In geocompr/geocompkg: Geocomputation with R Metapackage
class: large
What is this talk about?
This talk will introduce Geocomputation with R, a new book on using R to work with geographic data.
It covers some of the key contents from the book and provide some examples of what is possible when geographic datasets are approached from a 'data science' perspective, that ensures reproducibility.
I'll show how working with geographic data has become much easier and more user friendly over the years, and demonstrate the importance of geographic data for informing policy.
--
hamb_geo_talk = list(
intro = "about me + geo* data",
ecosystem = "prominent spatial packages",
live_demo = "actions speak louder than words",
evidence4policy = "ensuring impact"
)
str(hamb_geo_talk)
intro
'Team geocompr'
options(htmltools.dir.version = FALSE)
library(RefManageR)
BibOptions(check.entries = FALSE,
bib.style = "authoryear",
cite.style = 'alphabetic',
style = "markdown",
first.inits = FALSE,
hyperlink = FALSE,
dashed = FALSE)
my_bib = ReadBib("refs-geostat.bib", check = FALSE)
-
Jakub Nowosad: developer of GeoPAT + more. Lecturing 09:00 on Wednesday.
-
Jannes Muenchow, creator of RQGIS. Lecturing Weds 13:30 on GIS Bridges and Weds 15:30 on Spatial CV.
--
- Robin Lovelace, creator of stplanr, co-author of Efficent R Programming. Lecturing 11:00 Weds on spatial data and the tidyverse.
--
- Still looking for input before publication in early 2019.
whoami
system("whoami")
--
.pull-left[
- Environmental geographer
-
Learned R for PhD on energy and transport
-
Now work at the University of Leeds (ITS and LIDA)
-
Working on Geocomputation with R
devtools::install_github("r-rust/gifski")
system("youtube-dl https://youtu.be/CzxeJlgePV4 -o v.mp4")
system("ffmpeg -i v.mp4 -t 00:00:03 -c copy out.mp4")
system("ffmpeg -i out.mp4 frame%04d.png ")
f = list.files(pattern = "frame")
gifski::gifski(f, gif_file = "g.gif", width = 200, height = 200)
]
--
.pull-right[
Image credit: Jeroen Ooms + others
knitr::include_graphics("https://user-images.githubusercontent.com/1825120/39661313-534efd66-5047-11e8-8d99-a5597fe160ff.gif")
]
.pull-left[
Why geo*?
Geographic data is everywhere
underlies some of society's biggest issues
global analyses given local meaning
]
--
.pull-right[
knitr::include_graphics("https://raw.githubusercontent.com/npct/pct-team/master/figures/Leeds-network.png")
Example: Propensity to Cycle Tool (PCT) in Manchester: http://pct.bike/m/?r=greater-manchester
]
R can help tackle Big issues
--
knitr::include_graphics("https://media3.giphy.com/media/YkXNjAkG7CfEVx3gcy/giphy.gif?cid=3640f6095bd86a5e784739746ba086e4")
--
Geocomputation with R provides new tools
background-image: url("https://media1.giphy.com/media/Hw5LkPYy9yfVS/giphy.gif")
Geocomputation with R
- A book we are working on for CRC Press (to be published early 2019)
- Chapters 3 ~~and 4~~ of this book demonstrate foundations of geocomputing
What is Geocomputation?
.pull-left[
GeoComputation is about using the various different types of geodata and about developing relevant geo-tools within the overall context of a 'scientific' approach r Citep(my_bib, "openshaw_geocomputation_2000", .opts = list(cite.style = "authoryear")).
]
.pull-right[
knitr::include_graphics("http://www.ccg.leeds.ac.uk/people/s.openshaw/s.openshaw.png")
]
--
- But we do differ from early definitions in one important way:
At the turn of the 21st Century it was unrealistic to expect readers to be able to reproduce code examples, due to barriers preventing access to the necessary hardware, software and data r Citep(my_bib, "lovelace_geocomputation_2019", .opts = list(cite.style = "authoryear"))
Other aspects of the definition
--
What distinguishes geocomputation from the older quantitative geography, is its emphasis on "creative and experimental" GIS applications r Citep(my_bib, "longley_geocomputation_1998", .opts = list(cite.style = "authoryear")).
