized linear models