Visualizing statistical data is an essential step in data analysis. You’re probably familiar with using visualizations to transform raw numbers into clear and meaningful insights.
Effective visualizations allow us to identify patterns and trends. Which makes complex data more accessible and easier to interpret. This tutorial goes over five actionable tips for visualizing statistical data using code examples as needed.
Whether you’re looking to improve your data analysis skills or make your presentations more impactful, these tips will guide you in creating clear and insightful visualizations. Let’s get started.
Note: Here’s the Google Colab notebook for this tutorial.
1. Choose the Right Chart for Your Data
Choosing the correct chart type is the first step in effective data visualization. Selecting the appropriate chart type ensures that other developers and target audience can quickly get the key points you want to get across.
It’s essential to match your chart type to the features in your dataset to avoid misleading interpretations. You can use:
- Bar charts for comparing different categories or groups
- Histograms for displaying the distribution of a continuous variable
- Scatter plots for understanding correlations between two continuous variables
Let’s look at how to create different types of charts using seaborn for data visualization on a sample dataset.
Bar Plot for Categorical Data
As mentioned, a bar plot is used to compare different categories or groups. Here, we’ll visualize the average total bill amount by day of the week using the tips dataset.
import seaborn as sns
import matplotlib.pyplot as plt
# Load the dataset
data = sns.load_dataset("tips")
# Bar plot
sns.barplot(x="day", y="total_bill", data=data)
plt.title("Average Total Bill by Day")
plt.show()
Histogram for Continuous Data
A histogram helps in understanding the distribution of a continuous variable. Let’s visualize the distribution of total bill amounts.
# Histogram for distribution
sns.histplot(data["total_bill"], kde=True)
plt.title("Distribution of Total Bill Amounts")
plt.xlabel("Total Bill")
plt.ylabel("Frequency")
plt.show()
Scatter Plot for Correlation
Let’s create a scatter plot to visualize the relation between the total bill and the tip amounts.
# Scatter plot for correlation
sns.scatterplot(x="total_bill", y="tip", data=data)
plt.title("Total Bill vs Tip Amount")
plt.xlabel("Total Bill")
plt.ylabel("Tip")
plt.show()
2. Use Color to Enhance Data Comprehension
Color is a powerful tool in data visualization that improves the comprehension of your data. By strategically using color, you can highlight key trends, differentiate between categories, and draw attention to specific data points.
However, it’s important to use colors with caution. You should be careful not to overdo your charts with too many colors which can be overwhelming and may obscure the insights you’re trying to communicate.
Let’s explore how to use color palettes in Seaborn to differentiate categories and emphasize certain data points.
Using Color to Differentiate Categories
Let’s create a scatter plot—total bill vs. tip amount—using different colors to represent different days of the week. This makes it easier to compare how tips vary across different days.
import seaborn as sns
import matplotlib.pyplot as plt
# Load the dataset
data = sns.load_dataset("tips")
# Scatter plot with color differentiation by day
sns.scatterplot(x="total_bill", y="tip", hue="day", data=data, palette="Set2")
plt.title("Total Bill vs Tip Amount by Day")
plt.xlabel("Total Bill")
plt.ylabel("Tip")
plt.show()
In this example, the hue parameter is used to assign different colors to each day, making it easy to see how the relationship between total bill and tip changes across the week.
Using Custom Color Palettes to Emphasize Data
You can also use custom color palettes to emphasize certain data points. Below, we create a histogram of total bill amounts, using a custom color palette.
# Custom color palette for a histogram
custom_palette = sns.color_palette("coolwarm", as_cmap=True)
sns.histplot(data["total_bill"], kde=True, color=custom_palette(0.7))
plt.title("Distribution of Total Bill Amounts")
plt.xlabel("Total Bill")
plt.ylabel("Frequency")
plt.show()
Here, we use the coolwarm palette and use the custom palette when plotting the histogram.
3. Visualize Statistical Summaries
Summaries such as the mean, median, and quartiles help understand the central tendency, spread, and variability of your data. But sometimes, visualizing statistical summaries helps better understand the distribution of your data.
And for this, you can use box plots, violin plots, and error bars. Let’s create some of these plots for the tips dataset.
Using Box Plots
With a box plot, you can get a graphical summary of the distribution of a dataset, indicating the median, quartiles, and outliers. As an example, let’s visualize the distribution of total bills across different days of the week.
import seaborn as sns
import matplotlib.pyplot as plt
# Load the dataset
data = sns.load_dataset("tips")
# Box plot for statistical summaries
sns.boxplot(x="day", y="total_bill", data=data)
plt.title("Total Bill Distribution by Day")
plt.xlabel("Day")
plt.ylabel("Total Bill")
plt.show()
The box plot shows the median (the line inside the box), the interquartile range (the box), and any outliers (points outside the whiskers). This provides a clear summary of the data distribution across different days.
