Matplotlib: histogram with multiple bars

Question

I cannot seem to grasp how to draw multiple bar histograms. This is my code:

import matplotlib.pyplot as mpl

tags = 'manual (4 serv.)', 'Cromlech (average, 4 serv.)', 'Cromlech (0.925, 4 serv.)', 'Cromlech (0.925 improved, 7 serv.)', 'Cromlech (15 serv.)', 'Pangaea', 'ServiceCutter (4 serv.)'
a = (0.385, 0.4128, 0.406, 0.5394, 0.7674, 0.306, 0.3505)
b = (0.4025, 0.1935, 0.189, 0.189, 0.415, 0.238, 0.1714)
c = (1, 0.3619, 0.5149, 1, 0.4851, 0.4092, 0.4407)
d = (1, 0.3495, 0.4888, 1, 0.4861, 0.4985, 0.5213)
mpl.hist((a, b, c, d), 7, label=("decoupling", "op. cost", "op. similarity", "data similarity"))
mpl.legend()
mpl.xticks((1, 2, 3, 4, 5, 6, 7), (tags))
mpl.show()

What I'm trying to do is produce a plot with 7 data points, each characterized by a quadruple (decoupling, op. cost, op. similarity, data similarity). "a, b, c, d" respectively contains the value for decoupling, op. costs... I want to label each of the data points with one of the tags in the code.

I don't understand what I'm doing wrong. Could you help me?

Answer 1

It looks like you want to create a bar plot, not a histogram.

In this case, the grouping, the labels and the legend are easiest if you create a pandas dataframe , and use pandas plotting (pandas uses matplotlib for plotting):

import matplotlib.pyplot as plt
import pandas as pd

tags = ('manual (4 serv.)', 'Cromlech (average, 4 serv.)', 'Cromlech (0.925, 4 serv.)',
        'Cromlech (0.925 improved, 7 serv.)', 'Cromlech (15 serv.)', 'Pangaea', 'ServiceCutter (4 serv.)')
# insert some newlines in the tags to better fit into the plot
tags = [tag.replace(' (', '\n(') for tag in tags]
a = (0.385, 0.4128, 0.406, 0.5394, 0.7674, 0.306, 0.3505)
b = (0.4025, 0.1935, 0.189, 0.189, 0.415, 0.238, 0.1714)
c = (1, 0.3619, 0.5149, 1, 0.4851, 0.4092, 0.4407)
d = (1, 0.3495, 0.4888, 1, 0.4861, 0.4985, 0.5213)
# create a dataframe
df = pd.DataFrame({"decoupling": a, "op. cost": b, "op. similarity": c, "data similarity": d}, index=tags)
df.plot.bar(rot=0, figsize=(12, 5))
plt.tight_layout() # fit labels etc. nicely into the plot
plt.show()

Optionally, you can modify visual aspects. Here is an example:

ax = df.plot.bar(rot=0, figsize=(12, 5), width=0.75, cmap='turbo')
for spine in ['top', 'right']:
    ax.spines[spine].set_visible(False)
ax.set_axisbelow(True)
ax.grid(axis='y', color='grey', ls=':')
ax.set_facecolor('beige')
ax.set_ylim(0, 1)
plt.tight_layout()