Custom error bars in grouped bar chart

Iā€™d like to add error bars in grouped bar chart (actually a histogram), something like this:


The source data are melted, so I cannot add individual traces and specify their errors.
Is there a way how to achieve that?
My data look something like this:

pd.DataFrame({"name":["foo", "bar"]*2, "x":["A", "A", "B", "B"], "y":[5, 6, 7, 8]})

Thanks!

So I found the solution.

  1. Use bar plot instead of histogram (histogram does not have error_y argument).
  2. Simply set error_y argument to an array-like of error values, e. g.
figure = px.bar(
    dataframe,
    x="name",
    y="y",
    color='x',
    barmode='group',
    error_y=[6, 7, 5, 6, 7, 7]

Works well for melted data too, you just need to pass list with equal number of elements as there are bars in your graph. Easy :slight_smile: