get_valid_edges: Get valid edges

In michaelwitting/wormLipidPredictR: Lipid reaction prediction based on WormJam

Get valid edges

Description

The function get_valid_edges gets valid edges from a graph. A valid edge is defined as follows: (1) take from a graph g the two measured vertices with indices i, j; (2) obtain all shortest paths between the measured vertices; (3) assign edge = 1 (valid), if there is/are ONLY one/multiple unmeasured intermediate vertex/vertices or ZERO intermediate vertices within the shortest paths (4) do not assign edge = 1 (not valid), if there is AT LEAST one measured intermediate vertex within shortest path

Usage

get_valid_edges(vertices, paths, vertex_attr_logical)

Arguments

vertices	vertices of igraph object
paths	list of igraph.vs objects
vertex_attr_logical	logical vector storing information if a vertex is measured

Details

Measured vertices will be denoted via vertex_attr_logical.

The function get_valid_edges is a helper function for collapse_network.

Author(s)

Thomas Naake, [e-mail]

Examples

edges <- data.frame(from = c("A", "A", "B", "C", "E", "F"), 
    to = c("B", "E", "C", "D", "F", "D"))
g <- igraph::graph_from_data_frame(edges, directed = TRUE)
## assign data availability to nodes, TRUE for measured vertices
igraph::V(g)$data_available <- igraph::V(g)$name %in% c("A", "D", "E")

## check if graph is directed
is_directed <- igraph::is_directed(graph = g)

## obtain the vertex attributes and create a vector, vertex_attr_logical,
## that stores information if a measurement is available (TRUE) or
## not (FALSE)
vertex_attr <- igraph::vertex_attr(graph = g, name = "data_available")
vertex_attr_logical <- logical(length(vertex_attr))

if (is.numeric(vertex_attr) | is.character(vertex_attr))
    vertex_attr_logical[!is.na(vertex_attr)] <- TRUE
if (is.logical(vertex_attr))
    vertex_attr_logical[vertex_attr] <- TRUE

## obtain the vertices where information is available
vertices_measured <- igraph::V(g)[vertex_attr_logical]

vertex_1 <- vertices_measured[1]
vertex_2 <- vertices_measured[3]

## find the shortest path from vertex_1 to vertex_2
path_1 <- suppressWarnings(igraph::all_shortest_paths(graph = g, 
    from = vertex_1, to = vertex_2)$res)

## find the shortest path from vertex_2 to vertex_1
path_2 <- suppressWarnings(igraph::all_shortest_paths(graph = g, 
    from = vertex_2, to = vertex_1)$res)

## check and add paths if valid, add a valid path when between 
## shortest paths there are only "unmeasured" vertices
## apply the helper function add_valid_edges for path_1 and path_2
get_valid_edges(vertices = igraph::V(g), paths = path_1, 
    vertex_attr_logical = vertex_attr_logical)
get_valid_edges(vertices = igraph::V(g), paths = path_2,
    vertex_attr_logical = vertex_attr_logical)

michaelwitting/wormLipidPredictR documentation built on Jan. 30, 2025, 12:30 a.m.