--
It's about doing "practical work that is beneficial or useful" r Citep(my_bib, "openshaw_geocomputation_2000", .opts = list(cite.style = "authoryear")).
--
about harnessing the power of modern computers to do things with geographic data.
What's in the geocompr box?
.pull-left[
- Chapter 1: History + 'philosophy' = important
Foundations
- Starting from nothing
- Class definitions
- Spatial/attribute operations
- Projections
- Data IO
Extensions
- Advanced methods
- How to build your own functions
Applications
- A taster of what you can do
]
.pull-right[
knitr::include_graphics("https://raw.githubusercontent.com/Robinlovelace/geocompr/master/images/frontcover.png")
]
class: inverse, center, middle
ecosystem
Context
- Software for 'data science' is evolving
- In R, packages ggplot2 and dplyr have become immensely popular and now they are a part of the tidyverse
- These packages use the 'tidy data' principles for consistency and speed of processing (from
vignette("tidy-data")):
- Each variable forms a column.
- Each observation forms a row.
- Each type of observational unit forms a table
- Historically spatial R packages have not been compatible with the tidyverse
background-image: url("https://pbs.twimg.com/media/CvzEQcfWIAAIs-N.jpg")
background-size: cover
Enter sf
- sf is a recently developed package for spatial (vector) data
- Combines the functionality of three previous packages: sp, rgeos and rgdal
- Has many advantages, including:
- Faster data I/O
- More geometry types supported
- Compatibility with the tidyverse
A brief history of geographic data in R
R's predecesor was S, which was itself inspired by lisp r Citep(my_bib, "chambers_extending_2016", .opts = list(cite.style = "authoryear")).
This is geographic analysis in S r Citep(my_bib, "rowlingson_splancs_1993", .opts = list(cite.style = "authoryear")):
pts <- spoints(scan('cavities'))
uk()
pointmap(pts,add=T)
zoom()
uk(add=T)
pointmap(pts,add=T)
poly<-getpoly()
Still works today, 25 years later:
library(splancs)
#> Spatial Point Pattern Analysis Code in S-Plus
#> Version 2 - Spatial and Space-Time analysis
Live demo (try this)!
# install.packages("splancs"); library(splancs)
# example, interactive! (commented bits)
data(bodmin)
plot(bodmin$poly, asp=1, type="n")
pointmap(as.points(bodmin), add=TRUE)
# zoom()
# pointmap(as.points(bodmin), add=TRUE)
Observations
-
R' is robust and future-proof
-
See a video of Roger Bivand's talk on the subject (live demo of R 0.49) + GitHub repo
-
Rs capabilities have evolved substantially since then!
A brief history of geographic data viz in R
"The core R engine was not designed specifically for the display and analysis
of maps, and the limited interactive facilities it offers have drawbacks in this
area" r Citep(my_bib, "bivand_applied_2013", .opts = list(cite.style = "authoryear")).
--
Five years later...
--
"An example showing R's flexibility and evolving geographic capabilities is leaflet
r Citep(my_bib, "R-leaflet", .opts = list(cite.style = "authoryear")),
a package for making interactive maps that has been extended by the R community, as we'll see in Chapter 9"
r Citep(my_bib, "lovelace_geocomputation_2018", .opts = list(cite.style = "authoryear")).
Base R graphics: sf
library(sf)
library(spData)
plot(nz)
--
tmap
- A diverse dedicated mapping R package
library(tmap)
tm_shape(nz) +
tm_polygons("Median_income", palette = "RdYlBu")
Summary: recent packages for spatial data
install.packages("tidyverse")
The tidyverse now (mostly) works with spatial data.
