geographical convergent cross mapping

Usage

# S4 method for class 'sf'
gccm(
  data,
  cause,
  effect,
  libsizes,
  E = 3,
  tau = 1,
  k = E + 2,
  theta = 1,
  algorithm = "simplex",
  lib = NULL,
  pred = NULL,
  nb = NULL,
  threads = detectThreads(),
  parallel.level = "low",
  bidirectional = TRUE,
  trend.rm = TRUE,
  progressbar = TRUE
)

# S4 method for class 'SpatRaster'
gccm(
  data,
  cause,
  effect,
  libsizes,
  E = 3,
  tau = 1,
  k = E + 2,
  theta = 1,
  algorithm = "simplex",
  lib = NULL,
  pred = NULL,
  threads = detectThreads(),
  parallel.level = "low",
  bidirectional = TRUE,
  trend.rm = TRUE,
  progressbar = TRUE
)

Arguments

data: The observation data.
cause: Name of causal variable.
effect: Name of effect variable.
libsizes: A vector of library sizes to use.
E: (optional) Dimensions of the embedding.
tau: (optional) Step of spatial lags.
k: (optional) Number of nearest neighbors to use for prediction.
theta: (optional) Weighting parameter for distances, useful when algorithm is smap.
algorithm: (optional) Algorithm used for prediction.
lib: (optional) Libraries indices.
pred: (optional) Predictions indices.
nb: (optional) The neighbours list.
threads: (optional) Number of threads.
parallel.level: (optional) Level of parallelism, low or high.
bidirectional: (optional) whether to identify bidirectional causal associations.
trend.rm: (optional) Whether to remove the linear trend.
progressbar: (optional) whether to print the progress bar.

Value

A list

xmap: cross mapping prediction results
varname: names of causal and effect variable
bidirectional: whether to identify bidirectional causal associations

References

Gao, B., Yang, J., Chen, Z. et al. Causal inference from cross-sectional earth system data with geographical convergent cross mapping. Nat Commun 14, 5875 (2023).

Examples

columbus = sf::read_sf(system.file("case/columbus.gpkg", package="spEDM"))
# \donttest{
g = gccm(columbus,"hoval","crime",libsizes = seq(5,45,5),E = 6)
#> 
Computing: [========================================] 100% (done)                         
#> 
Computing: [========================================] 100% (done)                         
g
#>   libsizes hoval->crime crime->hoval
#> 1        5    0.1301861    0.2642894
#> 2       10    0.1762769    0.3961522
#> 3       15    0.2197481    0.4672470
#> 4       20    0.2324435    0.5144297
#> 5       25    0.2354728    0.5713321
#> 6       30    0.2360934    0.6048914
#> 7       35    0.2412944    0.6253026
#> 8       40    0.2383788    0.6364897
#> 9       45    0.2339483    0.6484721
plot(g, ylimits = c(0,0.85))

# }