How to annotate difference between bars?

Question

I'm trying to show difference between bars using annotation. Specifically, showing difference between all bars with respect to the first bar.

My code is shown below:

import plotly.graph_objects as go


lables = ['a','b','c']

values = [30,20,10]

difference = [ str(values[0] - x) for x in values[1:] ]


fig = go.Figure( data= go.Bar(x=lables,y=values,width = [0.5,0.5,0.5] ) )


fig.add_annotation( x=lables[0],y= values[0], 
                    xref="x",yref="y",
                    showarrow=True,arrowhead=7,
                    ax = 1200, ay= 0  )

fig.add_annotation( x = lables[1], y=values[0],
                    xref="x",yref="y",
                    showarrow=True,arrowhead=1,
                    ax = 0 , ay = 100,
                    text= difference[0]
                     )


fig.show()

The result graph looks like:

As you can see, I'm trying to use annotation to indicate the difference between a and b. But I don't know how to get the vertical distance between the horizontal line from a and the top of the b.

I'm trying to have an arrow point to the top of b and c from the horizontal line. I'm wondering is there a way to get this vertical distance or are there any other ways to achieve the same result?

Answer 1

The vertical distances can easily be obtained with:

diffs = [max(values) - v for v in values]

The only real challenge is getting every parameter of fig.add_annotations() right. The correct combination of yanchor, ax and ayref will give you this plot:

You can take a closer look at the details in the snippet below. And if you don't like the placement of the numbers, we can fix that too.

Complete code:

import plotly.graph_objects as go


labels = ['a','b','c']

values = [30,20,10]

diffs = [max(values) - v for v in values]

diff_labels = dict(zip(labels, diffs))
#print(diff_labels)

fig = go.Figure( data= go.Bar(x=labels,y=values,width = [0.5,0.5,0.5] ) )

for k, v in diff_labels.items():
    if v > 0:
        print(v)
        fig.add_annotation(x=k, y=max(values)-v, ax=0,
                           yanchor='bottom',
                           ay=max(values), ayref='y',
                           showarrow=True, arrowsize=2,
                           arrowhead=1,
                           text = v
                          )
        
fig.add_traces(go.Scatter(x=labels, y=[max(values)]*3, mode = 'lines',
                          line=dict(color='black', width=1)))
        
fig.show()