In November, I completed the 30 Day Map Challenge for the first time. I posted all of my submissions to Twitter/X and LinkedIn, and observed how the community reacted to each of my maps through likes, reposts, and comments. My most popular submission based on social media engagement was for Day 21: Raster. I’ve been using a technique for years called an “accessibility surface” to visualize proximity to locations. The accessibility surface is a raster dataset in which each grid cell represents the travel-time to that location from another given location, or the nearest location in a set of loca ..read more

My book Analyzing US Census Data: Methods, Maps, and Models in R, published last year, covers a lot of the data science tips and tricks I’ve learned over the years. In my academic and consulting work, I apply a lot of additional workflows that the book doesn’t cover. This year on the blog, I’d like to share some brief examples of workflows I’ve found useful with a focus on applications to Census and demographic data. In my consulting work, I’m commonly asked to build out maps, charts, or reports for a large number of cities or regions at once. The goal here is often to allow for rapid explorat ..read more

Travel-time isochrones are powerful analytical tools that represent the reachable area from a location for a given time and travel mode. In R, my package mapboxapi seamlessly integrates with R’s GIS infrastructure to allow for the use of Mapbox’s isochrones in spatial analysis workflows. In Python, there isn’t a package that directly connects Mapbox’s navigation toolkit to GeoPandas for spatial data analysis. However, these services are accessible via the routingpy Python package. In this blog post, I’ll present a workflow to help you connect GeoPandas with Mapbox’s isochrone services via rout ..read more

Earlier this month, I gave a two-part workshop series on analyzing the newly-released 2020 Decennial US Census Data with R. If you missed out on the workshop series, you can buy the videos and materials on my Workshops page. One topic I addressed was how to handle the impact of differential privacy on block-level accuracy in the new Census Data. Differential privacy refers to a method used by the Census Bureau to infuse “noise” into data products to preserve respondent confidentiality. Counts for larger areas and larger groups will still be accurate, but differential privacy makes smaller coun ..read more

Last month, version 8 of the LEHD Origin-Destination Employment Statistics (LODES) dataset was released. This long-awaited release includes data on workplaces, residences, and origin-destination flows for workers in 2020, along with a time series of these statistics back to 2002 enumerated at 2020 Census blocks. The latest release of the pygris package for Python enables programmatic access to these new data resources with its get_lodes() function. This new release also allows you to request Census geometry or longitude / latitude coordinates along with your LODES data, making data visualizati ..read more

Last week, I gave a workshop on working with geographic data in Python with the University of Michigan’s Social Science Data Analysis Network. The workshop focused on pygris, my new Python package for working with US Census Bureau geographic data resources. I was asked by multiple attendees if pygris works within ArcGIS Pro Python notebooks. I did not know the answer at the time, but it seemed like pygris - and other geospatial data packages in Python - combined with ArcGIS Pro could be quite powerful. I tested this process out, and it turns out that pygris and ArcGIS Pro work quite well toget ..read more

Spatial data science projects frequently require the calculation of proximity to resources. Analysts in fields like health care, real estate, retail, education, and more are commonly tasked with finding out what resources are near to a given location, and potentially develop strategies to fill identified gaps in resource access. In Chapter 7 of my book Analyzing US Census Data, I illustrate a workflow that shows how to analyze accessibility from Census tracts to major trauma hospitals in Iowa. In this post, I’ll show you how to reproduce the first part of that analysis in Python for the neighb ..read more

In Chapter 11 of my book Analyzing US Census Data, I explore a sampling of the variety of government datasets that are available for the United States. One of the most useful of these datasets is LODES (LEHD Origin-Destination Employment Statistics). LODES is a synthetic dataset that represents, down to the Census block level, job counts by workplace and residence as well as the flows between them. Given that LODES data are tabulated at the Census block level, analysts will often want to merge the data to Census geographic data like what is accessible in the pygris package. pygris includes a f ..read more