Pandas Styling

When working with data in pandas, it's often necessary to present your findings in a clear, visually appealing way. Pandas provides a rich set of styling tools that allow you to format your DataFrames for better visualization and communication of insights. This guide will walk you through the various styling capabilities in pandas, from basic techniques to advanced customization.

Introduction to Pandas Styling

The .style accessor in pandas allows you to format and style your DataFrames without changing the underlying data. This is particularly useful when you want to:

• Highlight specific values or patterns
• Format numbers appropriately (currencies, percentages, etc.)
• Create visually appealing reports
• Add color scales to represent data magnitude
• Generate publication-quality tables

Let's dive into how you can leverage these styling capabilities!

Basic Styling Techniques

Getting Started with Styling

To begin styling a DataFrame, you simply access the .style property of your DataFrame:

python

import pandas as pd
import numpy as np

# Create a sample DataFrame
df = pd.DataFrame({
    'A': np.linspace(1, 10, 5),
    'B': np.linspace(10, 20, 5),
    'C': ['a', 'b', 'c', 'd', 'e']
})

# Access the style property
styled_df = df.style
styled_df

Output:

Number Formatting

One of the most common styling needs is to format numbers with specific precision or as percentages, currencies, etc.

python

import pandas as pd
import numpy as np

# Create a sample DataFrame
df = pd.DataFrame({
    'Price': [1.2345, 2.3456, 3.4567],
    'Percentage': [0.1234, 0.2345, 0.3456],
    'Large Number': [1234567, 2345678, 3456789]
})

# Format with different styles
styled_df = df.style.format({
    'Price': '${:.2f}',
    'Percentage': '{:.2%}',
    'Large Number': '{:,.0f}'
})

styled_df

Output:

The output shows:

• 'Price' column formatted as currency with 2 decimal places
• 'Percentage' column formatted as percentage with 2 decimal places
• 'Large Number' column formatted with thousand separators

Conditional Formatting

Highlight Max and Min Values

You can easily highlight the maximum and minimum values in your DataFrame:

python

import pandas as pd
import numpy as np

# Create a sample DataFrame
df = pd.DataFrame(np.random.randn(5, 3), 
                  columns=['A', 'B', 'C'])

# Highlight the maximum values in green and minimum values in red
styled_df = df.style.highlight_max(color='lightgreen').highlight_min(color='pink')
styled_df

Output:

Custom Conditional Formatting

You can define custom highlighting conditions with the .applymap() and .apply() methods:

python

import pandas as pd
import numpy as np

df = pd.DataFrame(np.random.randn(5, 3), 
                  columns=['A', 'B', 'C'])

# Define a function for conditional styling
def color_negative_red(val):
    """
    Takes a scalar and returns a string with
    the CSS property `'color: red'` for negative
    values, black otherwise.
    """
    color = 'red' if val < 0 else 'black'
    return f'color: {color}'

# Apply the styling function to the entire DataFrame
styled_df = df.style.applymap(color_negative_red)
styled_df

Output:

Color Scales and Gradients

Background Color Scales

You can apply color scales to visualize the distribution of values:

python

import pandas as pd
import numpy as np

df = pd.DataFrame(np.random.randn(10, 3), 
                  columns=['A', 'B', 'C'])

# Apply a color background gradient
styled_df = df.style.background_gradient(cmap='viridis')
styled_df

Output:

Applying Color Scales by Columns

You can also apply color scales separately for each column:

python

import pandas as pd
import numpy as np

df = pd.DataFrame(np.random.randn(10, 3) * [10, 100, 1000], 
                  columns=['A', 'B', 'C'])

# Apply color background gradient by column
styled_df = df.style.background_gradient(cmap='coolwarm', axis=0)
styled_df

Output:

Bar Charts and Visual Representations

Adding Bar Charts

Pandas styling allows you to add in-cell bar charts for numerical comparisons:

python

import pandas as pd
import numpy as np

df = pd.DataFrame({
    'A': np.random.randint(0, 100, 5),
    'B': np.random.randint(0, 100, 5),
    'C': np.random.randint(0, 100, 5)
})

# Add bar charts to visualize data
styled_df = df.style.bar(subset=['A', 'B', 'C'], color='#d65f5f')
styled_df

Output:

Table Styling for Presentation

Table Properties

You can modify the appearance of the entire table:

python

import pandas as pd
import numpy as np

df = pd.DataFrame(np.random.randn(5, 3), 
                  columns=['A', 'B', 'C'])

# Style the entire table
styled_df = df.style.set_properties(**{
    'background-color': '#f5f5f5',
    'border': '1px solid black',
    'text-align': 'center',
    'font-weight': 'bold'
})
styled_df

Output:

Adding Captions and Headers

Add captions and style your table headers:

python

import pandas as pd
import numpy as np

df = pd.DataFrame(np.random.randn(5, 3), 
                  columns=['A', 'B', 'C'])

# Add caption and style headers
styled_df = df.style.set_caption("Sample Data Analysis")\
    .set_table_styles([
        {'selector': 'th', 
         'props': [('background-color', '#007bff'), 
                   ('color', 'white'),
                   ('font-weight', 'bold')]},
        {'selector': 'caption',
         'props': [('caption-side', 'top'),
                   ('font-size', '18px'),
                   ('font-weight', 'bold')]}
    ])
styled_df

Output:

