Compare neighbor time to total co-occurrence time

Compares the time pairs spend at neighboring bins to their total co-occurrence time (regardless of location). Calculates the proportion of co-occurrence time that is spent as spatial neighbors, which indicates pairs that prefer to be close to each other when feeding/drinking together.

Usage

synch_neighbor_compare(
  pair_results,
  neighbor_results,
  min_cooccurrence = 0,
  sort_by = "neighbor_ratio",
  decreasing = TRUE
)

Arguments

pair_results

List output from synch_pair_analysis(), containing total co-occurrence matrices

neighbor_results

List output from synch_neighbor_analysis(), containing neighbor-specific matrices

min_cooccurrence

Numeric, minimum co-occurrence time threshold to include a pair (default 0). Use this to filter out pairs with negligible total interaction.

sort_by

Character, column name to sort by. Common options: "neighbor_ratio" (default), "neighbor_time", "cooccurrence_time", or NULL for no sorting.

decreasing

Logical, if TRUE (default) sorts in descending order, if FALSE sorts ascending

Value

Data frame with columns:

• animal1: Character, first animal ID in pair

• animal2: Character, second animal ID in pair

• cooccurrence_time: Numeric, total time together anywhere

• neighbor_time: Numeric, time together at neighboring bins

• neighbor_ratio: Numeric, neighbor_time / cooccurrence_time

• day: Character, day identifier (only for multi-day input)

The neighbor_ratio ranges from 0 (never neighbors) to 1 (always at neighboring bins when together). Only includes pairs where cooccurrence_time > min_cooccurrence.

Examples

# Create toy data with animals at different bins
toy_data <- data.frame(
  animal = c(1, 2, 3, 1, 2),
  start = lubridate::ymd_hms(c(
    "2023-01-01 10:00:00", "2023-01-01 10:00:01",
    "2023-01-01 10:00:05", "2023-01-01 10:00:10",
    "2023-01-01 10:00:11"
  )),
  end = lubridate::ymd_hms(c(
    "2023-01-01 10:00:03", "2023-01-01 10:00:04",
    "2023-01-01 10:00:08", "2023-01-01 10:00:13",
    "2023-01-01 10:00:14"
  )),
  bin = c(1, 2, 3, 1, 2),
  start_weight = c(10.5, 8.3, 9.1, 10.2, 8.0),
  end_weight = c(10.2, 8.1, 8.9, 9.9, 7.8)
)

# Process matrices
matrices <- matrix_process(toy_data, type = "feed",
                          id_col = "animal", start_col = "start",
                          end_col = "end", bin_col = "bin",
                          start_weight_col = "start_weight",
                          end_weight_col = "end_weight",
                          bins_feed = 1:3)

# Analyze both total and neighbor patterns
pair_results <- synch_pair_analysis(matrices, type = "feed",
                                   id_col = "animal")
neighbor_results <- synch_neighbor_analysis(matrices,
                                           bin_layout = "1-2-3",
                                           type = "feed",
                                           id_col = "animal")

# Compare neighbor preference
comparison <- synch_neighbor_compare(pair_results, neighbor_results)
print(comparison)
#>   animal1 animal2 cooccurrence_time neighbor_time neighbor_ratio
#> 1       1       2                 6             6              1