You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Your case 1 above would actually be closer to waterfallmode="relative" to match the current barmode="relative" case... it's kind of ugly but it has a certain logic to it I'll admit :)
I am very interested in the possibility of a Plotly Stacked Waterfall. The image posted by jdb78 is exactly what I am looking for. I think this solution would fill a large gap in Python’s data visualization capabilities and attract more developers to Plotly.
Can you please advise me on how I can make this dream become a reality? :)
@whippysundew I think we still need a little more information about the kind of data it would be useful to represent with such a stack, and come to a consensus on the API for this that's most consistent with our other traces such as bar. After that we can look back at previous waterfall PRs such as #3531 to see where the code needs to be amended.
@alexcjohnson thank you for your insight. Let’s use software sales data as an example with waterfall columns as follows: YTD revenue (+), won business yet to hit revenue (+), net pipeline (pipeline + business at risk, + or -), total projected performance (+). The stack would be helpful to display different regions or lines of business while still representing the total performance. It would be very beneficial to allow positive or negative values in each column - say one region has a positive net pipeline while another is negative. The chart above shows how this can be displayed.
Definitely a useful feature to add, I was trying to showcase the capital returns of multiple companies, so you have initial assets in t0 and final assets in t1, just like the Plotly waterfall example. It would be great if those initial and final bars could be dissagregated into multiple stacked bars. Just a thought.
Activity
nicolaskruchten commentedon Dec 29, 2019
I see at least two variants:
xandmeasureand sign for ally(could be implemented with awaterfallmodeofstack)yvalues (of the same sign) for a givenx(could be implemented with no new attributes in principle?)nicolaskruchten commentedon Dec 29, 2019
In principle we could apply the same stacking/overlapping rules we have for
barin both cases... ugly but well-defined.And for case 1, we would just always use the
measurefrom the first trace in a given subplot, to avoid conflicts.jdb78 commentedon Dec 30, 2019
I like the multiple traces with
waterfallmode="stack"as it is consistent with other other trace types.if the y values are of different sign, we could show something like this:

We only stack for overlapping
x(consistent withbar) andmeasureWe allow
nullvalues forywhich prevents the any hover animationsWhat are your thoughts?
nicolaskruchten commentedon Jan 7, 2020
Your case 1 above would actually be closer to
waterfallmode="relative"to match the currentbarmode="relative"case... it's kind of ugly but it has a certain logic to it I'll admit :)whippysundew commentedon Jun 21, 2022
Hi Everyone,
I am very interested in the possibility of a Plotly Stacked Waterfall. The image posted by jdb78 is exactly what I am looking for. I think this solution would fill a large gap in Python’s data visualization capabilities and attract more developers to Plotly.
Can you please advise me on how I can make this dream become a reality? :)
Thank you
alexcjohnson commentedon Jun 27, 2022
@whippysundew I think we still need a little more information about the kind of data it would be useful to represent with such a stack, and come to a consensus on the API for this that's most consistent with our other traces such as bar. After that we can look back at previous waterfall PRs such as #3531 to see where the code needs to be amended.
whippysundew commentedon Jun 28, 2022
@alexcjohnson thank you for your insight. Let’s use software sales data as an example with waterfall columns as follows: YTD revenue (+), won business yet to hit revenue (+), net pipeline (pipeline + business at risk, + or -), total projected performance (+). The stack would be helpful to display different regions or lines of business while still representing the total performance. It would be very beneficial to allow positive or negative values in each column - say one region has a positive net pipeline while another is negative. The chart above shows how this can be displayed.
etiennecelery commentedon Jan 13, 2023
Definitely a useful feature to add, I was trying to showcase the capital returns of multiple companies, so you have initial assets in t0 and final assets in t1, just like the Plotly waterfall example. It would be great if those initial and final bars could be dissagregated into multiple stacked bars. Just a thought.
NolsN74 commentedon Nov 14, 2023
any news on this issue?
yogeshriyat commentedon Mar 26, 2024
Hi any update on this stacked waterfall chart?
[-]Stacked waterfall charts[/-][+]add stacked waterfall charts[/+]