UVisBox

Uncertainty Visualization Toolbox — a Python library for visualizing scientific uncertainty data, developed at the SCI Institute, University of Utah.

Implemented Methods

• Uncertainty Tube — Uncertainty in 3D trajectory data. arxiv
• Contour Boxplot — Summarizing isocontours in ensemble scalar fields. doi
• Functional Boxplot — Band envelopes for functional data.
• Curve Boxplot — Band envelopes for curve data.
• Squid Glyphs (2D & 3D) — Vector field uncertainty visualization. doi
• Cone Glyphs — 3D cone glyph visualization for vector field uncertainty.
• Uncertainty Lobes — Angular wedge representation of vector uncertainty.
• Probabilistic Marching Squares / Cubes / Triangles / Tetrahedra — Isosurface extraction with crossing probabilities.
• VSUP — A colormap designed for uncertain data. link

Installation

Requires Python >= 3.11, < 3.14. Uses Poetry for dependency management.

poetry install

Optional parallel computing dependencies (numba, joblib):

poetry install --with parallel

Dependencies

Core: numpy, scipy, matplotlib, scikit-learn, scikit-image, pyvista, basemap, xarray, dsapi

Quick Start

from uvisbox import uncertainty_tubes, contour_boxplot, functional_boxplot
from uvisbox import squid_glyph_2D, squid_glyph_3D, cone_glyph
from uvisbox import curve_boxplot, uncertainty_lobes
from uvisbox import probabilistic_marching_squares, probabilistic_marching_cubes

See the examples/ directory for usage demos, including:

Example	Description
uncertainty_tube_example.py	3D trajectory uncertainty tubes
contour_boxplot_example.py	Ensemble scalar field contours
functional_boxplot_example.py	Functional data boxplots
curve_boxplot_example.py	Curve data boxplots
squid_glyphs_2D_example.py	2D vector field glyphs
squid_glyphs_3D_example.py	3D vector field glyphs
uncertainty_lobes_2D_example.py	Angular uncertainty representation
probabilistic_marching_squares_example.py	Probabilistic contour extraction
custom_boxplot_style_example.py	Custom styling with BoxplotStyleConfig
wind_ensemble_example.py	Real climate data visualization
neural_UQ_example.py	Neural network uncertainty quantification
vsup_example.py	VSUP colormap demo

Run any example with:

poetry run python examples/uncertainty_tube_example.py

Project Structure

uvisbox/
├── Core/
│   ├── BandDepths/        # Band depth calculations (functional, contour, curve, vector)
│   ├── CellsCrossingProb/ # Cell crossing probability algorithms
│   ├── Colors/            # LAB color interpolation and ColorTree
│   ├── CommonInterface/   # Shared BoxplotStyleConfig
│   └── Interpolations/    # Linear interpolation utilities
├── Modules/
│   ├── UncertaintyTube/
│   ├── ContourBoxplot/
│   ├── FunctionalBoxplot/
│   ├── CurveBoxplot/
│   ├── SquidGlyphs/
│   ├── ConeGlyphs/
│   ├── UncertaintyLobes/
│   ├── ProbabilisticMarchingSquares/
│   ├── ProbabilisticMarchingCubes/
│   ├── ProbabilisticMarchingTriangles/
│   └── ProbabilisticMarchingTetrahedra/
└── Datasets/              # Dataset loaders (double gyre, hurricane, ERA5, etc.)

Each visualization module follows a 3-stage pipeline: Stats → Mesh → Visualization.

Testing

poetry run pytest

License

MIT