This is thanks to sf, a recent package (first release in 2016) that implements the open standard data model simple features. Get sf with:
install.packages("sf")
Raster data is also supported, in the more mature package raster:
install.packages("raster")
For datasets...:
install.packages("spData")
install.packages("rnaturalearth")
For more on this see: github.com/Robinlovelace/geocompr.
osmdata
A package by Mark Padgham and others:
Padgham, M., Lovelace, R., Salmon, M., Rudis, B., 2017. osmdata. The Journal of Open Source Software 2. https://doi.org/10.21105/joss.00305
library(osmdata)
hamburg = getbb("hamburg")
parks = opq(bbox = hamburg) %>%
add_osm_feature(key = "leisure", value = "park") %>%
osmdata_sf()
parks
--
A map can tell 1000 words
library(tmap)
ttm()
qtm(parks$osm_polygons)
class: inverse, center, middle
Demo: sf in the tidyverse
Reproducibility + collaboration
Collaboration is most important aspect of science (and reprex the most important R package!) (Jakub Nowosad, 2018, GEOSTAT)
Slides: https://geocompr.github.io/presentations/
Source code: https://github.com/geocompr/geostats_18
reprex::reprex(2 + 2)
2 + 2
#> [1] 4
Created on 2018-10-30 by the reprex package (v0.2.1)
Attaching packages
library(sf)
library(raster)
Issue: Conflicting function names
library(tidyverse)
Reading and writing spatial data
library(sf)
f = system.file(package = "spData", "shapes/world.shp")
world_sf = read_sf(f)
write_sf()/st_write() writes data st_write(prague_sf, 'data/prague_sf.gpkg'). See supported formats with: sf::st_drivers(). Details: Chapter 6 of our book: geocompr.robinlovelace.net/read-write.html
Structure of the sf objects
library(spData) # we'll use our own data
class(world)
world
# ---
## Structure of the sf objects
# world$name_long
# ```
#
# ```r
# world$name_long[1:3]
# ```
#
# ```r
# world$geom
# ```
#
# ```r
# print(world$geom, n = 3)
Attribute operations: filtering
world %>%
filter(name_long == "United Kingdom")
--
Base R equivalent:
world[world$name_long == "United Kingdom", ]
--
Question:
identical(
world %>% filter(name_long == "United Kingdom"),
world[world$name_long == "United Kingdom", ]
) # ?
Aggregation
world_cont = world %>%
group_by(continent) %>%
summarize(pop_sum = sum(pop, na.rm = TRUE))
print(world_cont, n = 1)
- The
st_set_geometry function can be used to remove the geometry column:
world_df = st_set_geometry(world_cont, NULL)
class(world_df)
Spatial operations
It's a big topic which includes:
- Spatial subsetting
- Spatial joining/aggregation
- Topological relations
- Distances
- Spatial geometry modification
- Raster operations (map algebra)
See Chapter 4 of Geocomputation with R
Spatial subsetting
lnd_buff = lnd[1, ] %>%
st_transform(crs = 27700) %>% # uk CRS
st_buffer(500000) %>%
st_transform(crs = 4326)
near_lnd = world[lnd_buff, ]
plot(near_lnd$geom)
- What is going with the country miles away?
Multi-objects
Some objects have multiple geometries:
st_geometry_type(near_lnd)
Which have more than 1?
data.frame(near_lnd$name_long,
sapply(near_lnd$geom, length))
Subsetting contiguous polygons
near_lnd_new = world %>%
st_cast(to = "POLYGON") %>%
filter(st_intersects(., lnd_buff, sparse = FALSE))
plot(st_geometry(near_lnd_new))
CRS: a key geo* concept
na_2163 = world %>%
filter(continent == "North America") %>%
st_transform(2163) #US National Atlas Equal Area
st_crs(na_2163)
na = world %>%
filter(continent == "North America")
png('slides/figs/coord_compare.png', width = 1000, height = 250)
par(mfrow = c(1, 2), mar=c(0,0,0,0))
plot(na[0]);plot(na_2163[0])
dev.off()
Making maps
- Basic maps of
sf objects can be quickly created using the plot() function:
plot(world[0])
plot(world["pop"])
png('slides/figs/plot_compare.png', width = 800, height = 300)
par(mfrow = c(1, 2), mar=c(0,0,1,0))
plot(world[0]);plot(world["pop"])
dev.off()
tmap
https://cran.r-project.org/web/packages/tmap/vignettes/tmap-nutshell.html
library(tmap)
tmap_mode("plot") #check the "view"
tm_shape(world, projection = "+proj=moll") +
tm_polygons("lifeExp", title = "Life expactancy",
style = "pretty", palette = "RdYlGn") +
tm_style("grey")
leaflet
library(leaflet)
leaflet(world) %>%
addTiles() %>%
addPolygons(color = "#444444", weight = 1, fillOpacity = 0.5,
fillColor = ~colorQuantile("YlOrRd", lifeExp)(lifeExp),
popup = paste(round(world$lifeExp, 2)))
library(widgetframe)
library('leaflet')
l = leaflet(world) %>%
addTiles() %>%
addPolygons(color = "#444444", weight = 1, fillOpacity = 0.5, fillColor = ~colorQuantile("YlOrRd", lifeExp)(lifeExp), popup = paste(round(world$lifeExp, 2)))
frameWidget(l, height = '400')
What about raster data?
