A data connection is used when you are analyzing massive amounts of data and you need to keep the underlying data in the database (in-db) rather than bringing it into Spotfire's internal data engine.
The data connection consists of two parts:
One or more data tables which may or may not be used by the analysis.
Both the connection data source and the data connection itself can be shared in the library, if desired.
Spotfire Analysis |
Data Connection |
Connection Data Source |
Shared in library |
Shared in library |
|
Shared in library |
Embedded in connection |
|
Embedded in analysis |
Shared in library |
|
Embedded in analysis |
Embedded in connection |
When your analyses use shared data connections you can easily update the data in a large number of analyses simultaneously by updating the data connection.
The data connections may either contain an embedded connection data source or a shared connection data source. When it comes to editing, a shared data source may temporarily be embedded in the connection, if necessary.
If your analysis contains a shared data connection you may still be able to use a cached version of the data connection if you are offline. However, this may result in inconsistencies with the expected data from the data connection on the library if the cached data is out of date.
Data Tables in the Connection
A data connection from a cube data source always consists of a single data table, which combines all of the information from the selected cube.
Data connections based on relational databases may contain one or more data tables. Each of the data tables may also be built by more than one of the source tables from the database, joined into a single view using structural relations. These relations may have been set up by the database administrator in the database but they may also be manually created in the Data Tables in Connection dialog.
The selection of which data tables should be available in the analysis is also done in the Data Tables in Connection dialog. In the image above the data connection will create two data tables and one of those is a view combined by two related source tables. Data connections shared in the library contain a specified set of data tables, but data connections that are embedded in the current analysis can be edited within the context of the analysis and the currently used data tables in the connection can be updated. You can always embed a data connection in the Data Connection Settings dialog if you need to make any changes to the data tables in this particular analysis.
See also:
Adding Data Connections to an Analysis