spatsoc is an R package for detecting spatial and temporal groups in GPS relocations. It can be used to build proximity-based social networks using gambit-of-the-group format and edge-lists. In addition, the randomization function provides data-stream randomization methods suitable for GPS data.
spatsoc
leverages data.table
to modify by reference and iteratively work on subsets of the input data. The first input for all functions in spatsoc
is DT
, an input data.table
. If your data is a data.frame
, you can convert it by reference using setDT(DF)
.
spatsoc
is designed to work in two steps: temporal followed by either spatial grouping or edge list generating. Considering your specific study species and system, determine a relevant temporal and spatial grouping threshold. This may be 5 minutes and 50 meters or 2 days and 100 meters or any other thresholds - the functions provided by spatsoc
are flexible to user input. In some cases, the spatial grouping function selected is only relevant with certain temporal grouping thresholds. For example, we wouldn’t expect a threshold of 5 minutes with group_polys
.
# Load packages
library(spatsoc)
library(data.table)
# Read data as a data.table
DT <- fread(system.file("extdata", "DT.csv", package = "spatsoc"))
# Cast datetime column to POSIXct
DT[, datetime := as.POSIXct(datetime)]
# Temporal groups
group_times(DT, datetime = 'datetime', threshold = '5 minutes')
## ID X Y datetime population minutes timegroup
## 1: A 715851.4 5505340 2016-11-01 00:00:54 1 0 1
## 2: A 715822.8 5505289 2016-11-01 02:01:22 1 0 2
## 3: A 715872.9 5505252 2016-11-01 04:01:24 1 0 3
## 4: A 715820.5 5505231 2016-11-01 06:01:05 1 0 4
## 5: A 715830.6 5505227 2016-11-01 08:01:11 1 0 5
## ---
## 14293: J 700616.5 5509069 2017-02-28 14:00:54 1 0 1393
## 14294: J 700622.6 5509065 2017-02-28 16:00:11 1 0 1394
## 14295: J 700657.5 5509277 2017-02-28 18:00:55 1 0 1449
## 14296: J 700610.3 5509269 2017-02-28 20:00:48 1 0 1395
## 14297: J 700744.0 5508782 2017-02-28 22:00:39 1 0 1396
# Spatial groups
group_pts(
DT,
threshold = 50,
id = 'ID',
coords = c('X', 'Y'),
timegroup = 'timegroup'
)
## ID X Y datetime population minutes timegroup
## 1: A 715851.4 5505340 2016-11-01 00:00:54 1 0 1
## 2: A 715822.8 5505289 2016-11-01 02:01:22 1 0 2
## 3: A 715872.9 5505252 2016-11-01 04:01:24 1 0 3
## 4: A 715820.5 5505231 2016-11-01 06:01:05 1 0 4
## 5: A 715830.6 5505227 2016-11-01 08:01:11 1 0 5
## ---
## 14293: J 700616.5 5509069 2017-02-28 14:00:54 1 0 1393
## 14294: J 700622.6 5509065 2017-02-28 16:00:11 1 0 1394
## 14295: J 700657.5 5509277 2017-02-28 18:00:55 1 0 1449
## 14296: J 700610.3 5509269 2017-02-28 20:00:48 1 0 1395
## 14297: J 700744.0 5508782 2017-02-28 22:00:39 1 0 1396
## group
## 1: 1
## 2: 2
## 3: 3
## 4: 4
## 5: 5
## ---
## 14293: 9909
## 14294: 9910
## 14295: 9911
## 14296: 9912
## 14297: 9913
See the vignette about using spatsoc in social network analysis.
Install GEOS
:
apt-get install libgeos-dev
pacman -S geos
dnf install geos geos-devel
brew install geos
data.table
igraph
sp
adehabitatHR
rgeos
group_times(DT, datetime, threshold)
DT
: input data.table
datetime
: date time column name in input data.tablethreshold
: threshold for groupingA data.table
with a date time formatted column. The input DT
will be returned with columns appended. The timegroup
column corresponds to the temporal group assigned to each row. Please note that the actual value of the time group is meaningless. Reordered data will return a different time group. What is meaningful, however, is the contents of each group. Each group will contain all rows nearest to the threshold provided.
The group_times
function expects either one column (POSIXct
) or two columns (IDate
and ITime
).