Raster data in R is evolving:
- You can convert vector data to the
Spatial* form using as(my_vector, "Spatial")before working on raster-vector interactions
- There are some initial efforts to bring raster data closer to the tidyverse/sf approaches, such as tabularaster, sfraster, or fasterize
- stars focusses on multidimensional space-time (raster/vector) is evolving.
- terra - first release expected early 2019
class: inverse, center, middle
Applications: R for sustainable transport research
background-image: url(https://pbs.twimg.com/media/DOH94nXUIAAgcll.jpg)
background-position: 50% 50%
class: center, bottom, inverse
Case study: congested cities
Transport: growing source of emissions
.pull-left[
knitr::include_graphics("https://raw.githubusercontent.com/Robinlovelace/erum18-transport/master/transport-projections-ipcc.png")
]
--
.pull-right[
-
People like to travel!
-
Does not 'saturate' with income
-
Hard to decarbonise via technology
]
'Leverage points' where R skills can help
A few examples:
--
- Visualising data: e.g geoplumber
--
- To the public - web viz pkgs help with that!
--
- Local issues. E.g. where's the most dangerous/unpopular road in Hamburg?
--
- Community cohesion
--
Challenge: balance between innovation and tool overload
Challenge: balance between transparency/simplicity and sophistication of analysis
That is a balance R is ideally set-up to strike
class: center, middle
Thanks, links, happy R day travels 🚶, 🚲 + 🚀!
-
Reproducible slides + app: github.com/Robinlovelace/erum18-transport
-
Transport chapter in Geocomputation with R (feedback welcome):
geocompr.robinlovelace.net
Slides created via the R package xaringan.
geocompr/geocompkg documentation built on July 5, 2025, 2:35 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
class: large
What is this talk about?
This talk will introduce Geocomputation with R, a new book on using R to work with geographic data. It covers some of the key contents from the book and provide some examples of what is possible when geographic datasets are approached from a 'data science' perspective, that ensures reproducibility. I'll show how working with geographic data has become much easier and more user friendly over the years, and demonstrate the importance of geographic data for informing policy.
--
hamb_geo_talk = list( intro = "about me + geo* data", ecosystem = "prominent spatial packages", live_demo = "actions speak louder than words", evidence4policy = "ensuring impact" )
str(hamb_geo_talk)
intro
'Team geocompr'
options(htmltools.dir.version = FALSE) library(RefManageR) BibOptions(check.entries = FALSE, bib.style = "authoryear", cite.style = 'alphabetic', style = "markdown", first.inits = FALSE, hyperlink = FALSE, dashed = FALSE) my_bib = ReadBib("refs-geostat.bib", check = FALSE)
-
Jakub Nowosad: developer of GeoPAT + more. Lecturing 09:00 on Wednesday.
-
Jannes Muenchow, creator of RQGIS. Lecturing Weds 13:30 on GIS Bridges and Weds 15:30 on Spatial CV.
--
- Robin Lovelace, creator of stplanr, co-author of Efficent R Programming. Lecturing 11:00 Weds on spatial data and the tidyverse.
--
- Still looking for input before publication in early 2019.
whoami
system("whoami")
--
.pull-left[ - Environmental geographer
-
Learned R for PhD on energy and transport
-
Now work at the University of Leeds (ITS and LIDA)
-
Working on Geocomputation with R
devtools::install_github("r-rust/gifski") system("youtube-dl https://youtu.be/CzxeJlgePV4 -o v.mp4") system("ffmpeg -i v.mp4 -t 00:00:03 -c copy out.mp4") system("ffmpeg -i out.mp4 frame%04d.png ") f = list.files(pattern = "frame") gifski::gifski(f, gif_file = "g.gif", width = 200, height = 200)
]
--
.pull-right[ Image credit: Jeroen Ooms + others
knitr::include_graphics("https://user-images.githubusercontent.com/1825120/39661313-534efd66-5047-11e8-8d99-a5597fe160ff.gif")
]
.pull-left[
Why geo*?
