Skip to content
Python module to implement BNA logic in a PostGIS database
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.



Python module to implement BNA logic on a PostGIS database. Uses a configuration file for customizability. The current implementation requires a user to run the old-style BNA on the command line to import datasets. This library can build the routeable network, process the connectivity routing, and calculate scores.


pyBNA is tested with Python 2.7. The following libraries are required:

  • psycopg2
  • tqdm
  • pyyaml
  • geopandas
  • munch
  • overpass
  • omsnx (version 0.9 until Python 3 is supported)

You can install these via pip:

pip install psycopg2 tqdm pyyaml geopandas munch overpass osmnx==0.9

*Special note for Windows users: Installing geopandas can be a real pain. If you don't already have a working version of Geopandas, we suggest following this excellent guide from Geoff Boeing.

The osmium package is also required for parsing OSM destinations from an .osm extract, but if you're downloading destinations directly from OSM you can safely skip this dependency.

A note on moving to Python 3

We are aware that Python 2 will be deprecated soon in favor of Python 3. Our intention is to make this change in pyBNA in the near future. We do not expect the switch to disrupt the pyBNA API in any way so this change should not be overly disruptive for most users, apart from having to run pyBNA with the latest Python 3 version.

tl;dr (Simple run)

The most simple BNA run, using stock datasets and no customization, can be completed in a few easy steps. The following assumes you already have a database running named "bna" on the local machine.

import pybna

# imports
i = pybna.Importer()

# stress
s = pybna.Stress()

# connectivity
bna = pybna.pyBNA()

Importing data

pyBNA includes a workflow to import data from publicly available sources (for the United States, at least). Automatic import of demographic data relies on US Census 2010 data. Street network and bicycle facility data is imported from OpenStreetMap.

For more guidance on the import process, see our import instructions.

Getting started

First, import pybna and create a pyBNA object by pointing it to the config file.

import pybna
bna = pybna.pyBNA(config="/home/spencer/dev/napa/bna/bna_vine_config.yaml")

Next, you can calculate the connectivity with


Lastly, you can generate block-level scores with


Configuration file

Most options in pyBNA are managed using a configuration file. This file is passed as an argument when creating the pyBNA object and tells pyBNA important things about your data and the assumptions you want to make in the analysis. The configuration file is written using YAML.

There's more information about the configuration file here

Travel sheds

Once you've completed the connectivity analysis, you can develop a low/high stress travel shed for any census block with

bna.travel_sheds([list, of, block, ids, here], my_travel_shed_table)
You can’t perform that action at this time.