Matplotlib

Overview

Matplotlib is the original and most widely used plotting library in the Python data ecosystem. It exposes both a high-level, MATLAB-style interface (pyplot) for quick charts and a lower-level, object-oriented API (Figure, Axes) for precise control over every element of a plot. Most other Python visualization libraries — pandas' built-in .plot(), seaborn, and many dashboarding tools — are built directly on top of Matplotlib or use it as a rendering backend.

Basic usage

Copy code
import matplotlib.pyplot as plt

fig, ax = plt.subplots()
ax.plot(dates, revenue, label="Revenue")
ax.set_xlabel("Date")
ax.set_ylabel("Revenue ($)")
ax.legend()
plt.savefig("revenue.png")

The object-oriented API (working with fig and ax explicitly) is generally recommended over the implicit plt.plot() shortcuts once a script needs more than a single simple chart, since it scales better to multi-panel figures and reusable plotting functions.

Typical workflow with a query engine

Matplotlib doesn't query data itself — it plots whatever arrays or DataFrame columns you give it. A common pattern is to run aggregation and filtering in SQL against a fast engine like DuckDB, pull back a small, already-summarized result, and hand that off to Matplotlib for charting, rather than loading a full raw dataset into memory just to plot a handful of aggregate values.

Copy code
import duckdb
import matplotlib.pyplot as plt

df = duckdb.sql(
    "SELECT date_trunc('month', order_date) AS month, sum(amount) AS revenue "
    "FROM read_parquet('orders.parquet') GROUP BY ALL ORDER BY ALL"
).df()

plt.plot(df["month"], df["revenue"])
plt.show()

This keeps the heavy lifting in the query engine and leaves Matplotlib to do what it's good at: rendering the final chart.