Matplotlib Mini Project (EDA charts pack)

Goal

Create a reusable set of plots you can apply to any dataset:

• Histogram (distribution)
• Bar chart (category counts)
• Line plot (trend)
• Scatter plot (relationship)

This mirrors how McKinney approaches exploratory data analysis (EDA) in the book: before modelling anything, look at the shape of your data (histogram), how categories compare (bar chart), how things move over time (line plot), and whether two variables relate (scatter plot). Four charts, four different questions.

diagram An EDA charts pack mermaid

  flowchart TD
  A["New dataset"] --> B["Distribution? -> Histogram"]
  A --> C["Compare categories? -> Bar chart"]
  A --> D["Change over time? -> Line plot"]
  A --> E["Relationship between 2 vars? -> Scatter plot"]

Four standard questions about a dataset, each answered by one chart type.

Sample dataset

Sample data

import pandas as pd
 
# Small example you can replace with any CSV
df = pd.DataFrame({
    "date": pd.date_range("2025-01-01", periods=10, freq="D"),
    "orders": [120, 140, 130, 160, 155, 170, 180, 175, 190, 200],
    "city": ["Pune", "Delhi", "Pune", "Delhi", "Pune", "Mumbai", "Pune", "Delhi", "Mumbai", "Pune"],
    "amount": [100, 140, 110, 160, 130, 90, 180, 150, 120, 200],
})

1) Histogram

Histogram

import matplotlib.pyplot as plt
 
plt.figure(figsize=(7, 4))
plt.hist(df["amount"], bins=8, edgecolor="black")
plt.title("Amount distribution")
plt.xlabel("Amount")
plt.ylabel("Count")
plt.tight_layout()
plt.show()

2) Bar chart (counts)

Bar counts

import matplotlib.pyplot as plt
 
counts = df["city"].value_counts()
 
plt.figure(figsize=(7, 4))
plt.bar(counts.index, counts.values)
plt.title("Orders count by city")
plt.xlabel("City")
plt.ylabel("Count")
plt.tight_layout()
plt.show()

3) Line plot (trend)

Trend

import matplotlib.pyplot as plt
 
plt.figure(figsize=(7, 4))
plt.plot(df["date"], df["orders"], marker="o")
plt.title("Orders trend")
plt.xlabel("Date")
plt.ylabel("Orders")
plt.xticks(rotation=20)
plt.tight_layout()
plt.show()

4) Scatter plot

Scatter

import matplotlib.pyplot as plt
 
plt.figure(figsize=(7, 4))
plt.scatter(df["orders"], df["amount"], alpha=0.7)
plt.title("Orders vs amount")
plt.xlabel("Orders")
plt.ylabel("Amount")
plt.tight_layout()
plt.show()

Deliverable

Save some plots with plt.savefig(...) and reuse this structure for your datasets.

From the book

McKinney reuses one pattern across every chart in this pack: build a Figure/Axes pair, call the right plotting method (.hist, .bar, .plot, .scatter), then finish with set_title/set_xlabel/set_ylabel and tight_layout(). Once that four-step habit is automatic, adding a new chart type is just swapping one method call.

Next

You’ve completed the Matplotlib phase — continue to the next phase to bring these charts together with pandas’ own .plot() shortcuts.

🧪 Try It Yourself

Exercise 1 – Distribution of a column

pch.exerciseTag pch.exerciseTitle

Exercise 2 – Counts by category

pch.exerciseTag pch.exerciseTitle

Exercise 3 – Trend and relationship together

pch.exerciseTag pch.exerciseTitle