Seaborn boxplot custom lables aside box

1
votes

I have the code segment given below, and it generates the provided boxplot. I would like to know how to add custom labels aside each box, so that the boxplot is even more digestible to the readers of my result. The expected diagram is also provided. I reckon there should be an easy way to get this done in Seaborn/Matplotlib.

What I exactly want is to add the following labels to each box (on left hand side as in shown in the example provided)

The code use to generate boxplot

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.ticker as MaxNLocator
from matplotlib import rcParams
from matplotlib.ticker import ScalarFormatter, FuncFormatter,FormatStrFormatter, EngFormatter#, mticker
%matplotlib inline
import seaborn as sns

range_stats = pd.read_csv(f'{snappy_data_dir}range_searcg_snappy_stats.csv')
data_stats_rs_txt = range_stats[range_stats['category'] == "t"]
data_stats_rs_seq = range_stats[range_stats['category'] == "s"]

fig, ax =plt.subplots(1,2)
rcParams['figure.figsize'] =8, 6
flierprops = dict(marker='x')
labels1 = ['R1', 'R2', 'R3', 'R4',  'R5']
sns.boxplot(x='Interval',y='Total',data=data_stats_rs_txt,palette='rainbow', ax=ax[0])

sns.boxplot(x='Interval',y='Total',data=data_stats_rs_seq,palette='rainbow', ax=ax[1])
ax[0].set(xlabel='Interval (s)', ylabel='query execution time (s)', title='Text format', ylim=(0, 290))
ax[1].set(xlabel='Interval (s)', ylabel='',  title='Proposed format',ylim=(0, 290), yticklabels=[])
plt.savefig("range-query-corrected.svg")
plt.savefig('snappy_compressed_rangesearch.pdf')

Resulted figure:

enter image description here

Expected figure with labels

enter image description here

2
pandasmatplotlibseabornboxplot

2 Answers

2
votes

This might help you, although it is not a fully correct way and is not a complete solution.

import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline

tips = sns.load_dataset('tips')

fig, axes = plt.subplots(1, 2, figsize=(12, 4))
sns.set_context('poster',font_scale=0.5)

sns.boxplot(x="day", y="total_bill", data=tips,palette='rainbow', ax=axes[0], zorder=0)

axes[0].text(0, 45, r"$B1$", fontsize=20, color="blue")
axes[0].text(0.9, 45, r"$B2$", fontsize=20, color="blue")
axes[0].text(2.2, 45, r"$B3$", fontsize=20, color="blue")
axes[0].text(3.1, 45, r"$B4$", fontsize=20, color="blue");

sns.boxplot(x="day", y="tip", data=tips,palette='rainbow', ax=axes[1], zorder=10)

Sample plot

1
votes
iris = sns.load_dataset("iris")
x_var = 'species'
y_var = 'sepal_width'
x_order = ['setosa', 'versicolor', 'virginica']
labels = ['R1','R2','R3']
max_vals = iris.groupby(x_var).max()[y_var].reindex(x_order)
ax = sns.boxplot(x=x_var, y=y_var, data=iris)

for x,y,l in zip(range(len(x_order)), max_vals, labels):
    ax.annotate(l, xy=[x,y], xytext=[0,5], textcoords='offset pixels', ha='center', va='bottom')

enter image description here