Hi all,
I have seen most of the discussions related to Sankey Diagram in the forum and elsewhere.
Have a question on the same -
Can we build Sankey Diagram to show the flow of count of id's from one 'Region' to 'Counties' in that region. I tried to use count but then I am facing an error when I try to write a calculation for
the 'pad' variable.
COUNT(if ([id]) = { FIXED [Region],[County]: MIN([id])} then 1 ELSE 49 END) this does not make sense.
For the examples that are shown the variable chosen is for example sales - which is a measure and it makes sense.
Can I get some advise if I can build a Sankey for my situation, ie. to show how the count of id's in a 'Region' get distributed over the 'County' in that region.
If not can you advise a chart type that I can use.
Thanking you all
Regards,
RS
When working with record counts in a Sankey diagram, I substitute a Count function wherever you would typically use a Sum function. My field calculations based upon Art of Visualization's very helpful video are below.
// VizSide
if [Table Name]="Orders" then "Region" else "Segment" end
// ToPad (SET TO DISCRETE BEFORE CREATING PADDING)
if [VizSide]="Region" then 1 else 49 end
// Padding - Right-click ToPad and select "Create bins," size: 1
// t
(INDEX()-25)/4
// Rank 1
RUNNING_SUM(COUNT([Orders]))/TOTAL(COUNT([Orders]))
// Rank 2
RUNNING_SUM(COUNT([Orders]))/TOTAL(COUNT([Orders]))
// Sigmoid
1/(1+EXP(1)^-[t])
// Curve
[Rank 1]+(([Rank 2]-[Rank 1])*[Sigmoid])
// Field sizing
WINDOW_AVG(COUNT([Orders]))
// Tooltip percentage
([Field sizing]/TOTAL(COUNT([Orders])))*2