Geographic data is everywhere
underlies some of society's biggest issues
global analyses given local meaning
]
--
.pull-right[
knitr::include_graphics("https://raw.githubusercontent.com/npct/pct-team/master/figures/Leeds-network.png")
Example: Propensity to Cycle Tool (PCT) in Manchester: http://pct.bike/m/?r=greater-manchester
]
R can help tackle Big issues
--
knitr::include_graphics("https://media3.giphy.com/media/YkXNjAkG7CfEVx3gcy/giphy.gif?cid=3640f6095bd86a5e784739746ba086e4")
--
Geocomputation with R provides new tools
background-image: url("https://media1.giphy.com/media/Hw5LkPYy9yfVS/giphy.gif")
Geocomputation with R
- A book we are working on for CRC Press (to be published early 2019)
- Chapters 3 ~~and 4~~ of this book demonstrate foundations of geocomputing
What is Geocomputation?
.pull-left[
GeoComputation is about using the various different types of geodata and about developing relevant geo-tools within the overall context of a 'scientific' approach
r Citep(my_bib, "openshaw_geocomputation_2000", .opts = list(cite.style = "authoryear")).
]
.pull-right[
knitr::include_graphics("http://www.ccg.leeds.ac.uk/people/s.openshaw/s.openshaw.png")
]
--
- But we do differ from early definitions in one important way:
At the turn of the 21st Century it was unrealistic to expect readers to be able to reproduce code examples, due to barriers preventing access to the necessary hardware, software and data
r Citep(my_bib, "lovelace_geocomputation_2019", .opts = list(cite.style = "authoryear"))
Other aspects of the definition
--
What distinguishes geocomputation from the older quantitative geography, is its emphasis on "creative and experimental" GIS applications
r Citep(my_bib, "longley_geocomputation_1998", .opts = list(cite.style = "authoryear")).
--
It's about doing "practical work that is beneficial or useful" r Citep(my_bib, "openshaw_geocomputation_2000", .opts = list(cite.style = "authoryear")).
--
about harnessing the power of modern computers to do things with geographic data.
What's in the geocompr box?
.pull-left[
- Chapter 1: History + 'philosophy' = important
Foundations
- Starting from nothing
- Class definitions
- Spatial/attribute operations
- Projections
- Data IO
Extensions
- Advanced methods
- How to build your own functions
Applications
- A taster of what you can do
]
.pull-right[
knitr::include_graphics("https://raw.githubusercontent.com/Robinlovelace/geocompr/master/images/frontcover.png")
]
class: inverse, center, middle
ecosystem
Context
- Software for 'data science' is evolving
- In R, packages ggplot2 and dplyr have become immensely popular and now they are a part of the tidyverse
- These packages use the 'tidy data' principles for consistency and speed of processing (from
vignette("tidy-data")):
- Each variable forms a column.
- Each observation forms a row.
- Each type of observational unit forms a table
- Historically spatial R packages have not been compatible with the tidyverse
background-image: url("https://pbs.twimg.com/media/CvzEQcfWIAAIs-N.jpg") background-size: cover
Enter sf
- sf is a recently developed package for spatial (vector) data
- Combines the functionality of three previous packages: sp, rgeos and rgdal
- Has many advantages, including:
- Faster data I/O
- More geometry types supported
- Compatibility with the tidyverse
A brief history of geographic data in R
R's predecesor was S, which was itself inspired by lisp r Citep(my_bib, "chambers_extending_2016", .opts = list(cite.style = "authoryear")).
This is geographic analysis in S r Citep(my_bib, "rowlingson_splancs_1993", .opts = list(cite.style = "authoryear")):
pts <- spoints(scan('cavities')) uk() pointmap(pts,add=T) zoom() uk(add=T) pointmap(pts,add=T) poly<-getpoly()
Still works today, 25 years later:
library(splancs) #> Spatial Point Pattern Analysis Code in S-Plus #> Version 2 - Spatial and Space-Time analysis
Live demo (try this)!
