Calculate a spatial lag — pg_spatial

Adds a weighted sum of neighboring cell values to a cell-level or cell-period data frame. The default is a row-standardized first-order queen neighborhood, so the resulting lag is the neighboring-cell mean.

Usage

pg_spatial_lag(
  data,
  value,
  pgid = "pgid",
  time = NULL,
  weights = NULL,
  method = c("queen", "rook"),
  wrap_longitude = TRUE,
  na_rm = FALSE,
  name = paste0(value, "_spatial_lag"),
  config = priogrid::pg_current_config()
)

Arguments

data: A data frame containing cell IDs and the variable to lag.
value: Name of the numeric value column.
pgid: Name of the PRIO-GRID cell-ID column.
time: Optional character vector naming temporal grouping columns.
weights: Optional table with pgid, neighbor_pgid, and weight. When omitted, weights are created from pg_neighbor_table().
method: Contiguity method used when weights is omitted.
wrap_longitude: Should default neighborhoods wrap across the antimeridian?
na_rm: Should missing and unavailable neighbor values be omitted?
name: Name of the new spatial-lag column.
config: A PRIO-GRID configuration object.

Value

data with the spatial-lag column appended.

Examples

config <- priogrid::pg_config(nrow = 18, ncol = 36)
x <- data.frame(pgid = 1:4, value = c(1, 2, 3, 4))
pg_spatial_lag(x, "value", config = config)
#>   pgid value value_spatial_lag
#> 1    1     1                 2
#> 2    2     2                 2
#> 3    3     3                 3
#> 4    4     4                 3