matplotlib: Drawing charts with Python

In short: matplotlib is the standard library for charts in Python: lines, bars, scatter plots and more, directly from lists or pandas tables.

In more detail: Numbers become understandable only through pictures. matplotlib produces charts for reports, notebooks and publications. seaborn (nicer statistical charts) and the plot functions of pandas build on it.

In Depth

First chart

import matplotlib.pyplot as plt
 
months = ["Jan", "Feb", "Mar", "Apr"]
revenue = [120, 150, 90, 180]
 
fig, ax = plt.subplots(figsize=(6, 3))
ax.plot(months, revenue, marker="o", label="Revenue")
ax.set_title("Revenue per month")
ax.set_xlabel("Month")
ax.set_ylabel("Euro")
ax.grid(True, alpha=0.3)
ax.legend()
fig.savefig("revenue.png", dpi=150, bbox_inches="tight")
plt.show()

With fig, ax = plt.subplots() you work object-oriented (recommended): fig is the figure, ax the drawing area.

Important chart types

• Line ax.plot: trends over time.
• Bar ax.bar / ax.barh: compare categories.
• Scatter ax.scatter: relationship of two quantities.
• Histogram ax.hist: distribution of values.
• Box plot ax.boxplot: spread and outliers.
• Pie ax.pie: shares (use sparingly, bars are usually more legible).
• Several panels: plt.subplots(2, 2).

With pandas

import pandas as pd
df = pd.read_csv("sales.csv")
df.groupby("month")["revenue"].sum().plot(kind="bar")

Often it takes one line. For interactive charts there are Plotly, Altair, Bokeh; in the browser Chart.js or D3 are suitable.

Good charts

Label axes (with units), choose a clear title, use colours sparingly and colour-blind-friendly, start bar charts at zero, avoid cluttered grids and 3D effects and show one message per chart. Save figures in sufficient resolution or as SVG/PDF (vector-based).

See also: Data visualisation, pandas, Conda and Jupyter, SVG vector graphics