I am plotting multiple boxplots along two different axes.
My code looks like:
fig, (ax1, ax2) = plt.subplots(2, sharex=True, sharey=False)
data_1 = [array1, array2, array3]
ax1.boxplot(data_1, whis=[5,95], showfliers=True)
data_2 = [array4, array5]
ax2.boxplot(data_2, whis=[5,95], showfliers=True)
ax2.set_xlim(0,4)
This produces a plot (substituting in my actual data) that looks like:
However, I would like the lower plot (on ax2) to shift to the right along the x-axis by one unit. That is, I would like to have the 2 lower boxplots plot at x=2 and x=3, such that they line up with the 2nd and 3rd upper boxplots. I would like to keep the xlabels the same and consistent for all x-axes.
Any ideas?
This should work for your example code. However this solution bypasses the sharex
aligment
In my opinion, the axis labeling when using box plots and sharex
is a little unintuitive.
%matplotlib inline
import matplotlib.pylab as plt
import numpy as np
np.random.seed(42)
# create random data
for i in range(1,6):
x = np.random.rand(10)
exec("array%s = x" % i)
widths = 0.3
fig, (ax1, ax2) = plt.subplots(2, sharex=True, sharey=False)
data_1 = [array1, array2, array3]
ax1.boxplot(data_1, widths=0.3, whis=[5,95], showfliers=True)
data_2 = [array4, array5]
positions = [2,3]
ax2.boxplot(data_2, positions=positions, widths=widths, whis=[5,95], showfliers=True)
ax2.set_xticks([1,2,3])
ax1.set_xticks([1,2,3])
ax2.set_xticklabels([1,2,3])
plt.xlim(0,4)