Data Visualization in Jupyter

# CHAPTER 12

Data Visualization in Jupyter #

1. Chapter Introduction #

2. Matplotlib: The Foundation #

Matplotlib is the original Python visualization library. It is incredibly powerful and customizable, though the syntax can be a bit verbose.

Cell 1:

import matplotlib.pyplot as plt
import numpy as np

# Data
months = ['Jan', 'Feb', 'Mar', 'Apr', 'May']
sales = [150, 200, 180, 250, 300]

# 1. Create the plot
plt.plot(months, sales)

# 2. Customize the plot
plt.title("Monthly Sales Growth")
plt.xlabel("Month")
plt.ylabel("Sales in USD")

# 3. Show the plot inline
plt.show()

*When you run this cell, a line chart will appear instantly below it.*

3. Creating Different Chart Types #

Matplotlib can create almost any standard chart type.

Cell 2:

# Bar Chart
plt.bar(months, sales, color='skyblue')
plt.title("Bar Chart Example")
plt.show()

# Scatter Plot (great for finding correlations)
ad_spend = [50, 70, 60, 90, 110]
plt.scatter(ad_spend, sales, color='red', marker='x')
plt.title("Ad Spend vs Sales")
plt.xlabel("Ad Spend ($)")
plt.ylabel("Sales ($)")
plt.show()

4. Seaborn: Beautiful Defaults #

While Matplotlib is great, it can look a bit outdated by default. Seaborn is a library built *on top* of Matplotlib. It creates gorgeous, modern, statistical graphics with much less code. It also integrates perfectly with Pandas DataFrames.

Cell 3:

import seaborn as sns
import pandas as pd

# Create a sample DataFrame
data = pd.DataFrame({
    'Category': ['A', 'A', 'B', 'B', 'C', 'C'],
    'Values': [10, 15, 20, 25, 30, 35]
})

# Seaborn automatically makes it look modern
# We pass the DataFrame directly to the 'data' parameter
sns.boxplot(x='Category', y='Values', data=data, palette='Set2')
plt.title("Seaborn Boxplot")
plt.show()

5. Jupyter Magic: %matplotlib inline #

In older versions of Jupyter, if you created a chart, it might open in a new pop-up window rather than inside the notebook.

To force charts to appear directly in the notebook (inline), data scientists traditionally placed a Magic Command at the top of their notebook: %matplotlib inline

*Note: Modern Jupyter Notebooks and JupyterLab usually do this automatically, but you will see this line in millions of older notebooks on GitHub. It's important to know what it does.*

6. Saving Visualizations #

Often, you need to put the chart you generated into a PowerPoint presentation or a Word document. You can save charts directly to your hard drive using code.

Cell 4:

plt.plot(months, sales)
plt.title("Final Report Chart")

# Save the figure BEFORE calling plt.show()
# It saves to the same folder as your notebook
plt.savefig('sales_report_chart.png', dpi=300, bbox_inches='tight')

plt.show()

*(Check your Jupyter file browser, you will now see salesreportchart.png!)*

7. Mini Project: Sales Analytics Dashboard #

Combine multiple plots into a single visual dashboard using Matplotlib's subplots feature.

Cell 5:

# Create a figure with 1 row and 2 columns
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 4))

# First subplot (Line chart)
ax1.plot(months, sales, marker='o', color='blue')
ax1.set_title("Revenue Trend")
ax1.set_ylabel("$ Revenue")

# Second subplot (Bar chart)
ax2.bar(months, ad_spend, color='orange')
ax2.set_title("Marketing Spend")
ax2.set_ylabel("$ Spend")

# Adjust spacing and display
plt.tight_layout()
plt.show()

8. Common Mistakes #

• Calling plt.savefig() after plt.show(): When plt.show() runs, it displays the image and then clears the canvas. If you put plt.savefig() after it, it will save a blank white square. Always save *before* you show.

• Not setting labels: A chart without a title, X-axis label, and Y-axis label is useless to a stakeholder. Always label your axes.

9. MCQs #

10. Interview Questions #

• Q: Explain the relationship between Matplotlib and Seaborn. Why would you use one over the other?

• Q: If you want to place two charts side-by-side in a single image in Jupyter, how do you approach it architecturally in Matplotlib?

11. Summary #

12. Next Chapter Recommendation #