• Have you ever had issues with tags not recalling marked data?
  • Have you named key columns but still lost tags when the data updated or changed?

A few months ago, I wrote this post explaining the purpose and utility of tags, so if you are unfamiliar with tags in Spotfire, give that post a quick read.  This post will specifically discuss and deal with making sure that the marked data is recalled.

In the earlier post, I noted the following….

…markings will NOT be retained in the tags after the file is closed and reopened unless keys are specified in the Edit menu, Data Table properties.  The user must choose one or more columns that define the uniqueness of a record in order for the markings to stay attached to the tags.

It’s only by defining the uniqueness of a record that Spotfire is able to recall a marked record.

Now, that’s all fine and good, but what if you data set changes?  A user I work with encountered this problem recently.  He had defined his key columns in order to retain marked data, but then his data set updated, and what he had defined as the key columns were no longer correct, and he lost all of his marked tags.

There is a solution to this problem — Lists.  Lists are another useful feature in Spotfire that I haven’t blogged about just yet, but you can bet I’ll be doing that soon.  List are available via a panel that you can turn on in the toolbar, and the panel will appear on the left hand side of the screen.  Lists are also used to capture and recall marked data, but they are different from tags.

The key differences in list are:

  • Lists are not stored in the analysis file.
  • Because lists are not stored in the analysis file, you will see lists in all analysis files created by a single user.
  • Marked data stored in lists does not disappear when the file is closed and reopened.
  • You can create tag collections from lists.  

That last bullet point is the most import.  Because you can create tag collections from lists and the marked data stored in lists does not disappear when the file is closed and reopened, lists can be used to restore tag collections.

In order for this to work, you must:

  • Create lists at the same granularity as the tags, which may not be possible depending on your data set.
  • Make sure the content of the list(s) is the same as the content of the tags.  This seems obvious, but if you make a last minute update to tags and don’t update the lists, not all tags will be restored, or they won’t be restored correctly.

The solution is a bit manual, but it works and is less manual than recreating lost tags.  Here’s a video that demonstrates.

Written by Julie Schellberg of Big Mountain Analytics, LLC
Residing in Whitefish, MT, an analytics partner with Ruths.ai. Specializing in Spotfire analytics, dabbling in Power BI and R.