Default Break By behavior should be Columns instead of Rows.
Summary: Chris Wallingford, Chris Wallingford, discusses an issue with the default behavior of pivot and line widgets in data visualization. When converting a pivot widget with a Running Sum (RSUM) over time to a line widget, the system defaults to breaking by rows, resulting in unexpected negative curvature in the line chart, leading clients to open support tickets. Chris suggests that defaulting to breaking by columns would align better with client expectations. He also notes a lack of clear use cases for charts breaking by rows, indicating a preference for column-based breaks in such visualizations.
Create a pivot widget of RSUM( [Total Quantity] ) over time (rows) and break by (columns) age range.
Scanning down each column, over rows, the numbers only increase. Now change it to a line widget.
The break by defaults from an RSUM within columns in the pivot to an RSUM within rows in the line chart. This is the same as not having an RSUM and doing a stacked area.
Our clients expect the line chart the results from changing the pivot to a line chart, to reflect the same output as the source pivot. As it stands now, they are surprised to see negative curvature in their RSUM, think something is broken, and open a support ticket. The default behavior should be to render Break by Columns instead of Rows.
As an aside, I'm having trouble coming up with, and would be interested in hearing, use cases for charts with break by set to rows.
Chris Wallingford
OP1 year ago · EditedImages lost during community platform updates...
Image of Break By setting showing the default Rows option selected.
Very busy image showing a Pivot widget with Running Sum columns vs a Running Sum row. Followed by Line widgets showing how the two Break By options manifest.
Defaulting to Columns aligns with what users expect.
Chris Wallingford
OP2 months agoSo very many clients complaining about this. More every week as we continue to migrate our clients into Sisense Cloud.
Consistent confusion as to why an RSUM({positive values}) would have negative curvature and Sisense just looking broken because of it.
Any chance of this default changing?