RosettaCodeData/Task/Dijkstras-algorithm/R/dijkstras-algorithm.r

166 lines
3.4 KiB
R

# Digraph class implementation using S3
create_digraph <- function(edges) {
# Extract vertex names from edges
vnames <- unique(c(edges[, 1], edges[, 2]))
# Create adjacency list as named list
adjmat <- list()
for (i in 1:nrow(edges)) {
key <- paste(edges[i, 1], edges[i, 2], sep = "->")
adjmat[[key]] <- as.numeric(edges[i, 3])
}
# Create digraph object
digraph <- list(
edges = adjmat,
verts = vnames
)
class(digraph) <- "Digraph"
return(digraph)
}
# Accessor functions
vertices <- function(g) {
return(g$verts)
}
edges <- function(g) {
return(g$edges)
}
# Get neighbours of a vertex
neighbours <- function(g, v) {
neigh <- list()
edge_names <- names(g$edges)
for (edge_name in edge_names) {
parts <- strsplit(edge_name, "->")[[1]]
if (parts[1] == v) {
neigh[[parts[2]]] <- g$edges[[edge_name]]
}
}
return(neigh)
}
# Dijkstra's shortest path algorithm
dijkstra_path <- function(g, source, dest) {
# Check if source is in graph
if (!(source %in% vertices(g))) {
stop(paste(source, "is not a vertex in the graph"))
}
# Easy case
if (source == dest) {
return(list(path = c(source), cost = 0))
}
# Initialize variables
inf <- Inf
verts <- vertices(g)
dist <- setNames(rep(inf, length(verts)), verts)
prev <- setNames(verts, verts)
dist[source] <- 0
Q <- verts
# Precompute neighbours for all vertices
neigh <- list()
for (v in verts) {
neigh[[v]] <- neighbours(g, v)
}
# Main loop
while (length(Q) > 0) {
# Find vertex with minimum distance
u <- Q[which.min(dist[Q])]
Q <- Q[Q != u]
if (dist[u] == inf || u == dest) break
# Update distances to neighbours
for (v_name in names(neigh[[u]])) {
if (v_name %in% Q) {
alt <- dist[u] + neigh[[u]][[v_name]]
if (alt < dist[v_name]) {
dist[v_name] <- alt
prev[v_name] <- u
}
}
}
}
# Reconstruct path
path <- c()
cost <- dist[dest]
if (prev[dest] == dest) {
return(list(path = path, cost = cost))
} else {
current <- dest
while (current != source) {
path <- c(current, path)
current <- prev[current]
}
path <- c(current, path)
return(list(path = path, cost = cost))
}
}
# Test data
testgraph <- matrix(c(
"a", "b", 7,
"a", "c", 9,
"a", "f", 14,
"b", "c", 10,
"b", "d", 15,
"c", "d", 11,
"c", "f", 2,
"d", "e", 6,
"e", "f", 9
), ncol = 3, byrow = TRUE)
# Convert weights to numeric
testgraph[, 3] <- as.numeric(testgraph[, 3])
# Test function
test_paths <- function() {
g <- create_digraph(testgraph)
src <- "a"
dst <- "e"
result <- dijkstra_path(g, src, dst)
path <- result$path
cost <- result$cost
cat("Shortest path from", src, "to", dst, ": ")
if (length(path) == 0) {
cat("no possible path")
} else {
cat(paste(path, collapse = " → "))
}
cat(" (cost", cost, ")\n")
# Print all possible paths
cat("\n")
cat(sprintf("%4s | %3s | %s\n", "src", "dst", "path"))
cat("----------------\n")
for (src in vertices(g)) {
for (dst in vertices(g)) {
result <- dijkstra_path(g, src, dst)
path <- result$path
cost <- result$cost
path_str <- if (length(path) == 0) {
"no possible path"
} else {
paste0(paste(path, collapse = " → "), " (", cost, ")")
}
cat(sprintf("%4s | %3s | %s\n", src, dst, path_str))
}
}
}
# Run the test
test_paths()