Given a character column representing the date time, convert it to POSIXct
or IDate
and ITime
:
DT[, datetime := as.POSIXct(datetime)]
DT[, c('idate', 'itime') := IDateTime(datetime)]
These are then provided to the function using the names of the column in the input data.
group_times(DT, datetime = 'datetime', threshold = '5 minutes')
or
group_times(DT, datetime = c('idate', 'itime'), threshold = '5 minutes')
The threshold
provided to group_times
should be related to the fix rate of the input dataset and to the specific study system and species. If relocations are recorded every two hours, a threshold = '2 hours'
will group all rows to the nearest two hour group (10am, 12pm, 2pm, 4pm, …). This, however, means that the relocations can be up to one hour apart from each other. Picking a smaller threshold, e.g.: threshold = '15 minutes'
may be more relevant in some cases. The flexibility of spatsoc
’s threshold argument means the user must carefully consider what threshold is reasonable to their specific system.
The threshold
of group_times
is considered only within the scope of 24 hours and this poses limitations on it:
threshold
must evenly divide into 60 minutes or 24 hoursthreshold
cannot be fractionalThe main column returned by group_times
is “timegroup”. It represents the temporal group of each row, where those nearest (either above or below) within the threshold are grouped. Its actual value does not have any meaning, but the contents of each group do. That means if the data is reordered, a row may have a different time group, but the other rows in that group should not change.
The extra columns are provided to help the user investigate, troubleshoot and interpret the timegroup.
threshold unit | column(s) added |
---|---|
minute | “minutes” column added identifying the nearest minute group for each row. |
hour | “hours” column added identifying the nearest hour group for each row. |
day | “block” columns added identifying the multiday block for each row. |
This message is returned to the user when a column matching those returned by group_times
is found in the input DT. This is commonly the case when group_times
is run multiple times consecutively.
This message is returned to the user when the threshold
is NULL. This is the default setting of threshold
and, at times, may be suitable. In this case, the date times in the datetime
column will be grouped exactly. Usually, a threshold should be provided.
This warning is returned to the user when the threshold
with unit days does not divide evenly into the range of days in DT. For example, if DT had data covering 30 days, and a threshold of ‘7 days’ was used, this warning would be returned. Note, this warning is returned for the range of days for the entire data set and not by year.
group_pts(DT, threshold, id, coords, timegroup, splitBy)
DT
: input data.table
threshold
: threshold for groupingid
: column name of IDs in DT
coords
: column names of x and y coordinates in DT
timegroup
: column name of time groupsplitBy
: (optional) column names of extra variables to group onThe input data.table
. It will returned with a column named group appended, which represents the spatial (and temporal if timegroup
is provided) group.
The threshold must be in the units of the coordinates.
The coordinates must be planar, such as UTM (of whichever zone your relocations are in).
group_lines(DT, threshold, projection, id, coords, timegroup, sortBy, splitBy, spLines)
DT
: input data.table
threshold
: threshold for groupingprojection
: projection of coordinates in DT
id
: column name of IDs in DT
coords
: column names of x and y coordinates in DT
timegroup
: (optional) column name of time groupsortBy
: column name of date time to sort rows for building linessplitBy
: (optional) column names of extra variables to group onsfLines
: alternatively, provide a sf LINESTRING object and id column nameSee 3.2.1.
The threshold
argument represents a buffer area around each line. When threshold = 0
, the lines are grouped by spatial overlap. If the threshold is greater than 0, the lines buffered, then grouped by spatial overlap.
The projection
argument expects a character string or numeric
defining the coordinate reference system to be passed to sf::st_crs
.
For example, for UTM zone 36S (EPSG 32736), the projection
argument is projection = "EPSG:32736"
or projection = 32736
.
See https://spatialreference.org for a list of EPSG codes.
Please note, R spatial has followed updates to GDAL
and PROJ for handling projections, see more at
https://r-spatial.org/r/2020/03/17/wkt.html.
The sortBy
argument expects a date time formatted column name, which is used to order the rows for each individual (and splitBy
).
group_polys(DT, area, hrType, hrParams, projection, id, coords, splitBy, spLines)
DT
: input data.table
area
: boolean argument if proportional area should be returnedhrType
: type of home range createdhrParams
: parameters relevant to the type of home range createdprojection
: projection of coordinates in DT
id
: column name of IDs in DT
coords
: column names of x and y coordinates in DT
splitBy
: (optional) column names of extra variables to group onsfPolys
: alternatively, provide a simple features POLGON or MULTIPOLYGON object and an id columnIf area = FALSE
, see 3.2.1. If area = TRUE
, the DT will not be appended with a group column instead a data.table
with IDs and proportional area overlap will be returned.
The default unit for area overlap is square meters.
The projection
argument expects a character string or numeric
defining the coordinate reference system to be passed to sf::st_crs
.
For example, for UTM zone 36S (EPSG 32736), the projection
argument is projection = "EPSG:32736"
or projection = 32736
.
See https://spatialreference.org for a list of EPSG codes.
