Seaborn: How to replace index with text in X-Axis in barplot?

Question

I have a dataset with category column which has integer values representing class label i.e 0,1,2.....

I have separate file which contains text labels for that category i.e against index 0, it contains classA and so on. I want to plot a barplot using seaborn with following code.

import seaborn as sns
train_df = pd.read_csv("unclean_text.csv", sep='\t')
label_text = pd.read_csv("labels.csv")
is_dup = train_df['category'].value_counts()

plt.figure(figsize=(8,4))
sns.barplot(is_dup.index, is_dup.values, alpha=0.8, color=color[1])
plt.show()

It correctly plots the barplot for frequency of each class.

But I want text labels on the x-axis instead of index values that are in label_text, which is also a column vector of length 19 (0-18). How to do it?

Answer 1

sns.barplot() will return the axis for the plot. You can use this to set your tick labels:

import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns

color = ['r', 'g', 'b']
train_df = pd.read_csv("unclean_text.csv", sep='\t')
label_text = pd.read_csv("labels.csv")
is_dup = train_df['category'].value_counts()

plt.figure(figsize=(8,4))
ax = sns.barplot(is_dup.index, is_dup.values, alpha=0.8, color=color[1])
ax.set_xlabel('Category')
ax.set_ylabel('Number of Occurrences')
ax.set_xticklabels(label_text['labels'], rotation='vertical', fontsize=10)
plt.show()

This assumes labels.csv is something like:

labels
cat0
cat1
cat2
cat3
cat4
cat5
cat6
cat7
cat8

etc..

Giving you an output: