- This chapter requires the following packages to be installed and attached:
- It also relies on spData, which loads datasets used in the code examples of this chapter:
Attribute data is non-spatial information associated with geographic (geometry) data.
A bus stop provides a simple example: its position would typically be represented by latitude and longitude coordinates (geometry data), in addition to its name.
The Elephant & Castle / New Kent Road stop in London, for example has coordinates of -0.098 degrees longitude and 51.495 degrees latitude which can be represented as
POINT (-0.098 51.495) in the
sfc representation described in Chapter 2.
Attributes such as the name attribute of the POINT feature (to use Simple Features terminology) are the topic of this chapter.
Another example is the elevation value (attribute) for a specific grid cell in raster data. Unlike the vector data model, the raster data model stores the coordinate of the grid cell indirectly, meaning the distinction between attribute and spatial information is less clear. To illustrate the point, think of a pixel in the 3rd row and the 4th column of a raster matrix. Its spatial location is defined by its index in the matrix: move from the origin four cells in the x direction (typically east and right on maps) and three cells in the y direction (typically south and down). The raster’s resolution defines the distance for each x- and y-step which is specified in a header. The header is a vital component of raster datasets which specifies how pixels relate to geographic coordinates (see also Chapter 4).
This teaches how to manipulate geographic objects based on attributes such as the names of bus stops in a vector dataset and elevations of pixels in a raster dataset.
For vector data, this means techniques such as subsetting and aggregation (see Sections 3.2.1 and 3.2.3).
Sections 3.2.4 and 3.2.5 demonstrate how to join data onto simple feature objects using a shared ID and how to create new variables, respectively.
Each of these operations has a spatial equivalent:
[ operator in base R, for example, works equally for subsetting objects based on their attribute and spatial objects; you can also join attributes in two geographic datasets using spatial joins.
This is good news: skills developed in this chapter are cross-transferable.
Chapter 4 extends the methods presented here to the spatial world.
After a deep dive into various types of vector attribute operations in the next section, raster attribute data operations are covered in Section 3.3, which demonstrates how to create raster layers containing continuous and categorical attributes and extracting cell values from one or more layer (raster subsetting). Section 3.3.2 provides an overview of ‘global’ raster operations which can be used to summarize entire raster datasets.
Geographic vector datasets are well supported in R thanks to the
sf class, which extends base R’s
Like data frames,
sf objects have one column per attribute variable (such as ‘name’) and one row per observation or feature (e.g., per bus station).
sf objects differ from basic data frames because they have a
geometry column of class
sfc which can contain a range of geographic entities (single and ‘multi’ point, line, and polygon features) per row.
This was described in Chapter 2, which demonstrated how generic methods such as
summary() work with
sf also provides generics that allow
sf objects to behave like regular data frames, as shown by printing the class’s methods:
methods(class = "sf") # methods for sf objects, first 12 shown
#>  aggregate cbind coerce #>  initialize merge plot #>  print rbind [ #>  [[<- $<- show
Many of these (
[) are for manipulating data frames.
rbind(), for example, is binds rows two data frames together, one ‘on top’ of the other.
$<- creates new columns.
A key feature of
sf objects is that they store spatial and non-spatial data in the same way, as columns in a
The geometry column of
sf objects is typically called
geom but any name can be used.
The following command, for example, creates a geometry column named g:
This enables geometries imported from spatial databases to have a variety of names such as
st_sf(data.frame(n = world$name_long), g = world$geom)
sf objects can also extend the
tidyverse classes for data frames,
Thus sf enables the full power of R’s data analysis capabilities to be unleashed on geographic data, whether you use base R or tidyverse functions for data analysis.
Rdatatable/data.table#2273 for discussion of compatibility between
sf objects and the fast
Before using these capabilities it is worth re-capping how to discover the basic properties of vector data objects.
Let’s start by using base R functions to learn about the
world dataset from the spData package:
world contains ten non-geographic columns (and one geometry list column) with almost 200 rows representing the world’s countries.
st_drop_geometry() keeps only the attributes data of an
sf object, in other words removing its geometry:
world_df = st_drop_geometry(world) class(world_df) #>  "tbl_df" "tbl" "data.frame" ncol(world_df) #>  10
Dropping the geometry column before working with attribute data can be useful; data manipulation processes can run faster when they work only on the attribute data and geometry columns are not always needed.
For most cases, however, it makes sense to keep the geometry column, explaining why the column is ‘sticky’ (it remains after most attribute operations unless specifically dropped).
Non-spatial data operations on
sf objects only change an object’s geometry when appropriate (e.g., by dissolving borders between adjacent polygons following aggregation).
Becoming skilled at geographic attribute data manipulation means becoming skilled at manipulating data frames.
For many applications, the tidyverse package dplyr offers an effective approach for working with data frames.
Tidyverse compatibility is an advantage of sf over its predecessor sp, but there are some pitfalls to avoid (see the supplementary
tidyverse-pitfalls vignette at geocompr.github.io for details).
Base R subsetting methods include the operator
[ and the function
The key dplyr subsetting functions are
slice() for subsetting rows, and
select() for subsetting columns.
Both approaches preserve the spatial components of attribute data in
sf objects, while using the operator
$ or the dplyr function
pull() to return a single attribute column as a vector will lose the attribute data, as we will see.
This section focuses on subsetting
sf data frames; for further details on subsetting vectors and non-geographic data frames we recommend reading section section 2.7 of An Introduction to R (R Core Team, Smith, and Team 2021) and Chapter 4 of Advanced R Programming (Wickham 2019), respectively.
[ operator can subset both rows and columns.
Indices placed inside square brackets placed directly after a data frame object name specify the elements to keep.
object[i, j] means ’return the rows represented by
i and the columns represented by
j typically contain integers or
FALSEs (indices can also be character strings, indicating row or column names).
object[5, 1:3], for example, means ’return data containing the 5th row and columns 1 to 3: the result should be a data frame with only 1 row and 3 columns, and a fourth geometry column if it’s an
j empty returns all rows or columns, so
world[1:5, ] returns the first five rows and all 11 columns.
The examples below demonstrate subsetting with base R.
Guess the number of rows and columns in the
sf data frames returned by each command and check the results on your own computer (see the end of the chapter for more exercises):
world[1:6, ] # subset rows by position world[, 1:3] # subset columns by position world[1:6, 1:3] # subset rows and columns by position world[, c("name_long", "pop")] # columns by name world[, c(T, T, F, F, F, F, F, T, T, F, F)] # by logical indices world[, 888] # an index representing a non-existent column
A demonstration of the utility of using
logical vectors for subsetting is shown in the code chunk below.
This creates a new object,
small_countries, containing nations whose surface area is smaller than 10,000 km2:
i_small = world$area_km2 < 10000 summary(i_small) # a logical vector #> Mode FALSE TRUE #> logical 170 7 small_countries = world[i_small, ]
i_small (short for index representing small countries) is a logical vector that can be used to subset the seven smallest countries in the
world by surface area.
A more concise command, which omits the intermediary object, generates the same result:
small_countries = world[world$area_km2 < 10000, ]
The base R function
subset() provides another way to achieve the same result:
small_countries = subset(world, area_km2 < 10000)
Base R functions are mature, stable and widely used, making them a rock solid choice, especially in contexts where reproducibility and reliability are key.
dplyr functions enable ‘tidy’ workflows which some people (the authors of this book included) find intuitive and productive for interactive data analysis, especially when combined with code editors such as RStudio that enable auto-completion of column names.
Key functions for subsetting data frames (including
sf data frames) with dplyr functions are demonstrated below.
select() selects columns by name or position.
