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.

Usage

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).

Spatial and temporal grouping

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')

# Spatial groups
group_pts(
  DT,
  threshold = 50,
  id = 'ID',
  coords = c('X', 'Y'),
  timegroup = 'timegroup'
)
ID X Y datetime timegroup group
A 715851.4 5505340 2016-11-01 00:00:54 1 1
A 715822.8 5505289 2016-11-01 02:01:22 2 2
A 715872.9 5505252 2016-11-01 04:01:24 3 3
A 715820.5 5505231 2016-11-01 06:01:05 4 4
A 715830.6 5505227 2016-11-01 08:01:11 5 5

Social network analysis

See the vignette about using spatsoc in social network analysis.

Installation

System dependencies

GEOS

Install GEOS:

Package dependencies

Functions

group_times

group_times(DT, datetime, threshold)

DT

A 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.

datetime format

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')

threshold recommendations

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.

Limitations of threshold

The threshold of group_times is considered only within the scope of 24 hours and this poses limitations on it:

Columns returned by group_times

The 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.

Warnings and messages

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

group_pts(DT, threshold, id, coords, timegroup, splitBy)

DT

The input data.table. It will returned with a column named group appended, which represents the spatial (and temporal if timegroup is provided) group.

threshold

The threshold must be in the units of the coordinates.

coords

The coordinates must be planar, such as UTM (of whichever zone your relocations are in).

group_lines

group_lines(DT, threshold, projection, id, coords, timegroup, sortBy, splitBy, spLines)

DT

See 3.2.1.

threshold

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.

projection

The projection argument expects a character string defining the projection to be passed to sp::CRS. For example, for UTM zone 36S (EPSG 32736), the projection argument is projection = "EPSG:32736". See \url{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 \url{https://www.r-spatial.org/r/2020/03/17/wkt.html}.

Because of changes to projection handling, group_polys, build_polys, group_lines and build_lines may return the warning once, or many times:

In proj4string(xy) : CRS object has comment, which is lost in output
Calls: group_polys ... do.call -> <Anonymous> -> proj4string -> proj4string

This warning is explicitly verbose, to ensure we are considering the updated use of CRS in sp and other spatial packages.

sortBy

The sortBy argument expects a date time formatted column name, which is used to order the rows for each individual (and splitBy).

group_polys

group_polys(DT, area, hrType, hrParams, projection, id, coords, timegroup, splitBy, spLines)

DT and area

If 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.

projection

The projection argument expects a character string defining the projection to be passed to sp::CRS. For example, for UTM zone 36S (EPSG 32736), the projection argument is projection = "EPSG:32736". See \url{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 \url{https://www.r-spatial.org/r/2020/03/17/wkt.html}.

Because of changes to projection handling, group_polys, build_polys, group_lines and build_lines may return the warning once, or many times:

In proj4string(xy) : CRS object has comment, which is lost in output
Calls: group_polys ... do.call -> <Anonymous> -> proj4string -> proj4string

This warning is explicitly verbose, to ensure we are considering the updated use of CRS in sp and other spatial packages.

hrType and hrParams

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

edge_dist(DT = NULL, threshold = NULL, id = NULL, coords = NULL, timegroup = NULL, splitBy = NULL, fillNA = TRUE)

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

edge_nn(DT = NULL, id = NULL, coords = NULL, timegroup = NULL, splitBy = NULL, threshold = NULL)

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

randomizations(DT, type, id, datetime, splitBy, iterations)

See the vignette Using spatsoc in social network analysis for details about the randomizations function (specifically the section 'Data stream randomization')

Package design

Don't I need to reassign to save the output?

(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.

Why does a function print the result, but columns aren't added to my DT?

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)

Summary information

Here are some useful code chunks for understanding the spatial and temporal extent of your data and the outputs of spatsoc functions.

Number of individuals

# Number of unique individuals
DT[, uniqueN(ID)]

# Number of unique individuals by timegroup
DT[, uniqueN(ID), by = timegroup]

Temporal range

# 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]

Spatial extent

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 outputs

After 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]