Skip to content
This repository has been archived by the owner on Feb 1, 2021. It is now read-only.

mansueto-institute/prclz-proto

 
 

Repository files navigation

p  r  c  l  z

"parcelize"

overview/

prclz is a library for:

  • generating vectors for street blocks in a neighborhood using street network information from OpenStreetMap
  • creating parcels/cadastral maps for building footprints and other land use features within each block
  • calculating structural and areal features (k-complexity of access graph, percentage of land used by buildings, etc) at the block and neighborhood level

structure/

  • prclz: main library

    • blocks: extract vector polygons representing street blocks from street network information
    • features: calculate per-block features
    • parcels: generate cadastral parcels tesselating each block given building/land use footprints
    • topology: planar graph implementation, with tools to calculate weak duals. largely taken from openreblock/topology.
  • reblock: tools to implement Steiner tree approximations of reblock graphs

  • smoketests: not quite unit tests, but visual tests to make sure things look right

  • osm-etl: scripts to perform ETL on *.osm.pbf files and extract relevant features.

  • midway: end-to-end workflow to be run on RCC's midway cluster

  • requirements: required packages

development setup/

  1. on midway, load the necessary modules
module load intel/18.0
module load gdal/2.2
module unload python
module load Anaconda3/5.1.0
module load parallel
  1. Set up a conda virtual environment, and activate it.
. ./midway/conda_setup.sh
conda create --name mnp
source activate mnp
  1. Install the requirements. (Due to version pinning/compatibility, some packages aren't in the standard conda repos, so we install them with pip).
conda install --name mnp -f -y -q -c anaconda -c conda-forge --file requirements/conda-requirements.txt
pip3 install -r requirements/pip-requirements.txt
  1. From the top-level directory, install prclz in editable mode.
pip3 install -e .
  1. Deactivate your virtual environment once you're done
conda deactivate

About

geospatial hierarchy and structure extraction

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 92.3%
  • Shell 4.5%
  • R 3.2%