Two side-by-side Bar Charts considering the linewidth

Question

I'm having an issue with creating two side-by-side bar charts using Matplotlib. Everything works fine if there is no linewidth, but as soon as I add one, the bars overlap.

Here's an example:

import matplotlib.pyplot as plt

fig, ax = plt.subplots(1, figsize=(3,2), dpi=300)
shift=0.2
ax.bar([0-shift,1-shift,2-shift], [2,2,2], 2*shift, lw=2, color='blue',ec='darkblue')
ax.bar([0+shift,1+shift,2+shift], [2,2,2], 2*shift, lw=2, color='red', ec='darkred')

I tried using axes.transform() to convert the linewidth into data and manually adjust my shift, but it didn't work.

I would like to achieve something like:

Answer 1

One approach is to use the double of the line width, and clip each bar with itself. Here is an example, that would also work for bars created via pandas or seaborn. The colors are a chosen a bit lighter to better see what's going on.

import matplotlib as mpl
import matplotlib.pyplot as plt

fig, ax = plt.subplots(1, figsize=(3, 2), dpi=300)
shift = 0.2
ax.bar([0 - shift, 1 - shift, 2 - shift], [2, 2, 2], 2 * shift, lw=4, color='lightblue', ec='blue')
ax.bar([0 + shift, 1 + shift, 2 + shift], [2, 2, 2], 2 * shift, lw=4, color='pink', ec='red')

for bars in ax.containers:
    if type(bars) == mpl.container.BarContainer:
        for bar in bars:
            bar.set_clip_path(bar)

plt.show()