Gerrymandering I

Feb-17 : In this lesson, we will practice reproducible geographic research with a research compendium.

Goals and Expectations

• Create a reproducible research compendium
• Plan a study to measure districting fairness in the context of gerrymandering
• Knit the analysis plan to html format and save to the docs folder
• Update the project readme.md and LICENSE files, and documentation of procedure_index.csv, data_index.csv, and data/metadata.
• Create a version release of your compendium on GitHub prior to analyzing any data
• Based on what you know of the reproducibility crisis and/or your own geographic research projects, post a short blog about the relative value or burden of starting a project with a compendium and analysis plan.
• Start implementing the study using R

Tutorial Application

In this tutorial, we will get an introduction to practicing open science by:

• Creating a reproducible research compendium from a GitHub template for reproducible research
• Using RStudio and R
• Planning and implementing a gerrymandering study papers for context and data for Alabama
• Reproducible R code for downloading and manipulating spatial data
• Learning vignette for Simple Features in R *updated with guidance on Geodesic area and perimeter calculations
• Starter example research compendium
• How to release a version

Software

• Base R: R project for statistical computing
• RStudio: posit downloads
• GitHub: Desktop Application

R packages for reproducible spatial analysis

• groundhog for reproducible computational environments (consistent versions of R and its packages)
• here for reproducible path names
• tidyverse includes dplyr for database-style data frames
• sf provides support for spatial vector data implementing the OSGeo simple features standards we are accustomed to
• stars spatial-temporal raster data in R
• tmap thematic maps, including static maps or interactive leaflet maps

Alternative packages

• sp alternative package for spatial vector data
• raster alternative raster data package
• mapsf alternative for static map cartography mapsf presentation slides with vignettes

Other Tutorial Resources

• R Markdown Code Chunks
• Spatial Data Science With applications in R by Pebesma and Bivand
• Geocomputation with R
• Spatial Data Science
• Introduction to R for Geospatial Data
• Introduction to Geospatial Raster and Vector Data with R

References

1. Nüst, D., and E. Pebesma. 2021. Practical Reproducibility in Geography and Geosciences. Annals of the American Association of Geographers 111 (5):1300–1310. DOI:10.1080/24694452.2020.1806028.
2. Wilson, J. P., K. Butler, S. Gao, Y. Hu, W. Li, and D. J. Wright. 2021. A Five-Star Guide for Achieving Replicability and Reproducibility When Working with GIS Software and Algorithms. Annals of the American Association of Geographers 111 (5):1311–1317. DOI:10.1080/24694452.2020.1806026.
3. Ibanez, L., W. J. Schroeder, and M. D. Hanwell. 2014. Practicing open science. In Implementing Reproducible Research, eds. V. Stodden, F. Leisch, and R. D. Peng, 241–280. Boca Raton: CRC Press.
4. Millman, K. J., and F. Perez. 2014. Developing Open-Source Scientific Practice. In Implementing Reproducible Research, eds. V. Stodden, F. Leisch, and R. D. Peng, 149–183. Boca Raton: CRC Press.
5. Nüst, D., C. Boettiger, and B. Marwick. 2018. How to Read a Research Compendium. arXiv:1806.09525.

Other templates and tools

1. Nüst and others are developing infrastructure for containerized executable compendia at o2r.info/
2. Docker containers allow researchers to reproduce and share their research processing environments as virtual machines which can be run on a server. www.docker.com/
3. Carl Boettiger’s research template
4. WORCS Workflow for Open Reproducible Code in Science R Package
5. Project TIER Protocol 4.0

Main Page