Plotly Express Boxplot boxmode group not working with x-axis as log

📊 Plotly Python
1

I’m trying to produce a plotly express boxplot using boxmod=‘group’ so I don’t have overlay of boxes and noticed that this works with a linear x-axis, but not when I change the xtype to logarithmic. The boxplot is created as follows:

boxfig = px.box(df, x=xaxis_column_name, y=yaxis_column_name, color=hue, log_x=True, boxmode='group')
boxfig.update_xaxes(title=xname, type=xtype)
boxfig.update_yaxes(title=yname, type=ytype)

Where xtype and ytype are either ‘linear’ or ‘log’.

X-axis as linear:

fig.update_xaxes(title=xname, type='linear')

image
image838×637 7.68 KB

X-axis as logarithmic:

fig.update_xaxes(title=xname, type='log')

image
image835×649 8.5 KB

What I’d like to see is the side by side boxes (as in the ‘linear’ example) but with a ‘log’ x-axis scale. Is this a bug? Am I going about this wrong?

2

Just by looking at these images: try changing the range of the xaxis for the second plot.

3

I’m not sure what you mean. Maybe showing the full plot would be helpful (see below). I didn’t originally because it’s hard to see that one is overlapping and the other is not when fully zoomed out.

Here’s maybe a better example.

Linear x-axis:
image

Logarithmic x-axis:
image

4

Now I am not sure what you mean. Could you share your code so that I can reproduce your figures?