# install.packages("splancs"); library(splancs) # example, interactive! (commented bits) data(bodmin) plot(bodmin$poly, asp=1, type="n") pointmap(as.points(bodmin), add=TRUE) # zoom() # pointmap(as.points(bodmin), add=TRUE)
Observations
-
R' is robust and future-proof
-
See a video of Roger Bivand's talk on the subject (live demo of R 0.49) + GitHub repo
-
Rs capabilities have evolved substantially since then!
A brief history of geographic data viz in R
"The core R engine was not designed specifically for the display and analysis
of maps, and the limited interactive facilities it offers have drawbacks in this
area" r Citep(my_bib, "bivand_applied_2013", .opts = list(cite.style = "authoryear")).
--
Five years later...
--
"An example showing R's flexibility and evolving geographic capabilities is leaflet
r Citep(my_bib, "R-leaflet", .opts = list(cite.style = "authoryear")),
a package for making interactive maps that has been extended by the R community, as we'll see in Chapter 9"
r Citep(my_bib, "lovelace_geocomputation_2018", .opts = list(cite.style = "authoryear")).
Base R graphics: sf
library(sf) library(spData) plot(nz)
--
tmap
- A diverse dedicated mapping R package
library(tmap) tm_shape(nz) + tm_polygons("Median_income", palette = "RdYlBu")
Summary: recent packages for spatial data
install.packages("tidyverse")
The tidyverse now (mostly) works with spatial data.
This is thanks to sf, a recent package (first release in 2016) that implements the open standard data model simple features. Get sf with:
install.packages("sf")
Raster data is also supported, in the more mature package raster:
install.packages("raster")
For datasets...:
install.packages("spData") install.packages("rnaturalearth")
For more on this see: github.com/Robinlovelace/geocompr.
osmdata
A package by Mark Padgham and others:
Padgham, M., Lovelace, R., Salmon, M., Rudis, B., 2017. osmdata. The Journal of Open Source Software 2. https://doi.org/10.21105/joss.00305
library(osmdata) hamburg = getbb("hamburg") parks = opq(bbox = hamburg) %>% add_osm_feature(key = "leisure", value = "park") %>% osmdata_sf() parks
--
A map can tell 1000 words
library(tmap) ttm() qtm(parks$osm_polygons)
class: inverse, center, middle
Demo: sf in the tidyverse
Reproducibility + collaboration
Collaboration is most important aspect of science (and reprex the most important R package!) (Jakub Nowosad, 2018, GEOSTAT)
Slides: https://geocompr.github.io/presentations/
Source code: https://github.com/geocompr/geostats_18
reprex::reprex(2 + 2)
2 + 2 #> [1] 4
Created on 2018-10-30 by the reprex package (v0.2.1)
Attaching packages
library(sf) library(raster)
Issue: Conflicting function names
library(tidyverse)
Reading and writing spatial data
library(sf) f = system.file(package = "spData", "shapes/world.shp") world_sf = read_sf(f)
write_sf()/st_write() writes data st_write(prague_sf, 'data/prague_sf.gpkg'). See supported formats with: sf::st_drivers(). Details: Chapter 6 of our book: geocompr.robinlovelace.net/read-write.html
Structure of the sf objects
library(spData) # we'll use our own data class(world) world
# --- ## Structure of the sf objects # world$name_long # ``` # # ```r # world$name_long[1:3] # ``` # # ```r # world$geom # ``` # # ```r # print(world$geom, n = 3)
Attribute operations: filtering
world %>% filter(name_long == "United Kingdom")
--
Base R equivalent:
world[world$name_long == "United Kingdom", ]
--
Question:
identical( world %>% filter(name_long == "United Kingdom"), world[world$name_long == "United Kingdom", ] ) # ?
