zuzana-vej opened a new issue #13717:
URL: https://github.com/apache/superset/issues/13717


   **Is your feature request related to a problem? Please describe.**
   Visually, higher cardinality (6+) dimensions are difficult to parse and pose 
a challenge for data consumers in attempting to interpret what metrics 
increased and/or decreased for a time-series line chart or bar chart with many 
series. 
   For data/chart producers, high-cardinality dimensions make it a challenge to 
display a legend (as it consumes too much real estate and ceases to be useful 
as an interactive visual exploration tool).  See bar chart below.
   Moreover, when setting a series limit the total value of the metric for the 
time period is affected and reflects only the top x series limited to - leading 
to confusion for users about what is an accurate total. 
   
   **Describe the solution you'd like**
   Time-series Superset charts (line & bar) would include an ability to group a 
dimension by top-N values (where N is set by the chart creator), of this 
dimension (this is most meaningful when ordered by metric for the biggest 
buckets of the metric cut by this dimension). The remaining dimension values 
would be aggregated in a single ‘Other’ grouping. This could be complicated by 
several dimension cuts to group by, but would be a modification and enhancement 
of the series limit feature. This type of feature is also a standard in BI 
tools such as Tableau.
   
   Example:
   ![Screen Shot 2021-03-19 at 3 11 13 
PM](https://user-images.githubusercontent.com/61221714/111847158-7fdc2200-88c5-11eb-90dc-2fa94b457b72.png)
   
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

For queries about this service, please contact Infrastructure at:
[email protected]



---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to