For example, you could select only two columns,
pop, with the following command:
Note: as with the equivalent command in base R (
world[, c("name_long", "pop")]), the sticky
geom column remains.
select() also allows selecting a range of columns with the help of the
# all columns between name_long and pop (inclusive) world2 = dplyr::select(world, name_long:pop)
You can remove specific columns with the
# all columns except subregion and area_km2 (inclusive) world3 = dplyr::select(world, -subregion, -area_km2)
Subset and rename columns at the same time with the
new_name = old_name syntax:
world4 = dplyr::select(world, name_long, population = pop)
It is worth noting that the command above is more concise than base R equivalent, which requires two lines of code:
Most dplyr verbs return a data frame, but you can extract a single column as a vector with
You can get the same result in base R with the list subsetting operators
[[, the three following commands return the same numeric vector:
pull(world, pop) world$pop world[["pop"]]
filter() is dplyr’s equivalent of base R’s
It keeps only rows matching given criteria, e.g., only countries with and area below a certain threshold, or with a high average of life expectancy, as shown in the following examples:
||Not equal to|
||Greater/Less than or equal|
||Logical operators: And, Or, Not|
Key to workflows using dplyr functions is the ‘pipe’ operator
%>% (or since R
4.1.0 the native pipe
|>), which takes its name from the Unix pipe
| (Grolemund and Wickham 2016).
Pipes enable expressive code: the output of a previous function becomes the first argument of the next function, enabling chaining.
This is illustrated below, in which only countries from Asia are filtered from the
world dataset, next the object is subset by columns (
continent) and the first five rows (result not shown).
The above chunk shows how the pipe operator allows commands to be written in a clear order:
the above run from top to bottom (line-by-line) and left to right.
The alternative to
%>% is nested function calls, which is harder to read:
Aggregation involves summarizing data with one or more ‘grouping variables’, typically from columns in the data frame to be aggregated (geographic aggregation is covered in the next chapter).
An example of attribute aggregation is calculating the number of people per continent based on country-level data (one row per country).
world dataset contains the necessary ingredients: the columns
continent, the population and the grouping variable, respectively.
The aim is to find the
sum() of country populations for each continent, resulting in a smaller data frame (aggregation is a form of data reduction and can be a useful early step when working with large datasets).
This can be done with the base R function
aggregate() as follows:
The result is a non-spatial data frame with six rows, one per continent, and two columns reporting the name and population of each continent (see Table 3.2 with results for the top 3 most populous continents).
aggregate() is a generic function which means that it behaves differently depending on its inputs.
sf provides the method
aggregate.sf() which is activated automatically when
x is an
sf object and a
by argument is provided:
world_agg2 object is a spatial object containing 8 features representing the continents of the world (and the open ocean).
group_by() %>% summarize() is the dplyr equivalent of
aggregate(), with the variable name provided in the
group_by() function specifying the grouping variable and information on what is to be summarized passed to the
summarize() function, as shown below:
The approach may seem more complex but it has benefits: flexibility, readability, and control over the new column names. This flexibility is illustrated in the command below, which calculates not only the population but also the area and number of countries in each continent:
In the previous code chunk
area (sqkm) and
n are column names in the result, and
n() were the aggregating functions.
These aggregating functions return
sf objects with rows representing continents and geometries containing the multiple polygons representing each land mass and associated islands (this works thanks to the geometric operation ‘union’, as explained in Section 5.2.7).
Let’s combine what we have learned so far about dplyr functions, by chaining multiple commands to summarize attribute data about countries worldwide by continent.
The following command calculates population density (with
mutate()), arranges continents by the number countries they contain (with
dplyr::arrange()), and keeps only the 3 most populous continents (with
top_n()), the result of which is presented in Table 3.2):
world_agg5 = world %>% st_drop_geometry() %>% # drop the geometry for speed dplyr::select(pop, continent, area_km2) %>% # subset the columns of interest group_by(continent) %>% # group by continent and summarize: summarize(Pop = sum(pop, na.rm = TRUE), Area = sum(area_km2), N = n()) %>% mutate(Density = round(Pop / Area)) %>% # calculate population density top_n(n = 3, wt = Pop) %>% # keep only the top 3 arrange(desc(N)) # arrange in order of n. countries
Combining data from different sources is a common task in data preparation.
Joins do this by combining tables based on a shared ‘key’ variable.
dplyr has multiple join functions including
inner_join() — see
vignette("two-table") for a full list.