Please note, R spatial has followed updates to GDAL
and PROJ for handling projections, see more at
https://r-spatial.org/r/2020/03/17/wkt.html.
Currently, spatsoc
offers two types of home ranges provided by the adehabitatHR
package: ‘mcp’ (mcp
) and ‘kernel’ (kernelUD
and getverticeshr
). The parameters must match the arguments of those functions.
Internally, we match arguments to the functions allowing the user to provide, for example, both the percent (provided to getverticeshr
) and grid arguments (provided to mcp
).
group_polys(
DT,
area = FALSE,
projection = utm,
hrType = 'mcp',
hrParams = list(grid = 60, percent = 95),
id = 'ID',
coords = c('X', 'Y')
)
edge_dist(DT = NULL, threshold = NULL, id = NULL, coords = NULL, timegroup = NULL, splitBy = NULL, fillNA = TRUE)
DT
: input data.table
threshold
: threshold for groupingid
: column name of IDs in DT
coords
: column names of x and y coordinates in DT
timegroup
: column name of time groupsplitBy
: (optional) column names of extra variables to group onfillNA
: boolean indicating if NAs should be returned for individuals that were not within the threshold distance of any other. If TRUE, NAs are returned. If FALSE, only edges between individuals within the threshold distance are returned.This is the non-chain rule implementation similar to group_pts
. Edges are defined by the distance threshold and NAs are returned for individuals within each timegroup if they are not within the threshold distance of any other individual (if fillNA
is TRUE).
See the vignette Using edge list generating functions and dyad_id for details about the edge_dist
function.
edge_nn(DT = NULL, id = NULL, coords = NULL, timegroup = NULL, splitBy = NULL, threshold = NULL)
DT
: input data.table
id
: column name of IDs in DT
coords
: column names of x and y coordinates in DT
timegroup
: column name of time groupsplitBy
: (optional) column names of extra variables to group onthreshold
: (optional) spatial distance threshold to set maximum distance between an individual and their neighbour.This function can be used to generate edge lists defined either by nearest neighbour or nearest neighbour with a maximum distance. NAs are returned for nearest neighbour for an individual was alone in a timegroup (and/or splitBy) or if the distance between an individual and it’s nearest neighbour is greater than the threshold.
See the vignette Using edge list generating functions and dyad_id for details about the edge_nn
function.
randomizations(DT, type, id, datetime, splitBy, iterations)
DT
: input data.table
type
: one of ‘daily’, ‘step’ or ‘trajectory’id
: Character string of ID column namedatetime
: field used for providing date time or time group - see detailssplitBy
: List of fields in DT to split the randomization process byiterations
: The number of iterations to randomizeSee the vignette Using spatsoc in social network analysis for details about the randomizations
function (specifically the section ‘Data stream randomization’)
(Almost) all functions in spatsoc
use data.table’s modify-by-reference to reduce recopying large datasets and improve performance. The exceptions are group_polys(area = TRUE)
, randomizations
and the edge list generating functions edge_dist
and edge_nn
.
Check that your data.table
has columns allocated (with data.table::truelength
) and if not, use data.table::setDT
or data.table::alloc.col
. This can happen if you are reading your data from RDS
or RData
files. See here.
if (truelength(DT) == 0) {
setDT(DT)
}
# then go to spatsoc
group_times(DT, datetime = 'datetime', threshold = '5 minutes')
or simply:
DT <- readRDS('path/to/data.Rds')
alloc.col(DT)
Here are some useful code chunks for understanding the spatial and temporal extent of your data and the outputs of spatsoc
functions.
# Number of unique individuals
DT[, uniqueN(ID)]
# Number of unique individuals by timegroup
DT[, uniqueN(ID), by = timegroup]
# Min, max datetime
DT[, range(datetime)]
# Difference between relocations in hours
DT[order(datetime),
.(difHours = as.numeric(difftime(datetime, shift(datetime), units = 'hours'))),
by = ID]
# Difference between relocations in hours
DT[order(datetime),
.(difMins = as.numeric(difftime(datetime, shift(datetime), units = 'mins'))),
by = ID]
Simple spatial extents can be calculated for all individuals or by individual.
# All individuals
DT[, .(minX = min(X),
maxX = max(X),
minY = min(Y),
maxY = max(Y),)]
# By individual
DT[, .(minX = min(X),
maxX = max(X),
minY = min(Y),
maxY = max(Y),),
by = ID]
spatsoc
outputsAfter using the grouping functions, we can determine the number of individuals in a temporal or spatial group.
# Number of unique individuals by timegroup
DT[, uniqueN(ID), by = timegroup]
# Number of unique individuals by group
DT[, uniqueN(ID), by = group]