For more, check out the open-access article: doi.org/10.1111/gean...

You can just as easily work with urban amenities/points of interest, building footprints, transit stops, elevation data, street orientations, speed/travel time, and routing. It recently reached version 2.0 with a slew of new features and enhancements.

If you haven't used it before OSMnx is a Python package to download, model, analyze, and visualize street networks and any other geospatial features from OpenStreetMap. You can download and model walking, driving, or biking networks with a single line of code then quickly analyze and visualize them.

All of these lessons have become central to the work my RAs do in the Urban Data Lab at USC. They're not always easy, but they make a clear improvement in research quality, clarity, and reusability that directly impacts our downstream empirical analyses and scientific theorizing.

What makes a good API, and why is it so hard (or is it just me)?

How your development pipeline can make or break your quality of life as an open-source developer

Dependency ecosystems and the fine line between dependency heaven and dependency hell

How we can advance reusable geospatial software

That's fantastic! I coauthored this surfacic networks paper with Marc, so I know you're all in for a treat! 😊

Congrats Marta! This looks fantastic.

And it has received hundreds of contributions from many other code contributors. Thank you to everyone who helped make this possible.

I hope you find the package as useful as I do. Now I'm looking forward to all your bug reports 😂

On a personal note, this has now been a labor of love for me for about 9 years. Wow. I initially developed this package to enable the empirical research for my dissertation. Since then, it has powered probably 2/3 of the articles I've published over the years.

You can just as easily work with urban amenities/points of interest, building footprints, transit stops, elevation data, street orientations, speed/travel time, and routing. Get started here: osmnx.readthedocs.io

If you haven't used it before, OSMnx is a Python package to easily download, model, analyze, and visualize street networks and any other geospatial features from OpenStreetMap. You can download and model walking, driving, or biking networks with a single line of code.

We did! I'll have to check my notes when I get off from teaching later... I can't remember the reliability numbers off the top of my head.

For more on this technique and how you can use to to scale up qualitative text analysis in urban research, check out the article: www.sciencedirect.com/science/arti...

To do so, we train a BERT large language model that incorporates manual hand-labeling early in the process to yield a semi-automated technique that's pretty good at identifying nuance in natural language.

We identify topics discussed in LA rental listings' text to understand the subtle language of discrimination and other gatekeeping obstacles to housing.

Using Los Angeles's housing crisis and rental market as a case study, we demonstrate how and when modern AI and NLP techniques can generate qualitative insights on par with traditional manual techniques, but at a far larger scale and requiring far less labor.

I coauthored an article recently in the journal Computers, Environment and Urban Systems (which is also a perfect Kraftwerk album title) with Madison Lore and Julia Harten which takes up this challenge.

But if you could train a model to do, say, topic labeling for you, you'd be able to (potentially) analyze nearly unlimited text data nearly instantly after that initial training work. The mixed methods holy grail.

One area where urban AI research seems promising is in mixed methods work. For example, it's hard to use traditional qualitative methods on really large text data sets because of the overwhelming manual labor involved.