# 5 Best Ways to Use Seaborn Library to Display Categorical Scatter Plots in Python

Rate this post

π‘ Problem Formulation: When working with categorical data in Python, visualizing relationships between variables becomes important for data analysis. Displaying categorical scatter plots is a frequent need to distinguish data points in different categories. We seek to utilize Python’s Seaborn library to generate scatter plots that effectively communicate the data’s structure, with varying categories clearly distinguished.

## Method 1: stripplot()

The `stripplot()` function in Seaborn creates a scatter plot where one variable is categorical. It is useful for plotting a distribution of values within categories as individual points, which can be jittered to reduce overlap. The function provides an excellent way to visualize the spread and density of data points within each category.

Here’s an example:

```import seaborn as sns
import matplotlib.pyplot as plt

# Sample data

# Create the plot
sns.stripplot(x="day", y="total_bill", data=tips, jitter=True)

plt.show()```

The above code snippet outputs a strip plot showing the distribution of total bills for different days of the week. It clearly demonstrates categorical distributions and can reveal dense clusters or gaps in the data.

## Method 2: swarmplot()

Seaborn’s `swarmplot()` function positions each scatter plot point on the categorical axis with an algorithm that avoids overlap, creating a clearer representation of the distribution of values. It is particularly useful for displaying all individual observations without cluttering, providing an unobscured view of the categorical data.

Here’s an example:

```import seaborn as sns
import matplotlib.pyplot as plt

# Sample data

# Create the plot
sns.swarmplot(x="day", y="total_bill", data=tips)

plt.show()```

This code produces a swarm plot where each data point’s position is adjusted along the categorical axis so they don’t overlap. The result is a more readable plot compared to simple strip plots, especially when dealing with numerous data points.

## Method 3: catplot()

Seabornβs `catplot()` is a versatile function capable of creating numerous kinds of categorical plots, including scatter plots, by specifying the kind parameter. This high-level interface provides the flexibility to create complex multivariate plots with ease.

Here’s an example:

```import seaborn as sns
import matplotlib.pyplot as plt

# Sample data

# Create the plot
sns.catplot(x="day", y="total_bill", kind="strip", data=tips)

plt.show()```

The `catplot()` function in this snippet is used to create a strip plot which plots the total bill against days of the week. Through catplot’s “kind” parameter, various other plots like box plots, violin plots, etc., can also be generated.

## Method 4: pointplot()

A `pointplot()` shows point estimates and confidence intervals using scatter plot glyphs. Itβs a great way to represent the central tendency of a numeric variable with scatter points, particularly useful for highlighting differences between levels.

Here’s an example:

```import seaborn as sns
import matplotlib.pyplot as plt

# Sample data

# Create the plot
sns.pointplot(x="day", y="total_bill", data=tips)

plt.show()```

The output from this code will display points representing the mean total bill for each day, connected by a line. This provides an effective comparison of the central tendency across categorical variables.

## Bonus One-Liner Method 5: pairplot()

Seaborn’s `pairplot()` function is a one-liner that can be used to quickly plot multiple pairwise scatter plots for categorical data, especially useful when dealing with pair-wise comparisons across several categories.

Here’s an example:

```import seaborn as sns

# Sample data

# Create the plot
sns.pairplot(iris, hue='species')

plt.show()```

The resulting visualization is a matrix of scatter plots for each pair of numerical variables in the data colored by species – an excellent tool for exploratory data analysis.

## Summary/Discussion

• Method 1: stripplot. Great for displaying individual datapoints. A downside is potential overlap without jitter.
• Method 2: swarmplot. Offers a clear representation without overlapping points but can be computationally intensive with large datasets.
• Method 3: catplot. Highly flexible with the ability to create various types of categorical plots, though it could be overkill for simple scatter plot needs.
• Method 4: pointplot. Excellent for comparing central tendencies across categories. Can obscure data distribution since it only shows aggregate statistics.
• Bonus Method 5: pairplot. Ideal for a quick overview of pairwise relationships, but may be too cluttered with numerous variables.