Using Violin Plots
A violin plot combines aspects of a box plot with a density plot, giving a richer view of the data distribution. Here’s how to visualize the total bill distribution across days with a violin plot.
# Violin plot for distribution and summaries
sns.violinplot(x="day", y="total_bill", data=data)
plt.title("Total Bill Distribution by Day (Violin Plot)")
plt.xlabel("Day")
plt.ylabel("Total Bill")
plt.show()
The violin plot provides a more detailed look at the distribution, showing the density of the data at different values, while also including the same summary statistics as the box plot.
4. Add Context with Annotations
Annotations are an effective way to add context to your visualizations—helping to highlight key data points or explain specific trends. You can use annotations to bring attention to specific information that might not be immediately obvious from the chart alone.
Typically, annotations can be used to label outliers, mark significant events, or clarify unusual patterns, making your data story clearer and more engaging.
Highlighting Multiple Data Points
Annotations can also be used to highlight multiple points in a plot, providing a better understanding of the data. For example, we can annotate the highest and lowest tips in our dataset using Matplotlib’s annotate() function.
Here, we highlight both the maximum and minimum tip amounts, using arrows for emphasis.
# Annotating the highest and lowest tips
max_tip = data.loc[data["tip"].idxmax()]
min_tip = data.loc[data["tip"].idxmin()]
plt.figure(figsize=(8, 6))
sns.scatterplot(x="total_bill", y="tip", data=data)
# Annotate max tip
plt.annotate(f"Max Tip: {max_tip['tip']}",
xy=(max_tip["total_bill"], max_tip["tip"]),
xytext=(max_tip["total_bill"] + 5, max_tip["tip"] + 2),
arrowprops=dict(facecolor='green', arrowstyle="->"))
# Annotate min tip
plt.annotate(f"Min Tip: {min_tip['tip']}",
xy=(min_tip["total_bill"], min_tip["tip"]),
xytext=(min_tip["total_bill"] - 10, min_tip["tip"] - 2),
arrowprops=dict(facecolor='red', arrowstyle="->"))
plt.title("Total Bill vs Tip Amount with Max/Min Annotations")
plt.xlabel("Total Bill")
plt.ylabel("Tip")
plt.show()
Whether you’re highlighting outliers, marking significant events, or simply adding explanatory notes, annotations are a simple yet effective way to make your data visualizations better.
5. Combine Multiple Plots for Comprehensive Analysis
Combining multiple plots allows you to compare different aspects of the data in a single, cohesive visual representation. This approach is particularly useful when you want to analyze multiple variables simultaneously or provide a more comprehensive view of the data.
You can use subplots or facet grids to display related charts side by side, making it easier to identify patterns that aren’t as obvious when viewing the plots individually.
Creating Subplots
Here, we’ll create a grid of subplots to compare the distributions of total bills and tips across different days of the week.
import seaborn as sns
import matplotlib.pyplot as plt
# Load the dataset
data = sns.load_dataset("tips")
# Create a figure with subplots
fig, axs = plt.subplots(2, 2, figsize=(12, 10))
# Subplot 1: Distribution of total bills
sns.histplot(data["total_bill"], kde=True, ax=axs[0, 0])
axs[0, 0].set_title("Distribution of Total Bill")
# Subplot 2: Distribution of tips
sns.histplot(data["tip"], kde=True, ax=axs[0, 1])
axs[0, 1].set_title("Distribution of Tips")
# Subplot 3: Total bill by day
sns.boxplot(x="day", y="total_bill", data=data, ax=axs[1, 0])
axs[1, 0].set_title("Total Bill by Day")
# Subplot 4: Tip by day
sns.boxplot(x="day", y="tip", data=data, ax=axs[1, 1])
axs[1, 1].set_title("Tip by Day")
# Adjust layout
plt.tight_layout()
plt.show()
In this example, we’ve created a 2×2 grid of subplots, allowing us to compare the distributions of total bills and tips and how these variables vary across different days of the week.
Wrapping Up
As you continue to work with data, remember that the best visualizations are those that clearly convey the story your data is telling. Be sure to experiment with different visualization techniques and styles to find what works best for your specific dataset.
By applying the tips covered in this tutorial, you can create powerful and informative visualizations that not only look great but also provide useful insights. Happy data visualization!
