Kyle Walker
@kylewalker.bsky.social
Demographics | Geospatial | Data Science | Open Source
It didn't make it into the CRAN release, but I've just implemented draggable legends in the dev version of {mapgl}.
It's quite useful when the legend is getting in the way of your analysis!
#rstats
It's quite useful when the legend is getting in the way of your analysis!
#rstats
January 14, 2026 at 5:59 PM
It didn't make it into the CRAN release, but I've just implemented draggable legends in the dev version of {mapgl}.
It's quite useful when the legend is getting in the way of your analysis!
#rstats
It's quite useful when the legend is getting in the way of your analysis!
#rstats
New post: learn how to use the new interactive legends and screenshot control in R's mapgl package.
These features are great for interactive demos and visualizing custom scenarios without Shiny.
The new release (v0.4.4) is on CRAN now, so go get it!
walker-data.com/posts/mapgl-...
These features are great for interactive demos and visualizing custom scenarios without Shiny.
The new release (v0.4.4) is on CRAN now, so go get it!
walker-data.com/posts/mapgl-...
January 13, 2026 at 7:57 PM
New post: learn how to use the new interactive legends and screenshot control in R's mapgl package.
These features are great for interactive demos and visualizing custom scenarios without Shiny.
The new release (v0.4.4) is on CRAN now, so go get it!
walker-data.com/posts/mapgl-...
These features are great for interactive demos and visualizing custom scenarios without Shiny.
The new release (v0.4.4) is on CRAN now, so go get it!
walker-data.com/posts/mapgl-...
In the latest mapgl release: a native screenshot control.
Workflow: spin up a quick map; use the new interactive legend to filter your data; drag a marker; take a screenshot; share with colleagues.
Uses html2canvas internally to capture the exact state of your map view.
#rstats
Workflow: spin up a quick map; use the new interactive legend to filter your data; drag a marker; take a screenshot; share with colleagues.
Uses html2canvas internally to capture the exact state of your map view.
#rstats
January 12, 2026 at 8:51 PM
In the latest mapgl release: a native screenshot control.
Workflow: spin up a quick map; use the new interactive legend to filter your data; drag a marker; take a screenshot; share with colleagues.
Uses html2canvas internally to capture the exact state of your map view.
#rstats
Workflow: spin up a quick map; use the new interactive legend to filter your data; drag a marker; take a screenshot; share with colleagues.
Uses html2canvas internally to capture the exact state of your map view.
#rstats
Here's how the new interactive categorical legends in {mapgl} work.
Click a category in the legend to turn the corresponding features on/off on the map.
Not groundbreaking, but you get this out of the box with one argument.
It used to take me forever to code this manually!
Click a category in the legend to turn the corresponding features on/off on the map.
Not groundbreaking, but you get this out of the box with one argument.
It used to take me forever to code this manually!
January 11, 2026 at 8:58 PM
Here's how the new interactive categorical legends in {mapgl} work.
Click a category in the legend to turn the corresponding features on/off on the map.
Not groundbreaking, but you get this out of the box with one argument.
It used to take me forever to code this manually!
Click a category in the legend to turn the corresponding features on/off on the map.
Not groundbreaking, but you get this out of the box with one argument.
It used to take me forever to code this manually!
Since I started developing the mapgl R package I've envisioned built-in interactive legends.
They're finally here in the upcoming package release: a filter slider for continuous legends, and click-to-filter for categorical legends.
Blog post coming this week!
They're finally here in the upcoming package release: a filter slider for continuous legends, and click-to-filter for categorical legends.
Blog post coming this week!
January 11, 2026 at 6:23 PM
Since I started developing the mapgl R package I've envisioned built-in interactive legends.
They're finally here in the upcoming package release: a filter slider for continuous legends, and click-to-filter for categorical legends.
Blog post coming this week!
They're finally here in the upcoming package release: a filter slider for continuous legends, and click-to-filter for categorical legends.
Blog post coming this week!
I'm as excited about Claude Code / Opus 4.5 as anyone.
It's helped me create things this year that I couldn't have done otherwise.
But I notice often it lacks "common sense."
Meaning, it makes decisions that "work" but shouldn't be made.
It's helped me create things this year that I couldn't have done otherwise.
But I notice often it lacks "common sense."
Meaning, it makes decisions that "work" but shouldn't be made.
January 10, 2026 at 5:13 PM
I'm as excited about Claude Code / Opus 4.5 as anyone.
