Omniscope's new feature, available in daily builds from May 2023, will bring joy to both analysts and report viewers.
The report viewers will be able to pick and try different fields for comparison grouping and measure, in addition to filtering and selection interaction.
This will enrich their UX and give them more options for drill-down, as they will be driving the exploration process and determining what kind of comparison and metrics they are interested in.
On the other hand, the analysts will love the feature because it provides them with new efficiency tools: they will need fewer tabs to deliver their analysis and, importantly, will have a more complete centralised Data Model repository, in addition to the existing data sources, derived data sources (mini workflows with dynamic results), variables and hierarchies.
If you are too excited to read the instructions - you can watch the video explaining it all.
What are Named Groupings and Named Measures
Before we get into the named groupings and named measures, let's clarify what is a grouping is and what is a measure.
A grouping is a way to break the data down into groups or 'data pots'.
If we had a bar view, then one grouping option that we could define is the X axis split and for example compare products, countries or categories.
Other places in Omniscope where the analyst may wish to define groupings are bar stacking and panning / tiling in most other views.
A measure is a metric we use to compare our data groupings - we could compare our products by using sales in $, weight in kg, units sold, unique value counts or any other method we need for the comparison.
Named groupings and Named measures are a method to centrally configure and manage some of report groupings and measures, while avoiding the need to edit the charts manually one by one. The mechanism can be set up so it allows both the administrator and the report viewer to control the measure/split changes.
The mechanism is configured in the Report's Data Model and can be made available to charts on one or more tabs. Another benefit is that the Data Model now more efficiently represents a central repository for dynamically calculated fields and chart-level formulae (with calculations responding to filtering and chart split/measure switching).
There are multiple use cases where the analyst - report creator may wish to utilise centralised grouping or measure control, however there are two distinct principles:
- Viewer making the split/measure switches: the report is configured so that Groupings and Measure choices are exposed and available to the viewer, as a method of enhancing their drill-down capabilities and optimising report structure, while avoiding the need to create multiple tabs with different split/measure combinations
- Analyst making the switches: report Administrator is utilising the central repository as a method to quickly customise multiple reports and create variations. The controls in this case are NOT exposed or accessible to the report viewer.
Centralised Data Model
Omniscope's ETL workflow can be used to create multiple reports in the same project. While these reports can share the data prepared in the same ETL process, they can also have a data model of their own, with the new dynamic calculations, mini workflows, hierarchies, variables and named measures/groupings.
To access these features open your report in edit mode and click on the Data Sources button (alternatively access it from the 3-dot Report menu).
All the data sources connected to the report in the workflow will appear as Input 1, 2, 3 etc. These inputs can be combined to create new 'derived data sources', which are result of the mini workflows, where further querying and calculations are performed on these inputs.
How to Create a Named Grouping or Grouping Choices
- Click on the Groupings tab in the Data Model dialogue and select the relevant data source.
- Add a Single field named grouping or a Grouping Choice, then click on the 'Add choice' button next to it to pick one or more fields.
Note: Each grouping field has a toggle button next to it, where more config choices can be found, similar to the chart's split 'pill' options e.g. 'fit to screen', bar numbers limit and 'othering' behaviour.
- Pick the choice type for this grouping: Single for the choice that allows the viewer to pick one split at a time (from multiple fields on offer) e.g. one field for the Bar X axis
Multi-select choice allows the viewer to pick more than one grouping field to be shown in the chart: in some cases you may wish to display on the X axis two field groupings e.g. first group the data by product, then by category. When you add this multiple choice option to a named grouping it will automatically convert it to 'grouping choice', meaning the viewer will have multiple field options to choose from.
Important: Both single and multi-choice options will require you to add the Named Grouping filter to the Filter devices on the tabs where you wish to expose the choices to the report viewer.
'None' choice type is used for those groupings that are not made available to the report viewer. This type of split can represent a field that is used in multiple charts and formulas in the report. If the analyst wishes to change the field and customise their report - this switch will happen in one location, but will affect all the charts and formulas where this named grouping choice is referenced.
How to Create a Named Measure or Measure Choices
The steps are similar to grouping process, with one interesting difference. You can define in the Named Measure (or Choice) one or more formula measures, that will be applied to the data subsets created by different chart settings.
If the formula is calculating a mean invoice value, and this formula is used in a Bar chart where the split field is [Product], then formula will produce different mean for each of the products in the dataset. If in another chart this same formula is applied with the [Country] field split - it will return a different mean for each country.
This formula will be available to all charts on the particular tab, or to all charts in the report. Any updates in the central repository will instantly be reflected in those charts.
Further to this - each Named Measure or Grouping can be tab-specific or Global - see the image below. This option can be accessed from the cog settings button next to each named choice.
Use case: in some scenarios it may be useful to expose the measure/grouping choices on one tab, and for that choice to have an effect on all referenced visuals in the same report. In some other cases, tabs can be created to have independent choices mechanism, and to allow the user to select different choices on multiple tabs.
Default choice allows the analyst to specify which field (or formula) the choice will revert to, if the choice is reset from the filtering menu.
"Allow none" option is a useful tool: it is applicable to those choices that will not be accessible to the report viewer, however there is also a use case in the Named Grouping. If 'none' choice is ticked it will give an option to the user to refer to the whole dataset, without any groupings, and instantly view the measure applied to all data.
There is also an easy way to create a Formula measure (formula created on a chart level), then add it to the Named Measures repository by clicking on the icon in the chart measures menu (see image below). This will also work if you wish to add a field to the named choices from this chart config menu.
Application: Reference Named Groupings, Measures and Formulae in Reports
Now that the Named Groupings and Measures are configured it is time to plug them into our report!
Step 1: in the chart configuration 'shelves' pick your choices from the menu in the Split, Measure, Stack, Size or Pane drop-down list.
(see image above - the purple Bar chart)
Step 2: Add the relevant Named Measure or Named Grouping to the Filter devices on the tab where you wish the report viewer to be able to interact with the chart settings.
(see image below)
Dynamic Axis Titles
If you choose to display the chart's X/Y axis titles, where applicable, you will notice how the axis titles are changing with every Named grouping and measure change, making these titles pick the new field or formula name. This is why you may wish to give the formulas and fields referenced in this context short and meaningful names.
Was this article helpful?
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
We appreciate your effort and will try to fix the article