These function names follow conventions used in the database language SQL (Grolemund and Wickham 2016, chap. 13); using them to join non-spatial datasets to
sf objects is the focus of this section.
dplyr join functions work the same on data frames and
sf objects, the only important difference being the
geometry list column.
The result of data joins can be either an
The most common type of attribute join on spatial data takes an
sf object as the first argument and adds columns to it from a
data.frame specified as the second argument.
To demonstrate joins, we will combine data on coffee production with the
The coffee data is in a data frame called
coffee_data from the spData package (see
?coffee_data for details).
It has 3 columns:
name_long names major coffee-producing nations and
coffee_production_2017 contain estimated values for coffee production in units of 60-kg bags in each year.
A ‘left join’, which preserves the first dataset, merges
Because the input datasets share a ‘key variable’ (
name_long) the join worked without using the
by argument (see
?left_join for details).
The result is an
sf object identical to the original
world object but with two new variables (with column indices 11 and 12) on coffee production.
This can be plotted as a map, as illustrated in Figure 3.1, generated with the
plot() function below:
For joining to work, a ‘key variable’ must be supplied in both datasets.
By default dplyr uses all variables with matching names.
In this case, both
world objects contained a variable called
name_long, explaining the message
Joining, by = "name_long".
In the majority of cases where variable names are not the same, you have two options:
- Rename the key variable in one of the objects so they match.
- Use the
byargument to specify the joining variables.
The latter approach is demonstrated below on a renamed version of
Note that the name in the original object is kept, meaning that
world_coffee and the new object
world_coffee2 are identical.
Another feature of the result is that it has the same number of rows as the original dataset.
Although there are only 47 rows of data in
coffee_data, all 177 country records are kept intact in
rows in the original dataset with no match are assigned
NA values for the new coffee production variables.
What if we only want to keep countries that have a match in the key variable?
In that case an inner join can be used:
world_coffee_inner = inner_join(world, coffee_data) #> Joining, by = "name_long" nrow(world_coffee_inner) #>  45
Note that the result of
inner_join() has only 45 rows compared with 47 in
What happened to the remaining rows?
We can identify the rows that did not match using the
setdiff() function as follows:
setdiff(coffee_data$name_long, world$name_long) #>  "Congo, Dem. Rep. of" "Others"
The result shows that
Others accounts for one row not present in the
world dataset and that the name of the
Democratic Republic of the Congo accounts for the other:
it has been abbreviated, causing the join to miss it.
The following command uses a string matching (regex) function from the stringr package to confirm what
Congo, Dem. Rep. of should be:
(drc = stringr::str_subset(world$name_long, "Dem*.+Congo")) #>  "Democratic Republic of the Congo"
To fix this issue, we will create a new version of
coffee_data and update the name.
inner_join()ing the updated data frame returns a result with all 46 coffee-producing nations:
coffee_data$name_long[grepl("Congo,", coffee_data$name_long)] = drc world_coffee_match = inner_join(world, coffee_data) #> Joining, by = "name_long" nrow(world_coffee_match) #>  46
It is also possible to join in the other direction: starting with a non-spatial dataset and adding variables from a simple features object.
This is demonstrated below, which starts with the
coffee_data object and adds variables from the original
In contrast with the previous joins, the result is not another simple feature object, but a data frame in the form of a tidyverse tibble:
the output of a join tends to match its first argument:
sfobject. The geometry column can only be used for creating maps and spatial operations if R ‘knows’ it is a spatial object, defined by a spatial package such as sf. Fortunately, non-spatial data frames with a geometry list column (like
coffee_world) can be coerced into an
sfobject as follows:
This section covers the majority of joining use cases. For more information, we recommend Grolemund and Wickham (2016), the join vignette in the geocompkg package that accompanies this book, and documentation of the data.table package.19 Another type of join is a spatial join, covered in the next chapter (Section 4.2.4).
Often, we would like to create a new column based on already existing columns.
For example, we want to calculate population density for each country.
For this we need to divide a population column, here
pop, by an area column, here
area_km2 with unit area in square kilometers.