It's helped me create things this year that I couldn't have done otherwise.
But I notice often it lacks "common sense."
Meaning, it makes decisions that "work" but shouldn't be made.
It's helped me create things this year that I couldn't have done otherwise.
But I notice often it lacks "common sense."
Meaning, it makes decisions that "work" but shouldn't be made.
All 13.5 million jobs in Texas, visualized by sector with block-level precision.
The just-released 2023 LODES data is one of the most underrated public datasets.
The just-released 2023 LODES data is one of the most underrated public datasets.
January 8, 2026 at 9:05 PM
All 13.5 million jobs in Texas, visualized by sector with block-level precision.
The just-released 2023 LODES data is one of the most underrated public datasets.
The just-released 2023 LODES data is one of the most underrated public datasets.
Announcing spopt-r: spatial optimization for regionalization, facility location, and market analysis in R.
My all-time favorite Python package is PySAL's spopt - and until now, it did not have an R equivalent.
Get started here: walker-data.com/spopt-r
My all-time favorite Python package is PySAL's spopt - and until now, it did not have an R equivalent.
Get started here: walker-data.com/spopt-r
January 7, 2026 at 1:46 PM
Announcing spopt-r: spatial optimization for regionalization, facility location, and market analysis in R.
My all-time favorite Python package is PySAL's spopt - and until now, it did not have an R equivalent.
Get started here: walker-data.com/spopt-r
My all-time favorite Python package is PySAL's spopt - and until now, it did not have an R equivalent.
Get started here: walker-data.com/spopt-r
Mapping US wind energy infrastructure with hexagonal binning and "fade-on-zoom" effects.
Learn how: gist.github.com/walkerke/a4b...
Learn how: gist.github.com/walkerke/a4b...
January 5, 2026 at 4:16 PM
Mapping US wind energy infrastructure with hexagonal binning and "fade-on-zoom" effects.
Learn how: gist.github.com/walkerke/a4b...
Learn how: gist.github.com/walkerke/a4b...
The interactive explorer in R's {geosam} package gets you started with @Meta's Segment Anything Model 3 for object detection from satellite imagery.
Shown: sand trap detection on a golf course from text, exemplar, and point prompts.
Try out {geosam}: walker-data.com/geos...
Shown: sand trap detection on a golf course from text, exemplar, and point prompts.
Try out {geosam}: walker-data.com/geos...
January 5, 2026 at 2:57 PM
The interactive explorer in R's {geosam} package gets you started with @Meta's Segment Anything Model 3 for object detection from satellite imagery.
Shown: sand trap detection on a golf course from text, exemplar, and point prompts.
Try out {geosam}: walker-data.com/geos...
Shown: sand trap detection on a golf course from text, exemplar, and point prompts.
Try out {geosam}: walker-data.com/geos...
The world's time zones, visualized on an interactive, spinnable globe.
Try it out- it'll default to your current time and (approximate) time zone: gist.github.com/walkerke/c4a...
Try it out- it'll default to your current time and (approximate) time zone: gist.github.com/walkerke/c4a...
January 4, 2026 at 4:29 PM
The world's time zones, visualized on an interactive, spinnable globe.
Try it out- it'll default to your current time and (approximate) time zone: gist.github.com/walkerke/c4a...
Try it out- it'll default to your current time and (approximate) time zone: gist.github.com/walkerke/c4a...
January 3, 2026 at 3:58 PM
Spatial optimization has applications in nearly every field but it's underutilized.
In 2026, I'm working on open source and lightning-fast tools to make spatial optimization accessible.
Shown: interactively solving a "location-allocation" problem in an app with instant results.
In 2026, I'm working on open source and lightning-fast tools to make spatial optimization accessible.
Shown: interactively solving a "location-allocation" problem in an app with instant results.
January 2, 2026 at 6:30 PM
Spatial optimization has applications in nearly every field but it's underutilized.
In 2026, I'm working on open source and lightning-fast tools to make spatial optimization accessible.
Shown: interactively solving a "location-allocation" problem in an app with instant results.
In 2026, I'm working on open source and lightning-fast tools to make spatial optimization accessible.
Shown: interactively solving a "location-allocation" problem in an app with instant results.
2025 was a massive year for me as an R developer.
This past year I've branched out and focused on bringing interfaces to some of the most exciting geospatial technologies to the R community.
Here's a summary:
This past year I've branched out and focused on bringing interfaces to some of the most exciting geospatial technologies to the R community.
