CODEWITHSUNDEEP
pythonmatplotlibseabornviolin-plot
Bilgin
Bilgin

Reputation: 519

How to plot Violin plot for multiple dataset?

I have multiple datasets with different size and I want to plot a violin plot from them. My dataset looks like below:

Input.CSV:

         city_A city_B  city C  city_D
cluster1   2       5      4      4
cluster2   3       3      2      8
cluster3   2       4      5      5
cluster4   3       5      4
cluster5           3      3
cluster6           5

Note: Each city has a different size and number of clusters.

I looked into a few posts such as here and I could not understand how to plot this dataset in one plot like:

enter image description here

Some of the example from seaborn or matplotlib is with fake data and my data is in CSV format as I showed above. It would be great if you can provide your help with code that use data like mine.

Upvotes: 1

Views: 7080

Answers (1)

Kartikeya Sharma
Kartikeya Sharma

Reputation: 1383

If you have multiple lists you want to plot, you can put them as list of lists and plot them. You can read the documentation here https://matplotlib.org/3.1.1/api/_as_gen/matplotlib.pyplot.violinplot.html

from matplotlib import pyplot as plt

A = [2, 5, 6, 10, 12, 8, 5]
B = [2, 2, 6,  8, 14, 5, 5]
C = [5, 7, 5, 13, 17, 7, 5]
D = [1, 4, 7, 12, 12, 5, 5]
E = [4, 1, 2, 11, 13, 7, 5]

fig, ax = plt.subplots(figsize=(5,5))
ax.violinplot([A,B,C,D,E][::-1],positions =[5,4,3,2,1],vert=False,showmeans=True)

def set_axis_style(ax, labels):
    ax.get_yaxis().set_tick_params(direction='out')
    ax.xaxis.set_ticks_position('bottom')
    ax.set_yticks(np.arange(1, len(labels) + 1))
    ax.set_yticklabels(labels)
    ax.set_ylim(0.25, len(labels) + 0.75)
    ax.set_ylabel('Sample name')

set_axis_style(ax,['A','B','C','D','E'][::-1])

enter image description here

Seaborn looks like a better and more aesthetic solution for the dataframe.

from matplotlib import pyplot as plt
import seaborn as sns

fig, axes = plt.subplots(figsize=(5,5))
sns.set(style="whitegrid")
sns.violinplot(data=df, ax = axes, orient ='h')

enter image description here

Upvotes: 2

Related Questions