Hi Expert,
I was wondering if I can get rid of “0” variance on the waterfall chart?
Thank you very much!
Turn on zero supression in the widget cube settings.
(Note: due to floating point arithmatic if the value is the result of a rule calculation or a more complex consolidation then the value might not be a true zero but a very small decimal value and would still show under zero supression. If this is the case you would need to apply or adjust rules to round small values to zero.)
@cw-ch-scott I have tried it but it only works for column with 0 value
Doesn’t work for non-0 column where variance = 0
But you are saying that it should work in both case? Whatever showing as 0 seems due to the floating point arithmatic issue?
Thanks!
In this case you could use the waterfall-single widget type
Yeah we need the normal waterfall one 
In this case the logic of the chart is correct. The purpose of the waterfall visualization is to show where variabce is coming from between 2 observations for a set of defined categories. Some of the categories aren’t a source of any variance and this is what the chart shows. If you don’t want to display these categories then don’t include them as columns.
The waterfall-single type assumes the first column as the starting point, the last as ending point and all in between as variances. If you use this widget type then zero supression on columns will automatically hide columns with no variance.
If using the standard waterfall then the sum of the first row is calculated as the starting column and the sum of the second row as the ending column. The variances between row 1 & row 2 items are then calculated as the waterfall steps. Note that you could still relatively easily with MDX supress steps out of the view with zero variance. However, this would also effect the starting and finishing values.
For what you want to do you need the single dimension waterfall. Simple as that.
1 Like