About Loading Data :: SAS® Visual Analytics 7.4: Administration Guide

Introduction

Users can easily import data. See Overview of Data Flow in SAS Visual Analytics in the SAS Visual Analytics: User’s Guide. This chapter documents administrative aspects of data loading.

SAS Visual Analytics uses data that is loaded to memory in a SAS LASR Analytic Server. Tables remain in memory until they are unloaded or the associated server stops. The following features help keep data available:

Optional Convenience Features
Feature	Trigger	Result (Automated Action)
Autoload	A time interval elapses	In-memory data synchronizes against a designated directory.
Autostart	A load or import is requested	The associated server starts.
Reload-on-start	A server starts	Participating tables reload.

Tip

See SAS Visual Analytics: Getting Started with Data on Windows for orientation and recommendations for beginners.

Load Methods

Load methods vary by data source.

Load Method	Data Source
Load Method	Spreadsheet or Delimited	SAS Data Set	SASHDAT File1	Data Server	Other2
Interactive load
Run a data query
Import from server
Import a local file
Autoload
1 Co-located HDFS or NFS-mounted MapR. See Distributed Server: Parallel Load.
2Data from Twitter, Google Analytics, or Facebook.

Reload Methods

Reload methods depend on how a table was initially loaded.

Reload Method	Eligible LASR Tables
Interactive reload	Tables that were interactively loaded (). Output from data queries.1 Output from LASR star schemas.1 SAS data sets that were imported from a server.
Reload-on-start	Participating tables from imports of local files. Participating tables from imports of Google Analytics, Facebook, or Twitter data.
1Any input LASR tables must be available (loaded).

Reloading data requires access to either the current source data or a backing store copy of the original source data.

Interactive reload runs against the current source data (using a job or query that was created by the initial load).

Note: Any data that was appended only in memory is not included in the reload.
Reload-on-start runs against a copy of the original source data (using a data provider library that functions as a backing store).

To make a table that is not reloadable available:

If the table was autoloaded, wait for the next run of the scheduled task.
Otherwise, repeat the action that initially loaded the data.

Data Compression

Effects of Compression

There is a trade-off between compression and performance. Compressing data conserves memory. However, it might take longer to retrieve data from a table that is compressed. See Data Compression in the SAS LASR Analytic Server: Reference Guide.

Support for Compression

Here is a summary of when compression occurs:

Administrators and data builders can request compression when they load a table.

Note: If you load a SASHDAT file, you cannot request compression. The compression setting that already exists in the source SASHDAT file is honored. An exception is that an encrypted SASHDAT file is always uncompressed when it is loaded. See On-Disk Encryption of SASHDAT Files.
Administrators and data builders can request compression when they add a table to co-located HDFS or NFS-mounted MapR.
Administrators can request or remove compression for a LASR table, by using the Change Source action. See Replace a Source Table.
Data builders can request compression when they import a table or run a data query that outputs to a LASR table, co-located HDFS, or NFS-mounted MapR.
Users who have access to the Advanced panel (in the designer or explorer) can request compression when they import a table. See the Build Data capability.
Administrators can use an extended attribute to request compression of autoloaded data. See VA.AutoLoad.Compress.Enabled.
In reload-on-start, compression is used for tables that were compressed when they were initially loaded.
Compression does not occur for small tables.
Compression does not occur for tables that are loaded from encrypted SASHDAT files.

Table and Column Names

In general, names can include spaces and special characters. Exceptions include the following:

For interactions with third-party data sources and operating systems, third-party name limitations apply.
For LASR table names, the period character (.) is not supported. If you load a SAS data set that has a period in its name, the period is replaced with an underscore (_).
SAS name limitations apply. See Summary of Extended Rules for Naming SAS Data Sets and SAS Variables in the SAS Language Reference: Concepts.

Note: When data is imported as a local file or autoloaded, any character that is not supported by SAS is replaced with an underscore.

Note: Column names in reports are case sensitive. If you change the case of a column name in a LASR table, then any reports that use that column will display an error message.