A polar scatter plot places each point by angle and distance from the center instead of by x and y. It is a natural fit for directional data like wind, bearings, and around-the-clock measurements, and both Matplotlib and Plotly support it.

Quick Example

Matplotlib needs a polar projection; Plotly has a dedicated scatterpolar trace.

# Matplotlib
import matplotlib.pyplot as plt
import numpy as np

fig = plt.figure()
ax = fig.add_subplot(projection="polar")
ax.scatter(theta_rad, r)

# Plotly
import plotly.graph_objects as go

fig = go.Figure(go.Scatterpolar(theta=theta_deg, r=r, mode="markers"))
fig.show()

Both take an angle and a distance from the center. This post builds the same wind-by-direction dataset in each.

Matplotlib polar scatter

Matplotlib draws polar plots through a projection. Create a polar axes and call scatter() as usual, but with the angle in radians.

import matplotlib.pyplot as plt
import numpy as np

angles_deg = np.arange(0, 360, 15)
a = np.deg2rad(angles_deg)

fig = plt.figure()
ax = fig.add_subplot(projection="polar")
ax.scatter(a, speed, s=70, c=speed, cmap="viridis",
           edgecolor="white", linewidth=0.8)

The plot defaults to zero pointing east and angles increasing counterclockwise. Real compass data is better served with North at the top and a clockwise sweep, so flip it and label the ticks with compass points.

compass = ["N", "NE", "E", "SE", "S", "SW", "W", "NW"]

ax.set_theta_zero_location("N")
ax.set_theta_direction(-1)
ax.set_xticks(np.deg2rad(np.arange(0, 360, 45)))
ax.set_xticklabels(compass)

Each point is placed by angle and radius, and the points are colored by speed through matplotlib's scatter c and cmap arguments.

Plotly polar scatter

Plotly's scatterpolar trace takes theta in degrees and r for the radius, so no conversion is needed.

import plotly.graph_objects as go

fig = go.Figure(
    go.Scatterpolar(
        r=speed,
        theta=angles_deg,
        mode="markers",
        marker=dict(size=12, color=speed, colorscale="Viridis",
                    showscale=True),
    )
)
fig.update_layout(
    polar=dict(
        angularaxis=dict(
            tickmode="array",
            tickvals=list(range(0, 360, 45)),
            ticktext=["N", "NE", "E", "SE", "S", "SW", "W", "NW"],
        )
    )
)
fig.show()

The result is interactive: hover a point to read its bearing and speed, and use the mode bar to zoom and pan. The chart below is embedded directly.

When to use a polar scatter

Reach for a polar scatter when the two coordinates really are an angle and a distance: wind direction and strength, sensor readings around a hub, or timestamps mapped to an angular view. If the axes are just two numbers, a plain x-y scatter is usually simpler and easier to interpret.

Tweak the angular axis

Both libraries default to non-compass conventions, so set the reference and direction to match your data.

Matplotlib:

ax.set_theta_zero_location("N")  # where 0 points
ax.set_theta_direction(-1)       # clockwise
ax.set_xticks(np.deg2rad(np.arange(0, 360, 45)))

Plotly:

fig.update_layout(
    polar=dict(
        angularaxis=dict(
            tickmode="array",
            tickvals=list(range(0, 360, 45)),
            ticktext=["N", "NE", "E", "SE", "S", "SW", "W", "NW"],
        )
    )
)

Using compass labels instead of raw degrees makes a wind or bearing chart readable at a glance.

Practical Tips

• Convert angles to radians for Matplotlib; Plotly's theta uses degrees.
• Color the markers with a third dimension (speed, magnitude) to add signal without a second plot.
• Orient the zero point and direction to match what the data means, whether that is North or something else.
• Use mode="markers" or "markers+lines" in Plotly depending on whether you want the angular path drawn.
• Use Matplotlib for a static print-ready figure and Plotly when readers should explore or zoom.

Similar Topics

• Scatter Plots in Matplotlib
• Interactive Sunburst Charts with Plotly
• Dumbbell Plots in Matplotlib and Plotly