Evaluate reproducible monitoring-network thinning scenarios

Usage

ps_network_thinning(
  points,
  retain = c(0.75, 0.5, 0.25),
  design = c("random", "spatial_coverage", "user"),
  method = "TPS",
  repeats = 10,
  subsets = NULL,
  template = NULL,
  mask = NULL,
  grid_res = NULL,
  seed = 1,
  progress = NULL
)

Arguments

points: Groundwater-head points.
retain: Fractions or exact retained counts.
design: Random, spatial-coverage, or user subsets.
method: Interpolation method.
repeats: Number of planned runs per fraction.
subsets: User-supplied lists of retained IDs.
template, mask, grid_res: Fixed surface controls.
seed: Deterministic seed.
progress: Optional callback.

Value

A potentiomap_network_thinning object.

Examples

data("synthetic_wells")
p <- ps_make_points(synthetic_wells[1:10, ], "x", "y", "gw_elevation",
                    "well_id", "EPSG:26916")
thin <- ps_network_thinning(p, retain = 0.7, method = "IDW",
                            repeats = 2, grid_res = 350, seed = 8)
thin$summary
#>   planned_runs unique_retained_sets duplicate_retained_sets successful_runs
#> 1            2                    2                       0               2
# Full-surface differences are descriptive, not error against truth.