Converting to and from a cocomo dataset

Converting to and from a `cocomo` dataset

The cocomo R package was developed in the CoCoMo research group at the Max Planck Institute of Animal Behavior. The package has many functions for the analysis of communication and movement of animal groups.

cocomo functions require that animal track data be stored in 4 separate objects: 2 matrix objects, one each representing the x and the y positions of all individuals, 1 vector of timestamps, and 1 dataframe containing ids and information about the individuals. We will refer to this dataset structure as a “cocomo dataset” throughout this vignette.

Convert a `trackframe` object to a cocomo dataset

To transform a trackframe object into a cocomo dataset, use the tf_as_cocomo function. This function returns a single list that contains the 4 objects that make up a cocomo dataset of animal tracks.

library(trackframe)

# Create a trackframe object with the locations of 3 individuals,
# each observed over 7 time intervals
tf <- as.trackframe(
  data.frame(
    timestamp = rep(as.POSIXct(Sys.time() + 1:7), times = 3),
    x = runif(21, 0, 10),
    y = runif(21, 0, 10),
    id = rep(1:3, each = 7)),
  crs = NA)

# Convert tf to a cocomo dataset
cocomo_list <- tf_as_cocomo(tf)

# Check out the cocomo dataset
attributes(cocomo_list)

## $names
## [1] "xs"  "ys"  "t"   "ids"

str(cocomo_list)

## List of 4
##  $ xs : num [1:3, 1:7] 0.808 2.898 4.023 8.343 7.329 ...
##   ..- attr(*, "dimnames")=List of 2
##   .. ..$ : chr [1:3] "1" "2" "3"
##   .. ..$ : NULL
##  $ ys : num [1:3, 1:7] 6.78 6.96 4.32 7.35 6.89 ...
##   ..- attr(*, "dimnames")=List of 2
##   .. ..$ : chr [1:3] "1" "2" "3"
##   .. ..$ : NULL
##  $ t  : POSIXct[1:7], format: "2026-06-15 09:57:01" "2026-06-15 09:57:02" ...
##  $ ids:'data.frame': 3 obs. of  1 variable:
##   ..$ id_code: int [1:3] 1 2 3

cocomo_list$xs

##        [,1]     [,2]     [,3]      [,4]       [,5]      [,6]     [,7]
## 1 0.8075014 8.343330 6.007609 1.5720844 0.07399441 4.6639350 4.977774
## 2 2.8976724 7.328820 7.725215 8.7460066 1.74940627 0.3424133 3.203857
## 3 4.0232824 1.956698 4.035381 0.6366146 3.88701313 9.7554784 2.898923

cocomo_list$ys

##       [,1]     [,2]      [,3]     [,4]     [,5]      [,6]     [,7]
## 1 6.783804 7.353196 1.9595673 9.805397 7.415215 0.5144628 5.302125
## 2 6.958239 6.885560 0.3123033 2.255625 3.008308 6.3646561 4.790245
## 3 4.321713 7.064338 9.4857658 1.803388 2.168999 6.8016292 4.988456

cocomo_list$t

## [1] "2026-06-15 09:57:01 UTC" "2026-06-15 09:57:02 UTC"
## [3] "2026-06-15 09:57:03 UTC" "2026-06-15 09:57:04 UTC"
## [5] "2026-06-15 09:57:05 UTC" "2026-06-15 09:57:06 UTC"
## [7] "2026-06-15 09:57:07 UTC"

cocomo_list$ids

##   id_code
## 1       1
## 2       2
## 3       3

So, this cocomo dataset is a list that contains 4 objects: the first is a matrix of the x (easting) positions of every individual (rows) across all timestamps (columns); the second is a matrix of the y (northing) positions of every individual (rows) across all timestamps (columns), the third is a vector of the 7 timestamps, and the fourth is a dataframe of the 3 individuals’ ids.

We can then use a function from the cocomo package on the objects in this cocomo dataset, for example:

# Install cocomo package
library(devtools)
install_github('livingingroups/cocomo')

library(cocomo)

## cocomo package contains both experimental and stable functions. Experimental functions have not been code reviewed and are likely to change in the future.

# Get the group centroid at each time step
get_group_centroid(xs = cocomo_list$xs,
  ys = cocomo_list$ys)

## $x_centr
## [1] 2.576152 5.876283 5.922735 3.651569 1.903471 4.920609 3.693518
## 
## $y_centr
## [1] 6.021252 7.101031 3.919212 4.621470 4.197507 4.560249 5.026942

Convert a cocomo dataset to a `trackframe` object

To convert a cocomo dataset to a trackframe object, we can use the dedicated cocomo_as_tf function, and specify exactly where to find the information for each of the key columns in the arguments.

# Convert cocomo dataset to trackframe object
tf2 <- cocomo_as_tf(
  xs = cocomo_list$xs,
  ys = cocomo_list$ys,
  t = cocomo_list$t,
  ids = cocomo_list$ids,
  crs = NA)

tf2

##                   time    easting  northing id
## 1  2026-06-15 09:57:01 0.80750138 6.7838043  1
## 2  2026-06-15 09:57:02 8.34333037 7.3531960  1
## 3  2026-06-15 09:57:03 6.00760886 1.9595673  1
## 4  2026-06-15 09:57:04 1.57208442 9.8053967  1
## 5  2026-06-15 09:57:05 0.07399441 7.4152153  1
## 6  2026-06-15 09:57:06 4.66393497 0.5144628  1
## 7  2026-06-15 09:57:07 4.97777389 5.3021246  1
## 8  2026-06-15 09:57:01 2.89767245 6.9582388  2
## 9  2026-06-15 09:57:02 7.32881987 6.8855600  2
## 10 2026-06-15 09:57:03 7.72521511 0.3123033  2
## 11 2026-06-15 09:57:04 8.74600661 2.2556253  2
## 12 2026-06-15 09:57:05 1.74940627 3.0083081  2
## 13 2026-06-15 09:57:06 0.34241333 6.3646561  2
## 14 2026-06-15 09:57:07 3.20385731 4.7902455  2
## 15 2026-06-15 09:57:01 4.02328238 4.3217126  3
## 16 2026-06-15 09:57:02 1.95669835 7.0643384  3
## 17 2026-06-15 09:57:03 4.03538117 9.4857658  3
## 18 2026-06-15 09:57:04 0.63661457 1.8033877  3
## 19 2026-06-15 09:57:05 3.88701313 2.1689988  3
## 20 2026-06-15 09:57:06 9.75547835 6.8016292  3
## 21 2026-06-15 09:57:07 2.89892295 4.9884561  3

class(tf2)

## [1] "trackframe" "data.frame"

attributes(tf2)[1:6]

## $names
## [1] "time"     "easting"  "northing" "id"      
## 
## $row.names
##  [1]  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21
## 
## $class
## [1] "trackframe" "data.frame"
## 
## $time
## [1] "time"
## 
## $easting
## [1] "easting"
## 
## $northing
## [1] "northing"

Alison Ashbury

2026-06-15