Nodes Overview
The nodes are organised into structured categories that reflect a coherent computational workflow.
Each category represents a distinct operational domain inside the node tree and defines how data is introduced, transformed, computed, and communicated.
The organisation aligns with Sverchok’s tree evaluation logic, where nodes execute when linked, pass structured data downstream, and remain modular and stateless wherever possible
(see:Sverchok Tree Evaluation System ).
1. Gathering
What they do
Gathering nodes import and structure external data sources so they can be processed inside the node tree.
They handle:
- GIS vectors (GeoJSON, Shapefile, GeoPackage)
- Raster DEM files (GeoTIFF)
- LAS point clouds
- OSM and street networks
- Street imagery
- CSV / JSON files
- API requests
- Pandas DataFrames and Series
Why they exist
Design computation depends on contextual data. These nodes translate heterogeneous external formats into structured computational objects (arrays, graphs, DataFrames, geometry), ensuring interoperability and reproducibility.
They define the entry point of most workflows.
2. Analysis
What they do
Analysis nodes compute metrics, evaluate spatial relationships, and extract structured information from datasets.
They operate on:
- DEM arrays (terrain attributes)
- NetworkX graphs (centrality, shortest path)
- GeoDataFrames and DataFrames (correlation, filtering, indexing)
- Visibility calculations (isovists)
- Machine learning pipelines (fit, predict, evaluate)
- External GIS engines (Whitebox Tools)
Why they exist
They convert raw datasets into measurable indicators and quantitative descriptors.
This layer separates data ingestion from computational reasoning and supports decision-making through metrics rather than direct geometry production.
3. Generation
What they do
Generation nodes transform computed data into geometry-ready or parametric structures.
They include:
- Coordinate transformations (Lat/Lon ↔︎ Projected)
- Face construction from vertices
- DataFrame reshaping (transpose, filter, mapping)
- Colormap encoding
- File format conversion
- Dictionary construction
Why they exist
They bridge analysis and geometry.
Metrics and attributes are converted into parameters, mesh structures, or visual encodings that can drive modelling and design exploration.
4. Visualisation
What they do
Visualisation nodes produce interpretative and deployable outputs.
They include:
- Seaborn statistical plots
- DEM plotting
- Plotly dashboards
- Bokeh dashboards
- Map dashboards
- Streamlit dashboard assembly
- WebVR connectors
- Local Python servers
Many dashboard nodes generate executable Python snippets rather than rendering directly in Blender.
Why they exist
Computation must remain inspectable and communicable.
These nodes expose analytical results, enable interaction, and support deployment outside the modelling environment.
Structural Logic
The categories reflect a computational separation:
- Data ingestion is isolated from computation
- Computation is isolated from transformation
- Transformation is isolated from representation
- Representation can be deployed independently
Data types remain explicit across the system:
- Geometry (Vertices, Faces)
- Graphs (NetworkX)
- Raster arrays
- DataFrames / Series
- JSON structures
- Dashboard code snippets