Combining Multiple Styling Features

You can chain multiple styling methods to create comprehensive visualizations:

python

import pandas as pd
import numpy as np

np.random.seed(24)
df = pd.DataFrame({
    'A': np.linspace(1, 10, 10),
    'B': np.linspace(11, 20, 10) + np.random.randn(10),
    'C': np.random.uniform(20, 30, 10)
})

# Apply multiple styles
styled_df = df.style\
    .format('{:.2f}')\
    .background_gradient(subset=['A'], cmap='Reds')\
    .background_gradient(subset=['B'], cmap='Blues')\
    .background_gradient(subset=['C'], cmap='Greens')\
    .highlight_max(color='yellow', axis=0)\
    .set_caption('Combined Styling Example')\
    .set_table_styles([
        {'selector': 'th',
         'props': [('background-color', '#4CAF50'),
                   ('color', 'white'),
                   ('font-weight', 'bold')]}
    ])

styled_df

Output:

Real-world Application: Financial Dashboard

Let's create a more comprehensive example showing how styling can be used in a financial context:

python

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt

# Create sample financial data
np.random.seed(42)
index = pd.date_range('2023-01-01', periods=10, freq='B')
data = {
    'Open': np.random.uniform(100, 150, 10),
    'High': np.random.uniform(150, 200, 10),
    'Low': np.random.uniform(50, 100, 10),
    'Close': np.random.uniform(100, 150, 10),
    'Volume': np.random.randint(1000000, 5000000, 10),
    'Change %': np.random.uniform(-0.05, 0.05, 10)
}

df = pd.DataFrame(data, index=index)

# Fix the data to ensure High >= Open >= Low and High >= Close >= Low
for i in range(len(df)):
    df.loc[df.index[i], 'High'] = max(df.loc[df.index[i], ['Open', 'Close', 'High']].max(), 
                                     df.loc[df.index[i], 'High'])
    df.loc[df.index[i], 'Low'] = min(df.loc[df.index[i], ['Open', 'Close', 'Low']].min(), 
                                    df.loc[df.index[i], 'Low'])

# Create a styled financial table
def color_change(val):
    color = 'green' if val > 0 else 'red'
    return f'color: {color}; font-weight: bold'

styled_df = df.style\
    .format({
        'Open': '${:.2f}',
        'High': '${:.2f}',
        'Low': '${:.2f}',
        'Close': '${:.2f}',
        'Volume': '{:,.0f}',
        'Change %': '{:.2%}'
    })\
    .applymap(color_change, subset=['Change %'])\
    .background_gradient(subset=['Volume'], cmap='Blues')\
    .bar(subset=['High', 'Low'], color=['lightgreen', 'pink'])\
    .set_caption('Stock Price Analysis - January 2023')\
    .set_table_styles([
        {'selector': 'th', 
         'props': [('background-color', '#2c3e50'), 
                   ('color', 'white'),
                   ('font-weight', 'bold'),
                   ('text-align', 'center')]},
        {'selector': 'caption',
         'props': [('caption-side', 'top'),
                   ('font-size', '16px'),
                   ('font-weight', 'bold'),
                   ('color', '#2c3e50')]}
    ])

styled_df

Output:

Exporting Styled DataFrames

Converting to HTML

You can export your styled DataFrame to HTML for web applications:

python

import pandas as pd
import numpy as np

df = pd.DataFrame(np.random.randn(5, 3), 
                  columns=['A', 'B', 'C'])

# Style the DataFrame
styled_df = df.style.format('{:.2f}')\
    .background_gradient(cmap='viridis')

# Export to HTML (can be saved to a file)
html = styled_df.to_html()
print(html[:300] + "...")  # Showing first 300 characters

Output:

<style type="text/css">
#T_12345_row0_col0, #T_12345_row0_col1, #T_12345_row0_col2, #T_12345_row1_col0, #T_12345_row1_col1, #T_12345_row1_col2, #T_12345_row2_col0, #T_12345_row2_col1, #T_12345_row2_col2, #T_12345_row3_col0, #T_12345_row3_col1, #T_12345_row3_col2, #T_12345_row4_col0, #T_12345_row4_col1, #T_12345_row4_col2 {
  background-color: ...

Summary

Pandas styling functionality offers a powerful way to enhance your DataFrames for better visualization and communication. In this guide, we covered:

• Basic number formatting and styling
• Conditional formatting and highlighting
• Color scales and gradients
• In-cell bar charts
• Table styling for presentations
• Combining multiple styling features
• Real-world applications
• Exporting styled DataFrames

With these tools, you can transform plain DataFrames into informative, visually appealing presentations that effectively communicate your data insights.

Additional Resources and Exercises

Additional Resources

• Official Pandas Styling Documentation
• Pandas Styling Cookbook
• Python for Data Analysis (Book by Wes McKinney)

Exercises

1. Basic Styling Exercise: Create a DataFrame with student grades and format it to show percentages. Highlight scores above 90% in green and below 60% in red.

2. Financial Data Exercise: Generate a mock stock portfolio and create a styled table showing gains and losses with appropriate color coding.

3. Sales Dashboard Challenge: Create a DataFrame with monthly sales data and style it to include:

   • Currency formatting
   • Color gradient for sales values
   • Bar charts comparing monthly performance
   • Highlighted top and bottom months
   • Custom header styling

4. Custom Style Function: Create a custom styling function that applies a 'traffic light' color scheme (red, yellow, green) based on threshold values.

5. Export Challenge: Style a DataFrame with various formatting features and export it to HTML for use in a web page.

By mastering pandas styling, you'll be able to create more effective data presentations that highlight key insights and make your analysis more impactful.