Jupyter notebook index
All four notebooks remain in their original repository paths. They serve different roles: canonical reproduction, validation, implementation demonstration, and optional 3-D exploration.
Canonical reproduction
reproduce_fire_vase_pipeline.ipynb
Purpose: Walk through environment setup, data-lake acquisition or rebuild, manuscript analysis, figure rendering, whole-pipeline checks, and release refresh.
Inputs: Repository environment; shared v0.1 data lake or local FIRED and gridMET source caches; project configuration.
Outputs: Current analysis products, manuscript figures, validation and reproducibility reports, and optionally a refreshed shareable data-lake package.
Place in the analysis: Primary notebook companion to the full reproduction vignette. The current quantitative generation is v2; historical stages remain explicitly labeled.
Scientific and software validation
validate_fire_vase_pipeline.ipynb
Purpose: Run one inspectable section per production validation module and the separate expected-failure contrast.
Inputs: Materialized FIRED/gridMET data lake, validation configuration, and the installed repository environment. The independent external-source section requires network access.
Outputs: Module-level PNG, CSV, HTML, and JSON evidence plus the collated suite report.
Place in the analysis: First-class validation notebook. It checks pipe and stream behavior, cube/HTML integrity, 2-D and 3-D geometry, climate attribution, external sources, and deliberate corruptions.
Implementation demonstration
06_vase_volume_basic.ipynb
Purpose: Demonstrate basic VASE masking and viewer overlays using both verb and pipe/grammar APIs.
Inputs: A small synthetic xarray cube constructed in the notebook and a
two-section VaseDefinition; no external research data required.
Outputs: Baseline and masked cube views plus a VASE-outline overlay.
Place in the analysis: Reusable implementation example, not a source of manuscript statistics or figures.
Optional 3-D exploration
07_vase_volume_3d_viz.ipynb
Purpose: Show point extraction and optional scientific 3-D visualization with PyVista.
Inputs: A small synthetic xarray cube and VaseDefinition created in the
notebook; PyVista is optional.
Outputs: Extracted point arrays and, when the optional dependency is available, an interactive 3-D scatter view.
Place in the analysis: Exploratory visualization example, not a canonical paper or validation workflow.
Preservation and execution notes
- Notebook paths are unchanged; no notebook was deleted, merged, or replaced with prose.
- The notebook index groups assets conceptually without moving them.
- Canonical commands also appear as scripts so automated reproduction does not depend on interactive execution.
- Large real-data inputs remain outside Git and are resolved through the documented data-lake path.