Visualization Module
The spine.vis module provides interactive visualization tools for detector inputs, reconstructed objects, truth labels, geometry, and training or evaluation products. It is primarily Plotly-based and is designed for inspecting sparse 3D reconstruction outputs rather than serving as the center of the package.
Visualization helpers, trace builders, and domain-aware drawers.
The top-level spine.vis namespace re-exports the public visualization
API while the implementation is organized internally into:
spine.vis.tracefor low-level Plotly trace buildersspine.vis.drawerfor higher-level object and detector drawersspine.vis.layoutfor shared Plotly and Matplotlib stylingspine.vis.metricfor metric-specific plotting helpers
Module Index
Use this package to inspect detector point clouds, overlay reconstructed or truth objects, and visualize model or evaluation outputs in notebooks and analysis workflows.
Shared Plotly layout, color, and styling helpers for visualization code. |
|
Visualization entrypoint for reconstructed and truth output objects. |
|
Tools to draw 2D or 3D point clouds. |
|
Module to draw 3D arrows. |
|
Defines functions used to draw finite-sized boxes. |
|
Module to convert a point cloud into a cone envelope. |
|
Module to draw cylinders. |
|
Module to convert a point cloud into an ellipsoidal envelope. |
|
Module to convert a point cloud into an convex hull envelope. |
|
Tools to draw voxelized data organized in clusts. |
|
Draw truth-particle labels on top of labeled point clouds. |
|
Draw graph connectivity on top of clustered point clouds. |
|
Draw lite data structures with compact geometric surrogates. |
|
Training-history visualization helpers. |
|
Detector-geometry drawers and helpers. |
|
Metric-specific visualization helpers. |