Plotting credal sets on a spider (radar) chart

The plot_credal_set() function automatically switches to a spider (radar) plot for credal sets with 4 or more classes. Each class corresponds to a spoke, and probability bounds are rendered as constant-width bars.

The same five credal set types are supported:

• ArraySingletonCredalSet – a closed envelope connecting point probabilities on each spoke.

• ArrayProbabilityIntervalsCredalSet – constant-width bars on each spoke from lower to upper bound.

• ArrayDistanceBasedCredalSet – the same bars plus a marker at the nominal distribution.

• ArrayConvexCredalSet – member distribution lines with a filled min/max envelope.

• ArrayDiscreteCredalSet – individual member distributions as distinct colored lines.

An optional ground_truth overlay (as an ArraySingletonCredalSet) can be added to any plot type.

from __future__ import annotations

import matplotlib.pyplot as plt
import numpy as np

from probly.plot import PlotConfig, plot_credal_set
from probly.representation.credal_set.array import (
    ArrayConvexCredalSet,
    ArrayDiscreteCredalSet,
    ArrayDistanceBasedCredalSet,
    ArrayProbabilityIntervalsCredalSet,
    ArraySingletonCredalSet,
)

NUM_CLASSES = 8
CLASS_LABELS = [f"Class {i}" for i in range(NUM_CLASSES)]

Singleton credal set

A single probability distribution shown as a closed envelope.

singleton = ArraySingletonCredalSet(
    array=np.array([[0.35, 0.20, 0.15, 0.10, 0.08, 0.05, 0.04, 0.03]]),
)
plot_credal_set(singleton, title="Singleton", labels=CLASS_LABELS)
plt.show()

Probability intervals

Constant-width bars show the per-class probability bounds.

intervals = ArrayProbabilityIntervalsCredalSet(
    lower_bounds=np.array([[0.05, 0.02, 0.02, 0.01, 0.01, 0.01, 0.01, 0.01]]),
    upper_bounds=np.array([[0.80, 0.30, 0.20, 0.15, 0.10, 0.10, 0.08, 0.05]]),
)
plot_credal_set(intervals, title="Probability Intervals", labels=CLASS_LABELS)
plt.show()

Probability intervals with ground-truth overlay

A ground-truth overlay is shown as a star marker.

gt = ArraySingletonCredalSet(
    array=np.array([[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]]),
)
plot_credal_set(
    intervals,
    title="Probability Intervals with Ground Truth",
    labels=CLASS_LABELS,
    ground_truth=gt,
)
plt.show()

Distance-based credal set

Interval bars derived from a nominal distribution and radius, with a marker at the nominal.

distance_based = ArrayDistanceBasedCredalSet(
    nominal=np.array([[0.30, 0.20, 0.15, 0.10, 0.08, 0.07, 0.05, 0.05]]),
    radius=np.array([0.05]),
)
plot_credal_set(distance_based, title="Distance-Based", labels=CLASS_LABELS)
plt.show()

Convex credal set

Multiple vertex distributions drawn as prominent lines with a filled min/max envelope.

convex = ArrayConvexCredalSet(
    array=np.array([
        [
            [0.50, 0.15, 0.10, 0.08, 0.07, 0.05, 0.03, 0.02],
            [0.10, 0.40, 0.15, 0.10, 0.10, 0.05, 0.05, 0.05],
            [0.15, 0.10, 0.35, 0.15, 0.10, 0.05, 0.05, 0.05],
        ],
    ]),
)
plot_credal_set(convex, title="Convex", labels=CLASS_LABELS)
plt.show()

Discrete credal set

Each member distribution is drawn as its own colored line.

discrete = ArrayDiscreteCredalSet(
    array=np.array([
        [
            [0.40, 0.20, 0.15, 0.10, 0.08, 0.03, 0.02, 0.02],
            [0.10, 0.35, 0.20, 0.15, 0.08, 0.05, 0.04, 0.03],
            [0.15, 0.10, 0.30, 0.20, 0.10, 0.07, 0.05, 0.03],
        ],
    ]),
)
plot_credal_set(discrete, title="Discrete", labels=CLASS_LABELS)
plt.show()

Custom styling

Adjust visual parameters through PlotConfig.

config = PlotConfig(fill_alpha=0.5, line_width=2.5, spider_bar_width=0.08)
plot_credal_set(
    intervals,
    title="Custom Styling",
    labels=CLASS_LABELS,
    config=config,
)
plt.show()