Here's a summary:
December 31, 2025 at 5:30 PM
2025 was a massive year for me as an R developer.
This past year I've branched out and focused on bringing interfaces to some of the most exciting geospatial technologies to the R community.
Here's a summary:
This past year I've branched out and focused on bringing interfaces to some of the most exciting geospatial technologies to the R community.
Here's a summary:
Spatial data science in 2026:
- Pull demographic data for all ~7k Census tracts in Texas;
- Create a territory design solution with spatial and demographic constraints;
- Explore the result on a WebGL map;
All in seconds.
Technology is no longer the bottleneck.
- Pull demographic data for all ~7k Census tracts in Texas;
- Create a territory design solution with spatial and demographic constraints;
- Explore the result on a WebGL map;
All in seconds.
Technology is no longer the bottleneck.
December 29, 2025 at 9:48 PM
Spatial data science in 2026:
- Pull demographic data for all ~7k Census tracts in Texas;
- Create a territory design solution with spatial and demographic constraints;
- Explore the result on a WebGL map;
All in seconds.
Technology is no longer the bottleneck.
- Pull demographic data for all ~7k Census tracts in Texas;
- Create a territory design solution with spatial and demographic constraints;
- Explore the result on a WebGL map;
All in seconds.
Technology is no longer the bottleneck.
My favorite #Python package to use is spopt, a library for spatial optimization.
It helps you with:
📊 Facility location planning;
📊 Sales territory design;
📊 Maximizing market share;
And much more! Check it out here:
pysal.org/spopt/
It helps you with:
📊 Facility location planning;
📊 Sales territory design;
📊 Maximizing market share;
And much more! Check it out here:
pysal.org/spopt/
December 27, 2025 at 2:56 PM
My favorite #Python package to use is spopt, a library for spatial optimization.
It helps you with:
📊 Facility location planning;
📊 Sales territory design;
📊 Maximizing market share;
And much more! Check it out here:
pysal.org/spopt/
It helps you with:
📊 Facility location planning;
📊 Sales territory design;
📊 Maximizing market share;
And much more! Check it out here:
pysal.org/spopt/
It’s been a week since I released the geosam package bringing Meta’s SAM3 image segmentation model to R.
This week, users have:
- Segmented trees in Belgium;
- Found paths in rural areas;
- Identified boats in harbors;
All in R!
Try out geosam: walker-data.com/geosam
This week, users have:
- Segmented trees in Belgium;
- Found paths in rural areas;
- Identified boats in harbors;
All in R!
Try out geosam: walker-data.com/geosam
December 24, 2025 at 4:28 PM
It’s been a week since I released the geosam package bringing Meta’s SAM3 image segmentation model to R.
This week, users have:
- Segmented trees in Belgium;
- Found paths in rural areas;
- Identified boats in harbors;
All in R!
Try out geosam: walker-data.com/geosam
This week, users have:
- Segmented trees in Belgium;
- Found paths in rural areas;
- Identified boats in harbors;
All in R!
Try out geosam: walker-data.com/geosam
While some have made the claim that "R is dead!" this year on social media...
Claude Code is going mainstream, and guess what language it's using.
#LongLiveR
#rstats
Claude Code is going mainstream, and guess what language it's using.
#LongLiveR
#rstats
December 24, 2025 at 2:31 PM
While some have made the claim that "R is dead!" this year on social media...
Claude Code is going mainstream, and guess what language it's using.
#LongLiveR
#rstats
Claude Code is going mainstream, and guess what language it's using.
#LongLiveR
#rstats
Interactive energy infrastructure analytics powered by the {geosam} R package.
Find solar farms on a satellite image in West Texas, click points to segment the imagery & extract shapes.
Save the result then get back an sf object in a single step with area calculated for you.
Find solar farms on a satellite image in West Texas, click points to segment the imagery & extract shapes.
Save the result then get back an sf object in a single step with area calculated for you.
December 22, 2025 at 9:14 PM
Interactive energy infrastructure analytics powered by the {geosam} R package.
Find solar farms on a satellite image in West Texas, click points to segment the imagery & extract shapes.
Save the result then get back an sf object in a single step with area calculated for you.
Find solar farms on a satellite image in West Texas, click points to segment the imagery & extract shapes.
Save the result then get back an sf object in a single step with area calculated for you.
