Skip to contents

Get the inter-domain edge list from a decomposition

Source: R/decompose_network.R
get_domain_graph.Rd

Returns the inter-domain edge list as a hydroloom edge list (hy_topo or hy_flownetwork). Two kinds of relationships appear in the result, selected by relations:

  • "flow" – pulled from nexus_registry. Each row whose to_domain_id is non-NA contributes one row to the result, recording where one domain hands off to the next at a hydro nexus. The result row carries that nexus's id in nexus_id and relation_type = "flow".

  • "contained" – pulled from each domain's containing_domain_id slot. Each non-NA value contributes one row, from the contained domain to its container. Containment is declared post-decomposition via set_containment() and does not pass through a hydro nexus, so the result row carries nexus_id = NA_character_ and relation_type = "contained".

The default returns both kinds; pass relations = "flow" for the inter-domain flow graph only, or relations = "contained" for containment relationships only.

The returned object is passed through classify_hy() so it carries the most-specific hydroloom class (hy_topo when the inter-domain graph is dendritic, hy_flownetwork when it is not). This lets downstream hydroloom functions like check_hy_graph() and sort_network() dispatch on it without extra conversion.

Usage

get_domain_graph(decomposition, relations = c("flow", "contained"))

Arguments

decomposition

object of class domain_decomposition.

relations

character vector. Which relation_type values to include. Default is both "flow" and "contained".

Value

hydroloom edge list (hy_topo or hy_flownetwork) with columns id, toid, nexus_id, relation_type.

See also

domain_decomposition for the wrapper object, decompose_network(), set_containment().

Examples

g <- sf::read_sf(system.file("extdata/walker.gpkg", package = "hydroloom"))

h <- hy(g) |>
  add_toids() |>
  add_levelpaths(name_attribute = "GNIS_ID",
    weight_attribute = "arbolate_sum")

d <- decompose_network(h)

get_domain_graph(d, relations = "flow")
#> # hydroloom dendritic edge list (self-referencing): 0 features
#> [1] id            toid          nexus_id      relation_type
#> <0 rows> (or 0-length row.names)