Try it yourself

Run this code directly in your browser. Click "Open in full editor" to experiment further.

Loading...

Click Run to see output

Or press Ctrl + Enter

How it works

Matplotlib 3D plots turn boring tables of (x, y, z) numbers into visualizations you can actually understand at a glance. 🌄

Why 3D Plots?

When your data has three variables that all relate to each other, a flat 2D chart can't show the whole story. 3D plots are perfect for:

Mathematical surfaces — visualizing functions like z = f(x, y)

Scientific data — temperature/pressure/altitude, terrain elevation

Machine learning — clustering visualized in feature space

Engineering — parametric curves, signal trajectories, motion paths

The Magic Line: `projection='3d'`

Matplotlib's 3D support lives inside mpl_toolkits.mplot3d, but you don't need to import it directly in modern Matplotlib. Just write:

ax = fig.add_subplot(111, projection='3d')

That single keyword argument flips the axes from flat 2D to a full 3D coordinate system. From there, you have a handful of go-to methods:

Method	What it draws
`ax.plot_surface(X, Y, Z)`	A solid colored surface (the "mountain")
`ax.plot_wireframe(X, Y, Z)`	A see-through mesh of lines
`ax.scatter(x, y, z)`	3D dots — great for clusters or point clouds
`ax.plot(x, y, z)`	A continuous 3D line — helices, trajectories
`ax.contour3D(X, Y, Z)`	Stacked contour rings, like a topographic map

The `meshgrid` Trick

For surfaces and wireframes, you don't pass simple lists — you need a 2D grid of every (x, y) coordinate pair. That's what np.meshgrid does:

x = np.linspace(-5, 5, 50)   # 50 x-values
y = np.linspace(-5, 5, 50)   # 50 y-values
X, Y = np.meshgrid(x, y)     # Two 50×50 matrices
Z = some_function(X, Y)      # A 50×50 height map

Now Z[i][j] is the height above the point (X[i][j], Y[i][j]). Matplotlib stitches them into a smooth surface.

Step-by-Step: Build a Surface Plot

1. Pick a function Z = f(X, Y) (or load real data shaped as a 2D array).

2. Create X, Y with np.meshgrid(...).

3. Compute Z from X and Y.

4. Make a figure: fig = plt.figure().

5. Add a 3D axis: ax = fig.add_subplot(111, projection='3d').

6. Draw: ax.plot_surface(X, Y, Z, cmap='viridis').

7. Label the axes and add a colorbar.

8. Call plt.show().

Pro Tips

Pick a good colormap: 'viridis', 'plasma', 'coolwarm', and 'terrain' all look great on surfaces.

Use `alpha=0.7` on scatter points so overlapping clusters stay readable.

Reduce mesh density with rstride / cstride if your wireframe looks too busy.

Add a colorbar with fig.colorbar(surf) — it shows what each color means.

Set viewing angle: ax.view_init(elev=30, azim=45) rotates the camera (great for screenshots).

When to Use Each Plot Type

Surface → smooth mathematical functions, terrain, heatmaps that need depth

Wireframe → showing structure without hiding the back of the surface

Scatter → discrete data points, ML clusters, sensor readings

Line → trajectories, parametric curves, time-evolving paths

Run the snippet above to see all four in action — you'll get a sinc surface, a 3-cluster scatter, a cos(x)·sin(y) wireframe, and a rising helix. ✨

Related examples

Data Science

Matplotlib 3D Plot in Python