I know we used to make fun of Google's AI overviews - but it's pretty impressive how much it's improved in a short period of time.
I released this R package Wednesday.
I released this R package Wednesday.
December 21, 2025 at 3:11 PM
I know we used to make fun of Google's AI overviews - but it's pretty impressive how much it's improved in a short period of time.
I released this R package Wednesday.
I released this R package Wednesday.
Detecting the letter "O" at @uoregon.bsky.social's athletic fields using Meta's Segment Anything Model 3, Mapbox imagery, and the {geosam} #rstats package.
SAM3 does quite well - all you need is the text prompt "the letter o"!
#GoDucks
SAM3 does quite well - all you need is the text prompt "the letter o"!
#GoDucks
December 20, 2025 at 9:06 PM
Detecting the letter "O" at @uoregon.bsky.social's athletic fields using Meta's Segment Anything Model 3, Mapbox imagery, and the {geosam} #rstats package.
SAM3 does quite well - all you need is the text prompt "the letter o"!
#GoDucks
SAM3 does quite well - all you need is the text prompt "the letter o"!
#GoDucks
Andrej Karpathy‘s comments on agency matter more than ever in a post-Opus 4.5 world
It’s not so much a matter of can you or can’t you anymore
It’s now a question of will you or won’t you
Agency in today’s labor market is the most important skill you can cultivate
It’s not so much a matter of can you or can’t you anymore
It’s now a question of will you or won’t you
Agency in today’s labor market is the most important skill you can cultivate
December 20, 2025 at 3:52 PM
Andrej Karpathy‘s comments on agency matter more than ever in a post-Opus 4.5 world
It’s not so much a matter of can you or can’t you anymore
It’s now a question of will you or won’t you
Agency in today’s labor market is the most important skill you can cultivate
It’s not so much a matter of can you or can’t you anymore
It’s now a question of will you or won’t you
Agency in today’s labor market is the most important skill you can cultivate
West Texas energy infrastructure identified from satellite imagery using zero-shot image segmentation.
Shown:
Well pads from simple text prompts;
Draw a box around a wind turbine, find the others in the view;
Click a solar farm, segment the area.
Shown:
Well pads from simple text prompts;
Draw a box around a wind turbine, find the others in the view;
Click a solar farm, segment the area.
December 19, 2025 at 9:33 PM
West Texas energy infrastructure identified from satellite imagery using zero-shot image segmentation.
Shown:
Well pads from simple text prompts;
Draw a box around a wind turbine, find the others in the view;
Click a solar farm, segment the area.
Shown:
Well pads from simple text prompts;
Draw a box around a wind turbine, find the others in the view;
Click a solar farm, segment the area.
Introducing {geosam}: an R toolkit for Meta's Segment Anything Model 3
- Detect objects in satellite imagery with text prompts
- Interactive Shiny tools with draw & click-to-identify
- Works with Mapbox, Esri, MapTiler
- Use regular images too!
walker-data.com/geosam/
- Detect objects in satellite imagery with text prompts
- Interactive Shiny tools with draw & click-to-identify
- Works with Mapbox, Esri, MapTiler
- Use regular images too!
walker-data.com/geosam/
December 17, 2025 at 5:23 PM
Introducing {geosam}: an R toolkit for Meta's Segment Anything Model 3
- Detect objects in satellite imagery with text prompts
- Interactive Shiny tools with draw & click-to-identify
- Works with Mapbox, Esri, MapTiler
- Use regular images too!
walker-data.com/geosam/
- Detect objects in satellite imagery with text prompts
- Interactive Shiny tools with draw & click-to-identify
- Works with Mapbox, Esri, MapTiler
- Use regular images too!
walker-data.com/geosam/
The "explore" tool in the new {geosam} R package puts Meta's powerful SAM3 directly in the hands of #rstats users.
Shown here: we supply the model six prompts to find in an image; it finds all six within an interactive viewer with pan/zoom.
Release coming soon!
Shown here: we supply the model six prompts to find in an image; it finds all six within an interactive viewer with pan/zoom.
Release coming soon!
December 17, 2025 at 3:54 AM
The "explore" tool in the new {geosam} R package puts Meta's powerful SAM3 directly in the hands of #rstats users.
Shown here: we supply the model six prompts to find in an image; it finds all six within an interactive viewer with pan/zoom.
Release coming soon!
Shown here: we supply the model six prompts to find in an image; it finds all six within an interactive viewer with pan/zoom.
Release coming soon!