Recognize negative values?

[ruthpozuelo]

Describe the bug
Don't know if this is a bug, a feature request or I simply don't know how to do it.

I am trying to create a diverging lollipop charts with positive and negative values. The negative values are plotted outside the plot segment and when I save the chart in power bi, they all get moved to the positive area (because they are plotted outside?).

To Reproduce
Steps to reproduce the behavior:

Expected behavior
I have tried to use a scale, different glyphs (that can be worked out, but it is not optimal), I would expect Charticulator to understand that there are negative values and plot them as such?

[zBritva]

It's not bug. It's limitation of current workaround (rendering legend ax axis).

[ruthpozuelo]

Thanks, so in order to plot the negative values we need to create two glyphs?

[zBritva]

I think the Data Axis is better for now.

I used list(0) to constantly define 0 and bound bottom side of rect to this anchor point.

See chart file
CAT__VAL.chart.txt

[ruthpozuelo]

Thanks, but how would that work when I want to do a lollipop chart?
My data:
data.csv.txt
Image of what i am trying to create:

[dm-p]

Hey @ruthpozuelo - this is not 100% perfect, but should get you on the right path:

Should be apparent from the screenshot, and I've also attached a .pbix (zipped), but here's how I solved using your data:

1. I created a measure for Pop Change as sorting didn't work too well with the column, but dataset looks as you might expect:
   

2. Add Country & Value Type columns and Pop Change measures to Charticulator dataset.

3. Sort data in visual header by Pop Change / Ascending:
   

4. Set plot segment to stack Y.

5. Add data axis to glyph, across the whole width, and add Pop Change measure. Additionally, drag the Pop Change measure on again, and change this to list(0) measure, as per @zBritva's suggestion and add to axis e.g.:
   
   

(you could also create a measure or column called Origin or similar, with a fixed value of 0 and add that to the dataset, which was how I approached initially if the list(0) approach is not ideal for you)

7. Add line, and set span from origin measure to Pop Change measure. This will ensure that for whatever direction of the origin, the line will be in the correct direction.

8. Add symbol and snap to the Pop Change measure on data axis. Should now look roughly as follows:
   

9. Add text mark, bound to origin measure on data axis. Set Anchor X to Right and Anchor Y to Middle. Add Pop Change to visibility condition and set to 'P', so this only shows for glyphs with that value:
   

10. Repeat step 9, but set Anchor X to left. When you drag Value Type to visibility, HOLD SHIFT TO CREATE A NEW SCALE, otherwise it will share the same condition as the previous one and won't work as desired (doc link, under 'Reusing a Scale'). Set this to show on 'N':
    

11. I also added a vertical line, bound to the Origin measure and set Plot Segment gap to 0%, so that the 0 line runs all the way up, e.g.:
    

[ruthpozuelo]

Fab.u.lous!! You are a 🎇!

[ruthpozuelo]

A million years later and I have finally tested your technique and it works as a charm, thanks @dm-p and zBritva.
In an attempt to understand how charticulator works, I have a question:
why do I have to set the zero baseline to make it work?

[dm-p]

As the line needs a start and an end point, it's giving the line a consistent anchor for all series that serves as a (data-bound) guide. This baseline could easily be another measure (e.g. if you were designing a dumbbell chart) and could also be a measure or column hard-coded to 0.

[Prakash-PBI]

I have tried with list(0) to show negative bars but still its not working for me,

Could you please add some more content.
Thanks

[FShawky]

Make sure your scale values are not set to auto-update. I've found that when you hit save, it goes back to the values it originally had.

[zBritva]

@FShawky It needs to uncheck Auto min value and Auto max value to keep axis with constant values: