pgNetworks

convert geographic data like address coordinates or other points of interest and lines like road networks or utility trenches into a graph data model for routing and analytics purposes.

what we're doing here:

1. Get Data

   1. Download
   2. Filter for road network
   3. Filter for POI
   4. Load to Postgres
      1. Vertex Data Model
      2. Linestring staging table

2. Create Graph

   1. Create connections between vertices and lines
   2. Segmentize linestrings into "true" edges
   3. Dissolve edges over nodes with degree 2

3. Create ancillary functions

   1. selector structures for parallel processing
   2. routing scenarios modeled as materialized views
   3. tbd

what we get:

• a proximity sorted inventory of point data
• a logical network that allows for graph data operations like routing or nearest neighbour searches
• a geometrical network that allows cartographic visualization and ad hoc spatial analysis with access to the graph data model

Overview:

folder database

folder etl

folder data