Guys, is there a way that the y axis gets auto fixed when I change the numbers? The problem is that one needs to maually change the axis.
Joe -- a big THANKS for the posts on the waterfall charts. I needed to construct one fora big dataset (28MM records) rather quickly. Following your posts I had it done in 15 minutes. Props to you and the others for helping we newbies get to the finish line quickly.
I have just stumbled on your article and was reviewing the technique you have used to create the waterfall charts. Impressive indeed!
Can you help me please? On your example from a previous version of Tableau, the representation of Ending Cash seems to be wrong. Do you know how I could get Ending Cash to be grounded? i.e., I would expect the bar for the Ending Cash to start from the axis going upward.
Waterfall.twbx.zip 117.3 KB
Good discussion, very useful.
So now I know how to do a waterfall with data shaped as follows -- basically, use category on the columns, count along the rows, do a running total along the rows, and then Gantt bar... and it works very nicely.
Category Count Time1 32000 Gains 4500 Losses 6500
I'd like to know how to do a waterfall when the data are shaped like this, ie, using only measures and their values
Time1 Gains Losses 32000 4500 6500
Intuitively it would seem that it could be done, but I don't see how (unless I missed something in the discussion).
I know I can reshape the input data outside of Tableau, but for a number of reasons I want to avoid that. Is there a way to do this, using only measures and their values?
Appreciate any guidance.
Any way to do a waterfall using different measures?
Suppose I have a set of different measures in my DB for each date/line:
- Gross Profit
- Net Profit
Now of course in Tableau this makes great sense as a data file, given that I can sum or average my revenue, cogs, etc. over any given period of time. Yet if I want to visualize this as a waterfall where Revenue builds down to Net Profit (which is often requested), is there any way to support this?
In Excel, the methodology would be to create a series of dummy values that are 'transparent' on which the rest of the data sits. I've been trying to think of how this might work in Tableau, but haven't been able to work out the dynamics
I have a similar question to Brian (if not the same):
I need some help to develop a financial waterfall in Tableau. The attached package workbook shows how it should look.
I have seen several solutions to do waterfalls in Tableau, but they work when the data is stacked (one column for all the measures, one column for all the Mesure names).
This is not a viable option in our case, as we have the data structured in a way in which each column is a measure. We can't stack the data because our DB has millions of records.
Note: the sample data used just has one row of data, but in reality, the DB has millions of records.
I have a need to do a waterfall on metrics that need to be aggregated at various levels of granularity (so that the data can continue to be sliced by different dimensions in Tableau and show accurate margins in waterfall viz at various levels of granularity) and then reshaped into Tableau's required format for waterfalls. My application data is in pivoted format in sql server db.
To date, in order to create a waterfall on computed margins I have aggregated various margin elements in my sql server db and then am doing an unpivot to get the margin elements in the format Tableau requires for a waterfall (amount_type, value as columns). The issue is that I cannot do the typical filtering in Tableau as the margins computed from required transactions elements and then pre-aggregated at the lowest level of granularity in my sql server data source (I had to choose a level to aggregate at, and am using that as my data source for the waterfall in Tableau), so I lose the ability to slice by various dimensions and have the margin elements be aggregated to the appropriate value by Tableau.
Do you, Joe, or anyone know of a simple way to solve this? Does Tableau 9 level of detail calcs provide a solution for this? We are currently on 8.2 and may not move to Tableau 9 for a while but wondered if anyone has encountered this issue or knows if Tableau 9 solves this somehow.
I can try to design an example without sharing application data if that helps but hoping my description of the problem will generate some questions or thoughts on the problem.