Aggregation
world_cont = world %>% group_by(continent) %>% summarize(pop_sum = sum(pop, na.rm = TRUE))
print(world_cont, n = 1)
- The
st_set_geometryfunction can be used to remove the geometry column:
world_df = st_set_geometry(world_cont, NULL) class(world_df)
Spatial operations
It's a big topic which includes:
- Spatial subsetting
- Spatial joining/aggregation
- Topological relations
- Distances
- Spatial geometry modification
- Raster operations (map algebra)
See Chapter 4 of Geocomputation with R
Spatial subsetting
lnd_buff = lnd[1, ] %>% st_transform(crs = 27700) %>% # uk CRS st_buffer(500000) %>% st_transform(crs = 4326) near_lnd = world[lnd_buff, ] plot(near_lnd$geom)
- What is going with the country miles away?
Multi-objects
Some objects have multiple geometries:
st_geometry_type(near_lnd)
Which have more than 1?
data.frame(near_lnd$name_long, sapply(near_lnd$geom, length))
Subsetting contiguous polygons
near_lnd_new = world %>% st_cast(to = "POLYGON") %>% filter(st_intersects(., lnd_buff, sparse = FALSE)) plot(st_geometry(near_lnd_new))
CRS: a key geo* concept
na_2163 = world %>% filter(continent == "North America") %>% st_transform(2163) #US National Atlas Equal Area st_crs(na_2163)
na = world %>% filter(continent == "North America") png('slides/figs/coord_compare.png', width = 1000, height = 250) par(mfrow = c(1, 2), mar=c(0,0,0,0)) plot(na[0]);plot(na_2163[0]) dev.off()
Making maps
- Basic maps of
sfobjects can be quickly created using theplot()function:
plot(world[0])
plot(world["pop"])
png('slides/figs/plot_compare.png', width = 800, height = 300) par(mfrow = c(1, 2), mar=c(0,0,1,0)) plot(world[0]);plot(world["pop"]) dev.off()
tmap
https://cran.r-project.org/web/packages/tmap/vignettes/tmap-nutshell.html
library(tmap) tmap_mode("plot") #check the "view" tm_shape(world, projection = "+proj=moll") + tm_polygons("lifeExp", title = "Life expactancy", style = "pretty", palette = "RdYlGn") + tm_style("grey")
leaflet
library(leaflet) leaflet(world) %>% addTiles() %>% addPolygons(color = "#444444", weight = 1, fillOpacity = 0.5, fillColor = ~colorQuantile("YlOrRd", lifeExp)(lifeExp), popup = paste(round(world$lifeExp, 2)))
library(widgetframe) library('leaflet') l = leaflet(world) %>% addTiles() %>% addPolygons(color = "#444444", weight = 1, fillOpacity = 0.5, fillColor = ~colorQuantile("YlOrRd", lifeExp)(lifeExp), popup = paste(round(world$lifeExp, 2))) frameWidget(l, height = '400')
What about raster data?
Raster data in R is evolving:
- You can convert vector data to the
Spatial*form usingas(my_vector, "Spatial")before working on raster-vector interactions - There are some initial efforts to bring raster data closer to the tidyverse/sf approaches, such as tabularaster, sfraster, or fasterize
- stars focusses on multidimensional space-time (raster/vector) is evolving.
- terra - first release expected early 2019
class: inverse, center, middle
Applications: R for sustainable transport research
background-image: url(https://pbs.twimg.com/media/DOH94nXUIAAgcll.jpg) background-position: 50% 50% class: center, bottom, inverse
Case study: congested cities
Transport: growing source of emissions
.pull-left[
knitr::include_graphics("https://raw.githubusercontent.com/Robinlovelace/erum18-transport/master/transport-projections-ipcc.png")
]
--
.pull-right[
-
People like to travel!
-
Does not 'saturate' with income
-
Hard to decarbonise via technology
]
'Leverage points' where R skills can help
A few examples:
--
- Visualising data: e.g geoplumber
--
- To the public - web viz pkgs help with that!
--
- Local issues. E.g. where's the most dangerous/unpopular road in Hamburg?
--
- Community cohesion
--
Challenge: balance between innovation and tool overload
Challenge: balance between transparency/simplicity and sophistication of analysis
That is a balance R is ideally set-up to strike
class: center, middle
Thanks, links, happy R day travels 🚶, 🚲 + 🚀!
-
Reproducible slides + app: github.com/Robinlovelace/erum18-transport
-
Transport chapter in Geocomputation with R (feedback welcome): geocompr.robinlovelace.net
Slides created via the R package xaringan.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.