Using base R, we can type:
world_new = world # do not overwrite our original data world_new$pop_dens = world_new$pop / world_new$area_km2
unite() from the tidyr package (which provides many useful functions for reshaping datasets, including
pivot_longer()) pastes together existing columns.
For example, we want to combine the
region_un columns into a new column named
Additionally, we can define a separator (here: a colon
:) which defines how the values of the input columns should be joined, and if the original columns should be removed (here:
world_unite = world %>% unite("con_reg", continent:region_un, sep = ":", remove = TRUE)
separate() function does the opposite of
unite(): it splits one column into multiple columns using either a regular expression or character positions.
This function also comes from the tidyr package.
The dplyr function
rename() and the base R function
setNames() are useful for renaming columns.
The first replaces an old name with a new one.
The following command, for example, renames the lengthy
name_long column to simply
setNames() changes all column names at once, and requires a character vector with a name matching each column.
This is illustrated below, which outputs the same
world object, but with very short names:
It is important to note that attribute data operations preserve the geometry of the simple features.
As mentioned at the outset of the chapter, it can be useful to remove the geometry.
To do this, you have to explicitly remove it.
Hence, an approach such as
select(world, -geom) will be unsuccessful and you should instead use
In contrast to the vector data model underlying simple features (which represents points, lines and polygons as discrete entities in space), raster data represent continuous surfaces. This section shows how raster objects work by creating them from scratch, building on Section 2.3.2. Because of their unique structure, subsetting and other operations on raster datasets work in a different way, as demonstrated in Section 3.3.1.
The following code recreates the raster dataset used in Section 2.3.4, the result of which is illustrated in Figure 3.2.
This demonstrates how the
rast() function works to create an example raster named
elev (representing elevations).
elev = rast(nrows = 6, ncols = 6, resolution = 0.5, xmin = -1.5, xmax = 1.5, ymin = -1.5, ymax = 1.5, vals = 1:36)
The result is a raster object with 6 rows and 6 columns (specified by the
ncol arguments), and a minimum and maximum spatial extent in x and y direction (
vals argument sets the values that each cell contains: numeric data ranging from 1 to 36 in this case.
Raster objects can also contain categorical values of class
factor variables in R.
The following code creates the raster datasets shown in Figure 3.2:
The raster object stores the corresponding look-up table or “Raster Attribute Table” (RAT) as a list of data frames, which can be viewed with
?cats() for more information).
Each element of this list is a layer of the raster.
It is also possible to use the function
levels() for retrieving and adding new or replacing existing factor levels:
?coltab). Importantly, saving a raster object with a color table to a file (e.g., GeoTIFF) will also save the color information.
Raster subsetting is done with the base R operator
[, which accepts a variety of inputs:
- Row-column indexing
- Cell IDs
- Coordinates (see Section 4.3.1)
- Another spatial object (see Section 4.3.1)
Here, we only show the first two options since these can be considered non-spatial operations. If we need a spatial object to subset another or the output is a spatial object, we refer to this as spatial subsetting. Therefore, the latter two options will be shown in the next chapter (see Section 4.3.1).
The first two subsetting options are demonstrated in the commands below —
both return the value of the top left pixel in the raster object
elev (results not shown):
# row 1, column 1 elev[1, 1] # cell ID 1 elev
Subsetting of multi-layered raster objects will return the cell value(s) for each layer.
two_layers = c(grain, elev); two_layers returns a data frame with one row and two columns — one for each layer.
To extract all values or complete rows, you can also use
Cell values can be modified by overwriting existing values in conjunction with a subsetting operation.
The following code chunk, for example, sets the upper left cell of
elev to 0 (results not shown):
elev[1, 1] = 0 elev
Leaving the square brackets empty is a shortcut version of
values() for retrieving all values of a raster.
Multiple cells can also be modified in this way:
elev[1, c(1, 2)] = 0
Replacing values of multilayered rasters can be done with a matrix with as many columns as layers and rows as replaceable cells (results not shown):
terra contains functions for extracting descriptive statistics for entire rasters.
