Release Notes v6.12

What's new?

• SWITCH pattern, which generates a value based on the evaluation of an ordered sequence of possible values against the same field, each with its own result.
• Added an optional REDSHIFT_IMPORT_MAXERROR_PARAMETER environment variable, which if set will dictate the Redshift MAXERROR parameter for all Redshift S3 imports.
• Added an optional LOG_LEVEL environment variable, whose allowable values are error, warn, info, and verbose. This variable will dictate the application logging level and is cumulative; for example, if set to "warn" then the application will log errors and warnings. The default value is "info".
• For users who are granted access to view Ursa Studio-generated SQL, a Request History screen is available in the My Account area of Analytics Portal, which shows recent requests made from Analytics Portal, along with details that might help with troubleshooting.

  There are a variety of different request types in Analytics Portal, which are described in the request history with terms like measure-explorer or case-review-count. Within each request type, the method used to serve the request is also noted, which can help shed light on the many layers of caching and optimization that Ursa Studio uses to serve requests to Analytics Portal.

  Ursa Studio keeps a cache in the database of responses for requests made by Analytics Portal users. When another user, or the same user in a new browser session, requests the same analysis, the results are served immediately from this cache, instead of requiring the analysis to be re-computed. The results of all the charts in Analytics Portal are saved in this cache, as are the row counts as displayed in the tabs at the top of the case review data table. Requests served by this mechanism will show "cache" as the method.

  Long-form measures and interval measures display a view of the data that has been rolled up to the case level; the results reported to the user in Case Review thus are not simply one row per row of the underlying measure table. In order to eliminate the need to calculate this grouping on a per-request basis, every report run creates a companion table to the underlying measure table whose grain size is one row per case, and this table will be used whenever possible so as to minimize processing. Requests served by this table will show "case-grain" as the method.

  Measure explorer (as well as similar chart-driven screens such a report home) leverages a dimensional table, also automatically created during the report run, which is a cube across every field designated on the measure as being analytic. This table is often much faster to query and is used whenever possible. Requests served by this mechanism will show "analytic-cube" as the method.

  There are circumstances where it would be impossible or unnecessary to use these optimized tables; in these circumstances Ursa Studio automatically routes this query to the raw measure table, and "standard" will be shown as the method.

What improvements have we made to existing features?

• Allow range import of multiple files even if date imputation is not activated.
• Allow range imports to work with "Latest file matching" date resolution for import objects.
• Data effective date is no longer a required field for import objects. If this value is left blank, the date of the import will be considered the data effective date. This now allows for import objects without imputed dates to be run directly from the Object Workshop dashboard, or added to a Saved ELT. If a multi-file range import is set to impute the data effective date from the filename and a mid-range file has an invalid or missing date, the effective date will be set to the date of the import.
• Added an optional email notification list for scheduled events.
• Moved "Logout" and "My Account" links in the topbar to a dropdown menu to be spawned by clicking the user avatar in the topbar. Added an "About Ursa" option to the same topbar to show user details about the deployment including version number, database type, and deployment type.
• Allow users to save cohorts from Case Compare tabs.
• Calculations for "case compare" mode in case review will ignore hidden columns.