Chapter 12: Spatial Statistical Learning
Boost your spatial models! Learn about spatial autocorrelation, cross-validation, and machine learning with `mlr3`, using a real-world landslide prediction case study.
Chapter 12: Spatial Statistical Learning
Boost your spatial models! Learn about spatial autocorrelation, cross-validation, and machine learning with `mlr3`, using a real-world landslide prediction case study.
New R package alert: mbg for model-based geostatistics
Run spatial ML & geostatistical models to estimate continuous surfaces from point data + raster covariates.
Built on sf, terra, data.table, caret, and R-INLA.
Final part of our Spatial ML with R series !
We explore spatial cross-validation with sperrorest & blockCV — tools outside the usual ML frameworks
URL: https://geocompx.org/post/2025/sml-bp6/
Okay - further to my early rants about CDSE data, it aint as bad as I thought it also prompted me to properly sort out my approach to scaling/offsets which had been driving me mad! So if anyone cares for another way to download data from CDSE, with #rspatial / #gdal I made a gist:
https://gist.github.com/h-a-graham/86cd3403445cf163ce958efa2d29c621
There are still some improvements to be made for sure.
FYI @Micha_Silver
Chapter 11: Writing Geoalgorithms
Focuses on developing reusable and reproducible code for spatial tasks in R. Demonstrates algorithm design using examples like calculating polygon centroids.
¡Nueva edición del libro Usa R como Sistema de Información Geográfica!
Me da mucho gusto compartir que ya está disponible la nueva edición de este libro, escrito originalmente por Jean-François Mas, y que en esta versión suma a nuevos colaboradores y colaboradoras, ampliándose con capítulos adicionales y casos de uso aplicados.
Puedes descargarlo gratuitamente en formato digital aquí:
https://u.pcloud.link/publink/show?code=XZs74u5ZSy6RJH0Oj6kLUd43oBlXz0lUcX1V
The first major version of {mapsf} has arrived on CRAN!
mapsf is a thematic mapping R package. Its goal is to be simple and lightweight while offering all the necessary features to create beautiful statistical maps.
This release includes a revamped theming system, an updated cheat sheet and improved PNG and SVG exports and solves some long-lasting display bugs.
More details in this blog post: https://rcarto.github.io/posts/mapsf_v1.0.0
New website: https://riatelab.github.io/mapsf
A heads-up about the Geocomputation with R book: some copies were mistakenly printed in black & white instead of full color. If you received one, please contact me or the publisher for a replacement. A new, correct copy will be sent to you!
Publisher: https://www.routledge.com/contacts/customer-service
I finally manage to watch @paleolimbot presentation at @RConsortium on "scaling the #Rspatial ecosystem" !
https://www.youtube.com/watch?v=tjNEoIYr_ag
Quick subjective key points:
- Use the database Luke and learn a bit SQL (I was already converted)
- the diversity of R packages to do some workflows also represent the diversity of standards (s.f.) and steps to reach similar results
- wkt_filter seems very nice (I was using "query" and GDAL/SQL instead)
Chapter 10: Bridges to GIS Tools
Shows how to connect R with external GIS tools like QGIS, GRASS, and SAGA. Also includes guidance on working with GDAL, spatial databases, and cloud-based services.
New blog post! Part 5 of our series on spatial ML with #RStats explores specialized packages: RandomForestsGLS, spatialRF, and meteo -- tools beyond caret, tidymodels, & mlr3.
TOMORROW! Scaling the r-spatial ecosystem for the modern composable data pipeline - with Dewey Dunnington, Senior Software Engineer at Wherobots
Tues, June 24, 1pm ET
Sign up now! https://r-consortium.org/webinars/scaling-the-r-spatial-ecosystem-for-the-modern-composable-data-pipeline.html
Registration is open for Spatial Data Science across Languages (SDSL) 2025 – Sept 17–18 (+19), Salzburg, Austria.
Connect R, Python, Julia & more in spatial science.
https://forms.gle/E9fpG88V2VQQKmjk9 -- Apply for on-site by mid-July – limited spots.
Chapter 9: Mapping & Visualization
Introduces tools for making static, animated, and interactive maps in R. Covers the tmap package and web-based options for sharing geographic data visually.
From Brittany Barker: ‘My "GIS and Mapping in R" workshop for the Cascadia R Conference . . . is available at GitHub and includes four exercises that focus on using "sf", "terra", "ggplot2", and "leaflet" for geospatial analyses and creating static and interactive maps’
#RStats #RSpatial
(Barker is an asst research professor at Oregon State University in Portland)
https://github.com/bbarker505/CASCADIA_R_Intro_to_GIS_2025
Ooh CNN-based cloud masking baked right into a VRT file. This should be fun! #rstats #rspatial #gdal #python
This is using https://github.com/DPIRD-DMA/OmniCloudMask and appears to be a massive improvement on the standard Sentinel 2A SCL band!
I should start collecting all the ways in which #RSpatial can go wrong spectacularly.
Chapter 8: Geographic Data I/O
Covers how to read and write spatial data in various formats, access open geoportals, and work with geographic web services in R. Includes tips on metadata and exporting maps.
I gave a talk on measuring spatial autocorrelation in spatial machine learning at the #AGILE0GI conference.
Slides: https://jakubnowosad.com/agile-gi2025/
Always glad to discuss spatial ML or autocorrelation --- feel free to reach out!