Printing a raster object to the console by typing its name returns minimum and maximum values of a raster.
summary() provides common descriptive statistics – minimum, maximum, quartiles and number of
NAs for continuous rasters and a number of cells of each class for categorical rasters.
Further summary operations such as the standard deviation (see below) or custom summary statistics can be calculated with
global()functions with a multi-layered raster object, they will summarize each layer separately, as can be illustrated by running:
freq() function allows to get the frequency table of categorical values.
Raster value statistics can be visualized in a variety of ways.
Specific functions such as
pairs() work also with raster objects, as demonstrated in the histogram created with the command below (not shown):
Descriptive raster statistics belong to the so-called global raster operations. These and other typical raster processing operations are part of the map algebra scheme, which are covered in the next chapter (Section 4.3.2).
Some function names clash between packages (e.g., a function with the
extract() exist in both terra and
tidyr packages). In addition to not loading packages by
referring to functions verbosely (e.g.,
another way to prevent function names clashes is by unloading the
offending package with
detach(). The following command, for
example, unloads the terra package (this can also be
done in the package tab which resides by default in the
right-bottom pane in RStudio):
detach(“package:terra”, unload = TRUE, force = TRUE). The
force argument makes sure that the package will be detached
even if other packages depend on it. This, however, may lead to a
restricted usability of packages depending on the detached package, and
is therefore not recommended.
For these exercises we will use the
us_states_df datasets from the spData package.
You must have attached the package, and other packages used in the attribute operations chapter (sf, dplyr, terra) with commands such as
library(spData) before attempting these exercises:
us_states is a spatial object (of class
sf), containing geometry and a few attributes (including name, region, area, and population) of states within the contiguous United States.
us_states_df is a data frame (of class
data.frame) containing the name and additional variables (including median income and poverty level, for the years 2010 and 2015) of US states, including Alaska, Hawaii and Puerto Rico.
The data comes from the United States Census Bureau, and is documented in
E1. Create a new object called
us_states_name that contains only the
NAME column from the
us_states object using either base R (
[) or tidyverse (
What is the class of the new object and what makes it geographic?
E2. Select columns from the
us_states object which contain population data.
Obtain the same result using a different command (bonus: try to find three ways of obtaining the same result).
Hint: try to use helper functions, such as
matches from dplyr (see
E3. Find all states with the following characteristics (bonus find and plot them):
- Belong to the Midwest region.
- Belong to the West region, have an area below 250,000 km2and in 2015 a population greater than 5,000,000 residents (hint: you may need to use the function
- Belong to the South region, had an area larger than 150,000 km2 or a total population in 2015 larger than 7,000,000 residents.
E4. What was the total population in 2015 in the
What was the minimum and maximum total population in 2015?
E5. How many states are there in each region?
E6. What was the minimum and maximum total population in 2015 in each region? What was the total population in 2015 in each region?
E7. Add variables from
us_states, and create a new object called
What function did you use and why?
Which variable is the key in both datasets?
What is the class of the new object?
us_states_df has two more rows than
How can you find them? (hint: try to use the
E9. What was the population density in 2015 in each state? What was the population density in 2010 in each state?
E10. How much has population density changed between 2010 and 2015 in each state? Calculate the change in percentages and map them.
us_states_df create a new object called
The new object should have only two variables -
Change the name of the
median_income_15 column to
E13. Calculate the change in the number of residents living below the poverty level between 2010 and 2015 for each state. (Hint: See ?us_states_df for documentation on the poverty level columns.) Bonus: Calculate the change in the percentage of residents living below the poverty level in each state.
E14. What was the minimum, average and maximum state’s number of people living below the poverty line in 2015 for each region? Bonus: What is the region with the largest increase in people living below the poverty line?
E15. Create a raster from scratch with nine rows and columns and a resolution of 0.5 decimal degrees (WGS84). Fill it with random numbers. Extract the values of the four corner cells.
E16. What is the most common class of our example raster
E17. Plot the histogram and the boxplot of the
dem.tif file from the spDataLarge package (
system.file("raster/dem.tif", package = "spDataLarge")).