Charts and Insights

Introduction

Numbers in a table are hard to scan. A chart shows the pattern at a glance. pandas works with a plotting library called matplotlib (install it with pip install matplotlib).

A bar chart in three lines

import matplotlib.pyplot as plt

totals = df.groupby("category")["revenue"].sum().sort_values(ascending=False)
totals.plot(kind="bar", title="Revenue by category")
plt.ylabel("Revenue")
plt.tight_layout()
plt.savefig("revenue.png")   # or plt.show() in a notebook

This draws one bar per category, tallest first. The bars show that Bags earns the most, then Stationery, then Kitchen.

Choosing a chart

• Bar chart: compare amounts across categories.
• Line chart: show change over time.
• Histogram: show how values are spread out, such as all prices.
• Scatter plot: check whether two number columns move together.

Avoid pie charts with many slices, because they are hard to compare.

Make charts honest

• Give every chart a title and label the axes, with units.
• Start bar charts at zero. Cutting the axis makes small differences look huge.
• Do not show more detail than the data supports.

From numbers to insight

An insight is a short, useful statement backed by data. Compare:

• Weak: "Here is a chart of revenue."
• Better: "Bags brought in 149.50 of the 304.50 total revenue in this small sample, nearly half."

Always add the limits. Our table has only six rows, so it is a practice sample and not proof of anything about a real shop. Say how many rows you used, and what you would check next.

Try it yourself

Load or build a small table of your own. Group it by one column, draw a bar chart, and write two sentences: what the chart shows and one limitation.

Summary

• Chart a grouped result with .plot(kind="bar") and save or show it.
• Match the chart type to the question, and label everything clearly.
• Finish with an honest insight that states its limits.