What’s New in SAS LASR Analytic Server

Overview

SAS LASR Analytic Server 2.6 includes the following changes:
  • encryption for SASHDAT tables
  • enhancements to the IMSTAT procedure
  • enhancement to distributed servers for high-volume access to smaller tables
  • support for the MapR Distribution for Apache Hadoop
  • support for Teradata with SAS In-Memory Statistics

Encryption for SASHDAT Tables

The following enhancements enable you to work with encryption for SASHDAT tables:
  • The SAS Data in HDFS (SASHDAT) engine is enhanced to support ENCRYPT=AES and ENCRYPTKEY= data set options. For environments that use a SAS Metadata Server, the engine supports a SIGNER= option and a SIGNERFILEPOLICY data set option. The SIGNER= option enables the engine to perform the passphrase exchange with the SAS Metadata Server that is needed to encrypt and decrypt the data. The SIGNERFILEPOLICY option ensures that if encryption settings are registered with the library in metadata, that they are applied to the SASHDAT file.
  • The LASR procedure supports the ENCRYPTKEY= option that can be used to specify the passphrase for the SASHDAT table when it loads the table to memory. In metadata environments, the procedure supports the SIGNER= and SIGNERFILEPOLICY options that are used to apply encryption settings that are registered with a library in metadata.
  • The IMSTAT procedure is enhanced to support the ENCRYPTKEY= option and SIGNERFILEPOLICY option in the SAVE statement.

Enhancements to the IMSTAT Procedure

The following enhancements have been made to the IMSTAT procedure:
  • The HYPERGROUP statement can perform a number of analytics on data interpreted as a graph (vertices and edges). Features include several forms of structural analysis, the calculation of vertex centrality measures, and layout in 2-D or 3-D space.
  • The TRANSFORM statement can perform a variety of analytic data preparation tasks. The features include imputation, outlier treatment, functional transformation, binning, and output.

High-Volume Access to Smaller Tables

High-volume access to smaller tables in a distributed server is enhanced. When loading tables to the server, administrators can specify a number of machines to act as non-distributed servers for a table. Requests for the table are load balanced to the non-distributed servers.

Support for the MapR Distribution for Apache Hadoop

The MapR distribution includes the MapR NFS service. The following list highlights the interactions with SAS LASR Analytic Server:
  • The machines that are used for SAS LASR Analytic Server are configured to mount the cluster using NFS.
  • The SAS Data in HDFS engine can connect to the root node of the SAS LASR Analytic Server and distribute data as SASHDAT files. The data are transparently written over the NFS connection to the MapR cluster.

Support for Teradata with SAS In-Memory Statistics

The second maintenance release for SAS In-Memory Statistics 2.5 adds support for reading data from a Teradata appliance into memory on a SAS LASR Analytic Server. Support for reading data from Hadoop continues.

Documentation Enhancements

The sample code for showing the sequences table in the ARM statement section of the IMSTAT procedure is changed to include the ITEMSETFILTER=ALLITEMS option.
The syntax for the PREDICT statement in the RECOMMEND procedure is updated to show that the USERLIST= or USERDATA= options are required.
Copyright © SAS Institute Inc. All rights reserved.