Translated minimum_spanning_tree() (#3)

NAMESPACE

@@ -57,6 +57,7 @@ importFrom(matlab2r,isfield)
 importFrom(matlab2r,isspace)
 importFrom(matlab2r,max)
 importFrom(matlab2r,min)
+importFrom(matlab2r,nargin)
 importFrom(matlab2r,ones)
 importFrom(matlab2r,questdlg)
 importFrom(matlab2r,rand)

R/minimum_spanning_tree.R

@@ -1,51 +1,53 @@
-minimum_spanning_tree <- function(C1, C2) stop("needs translation")
+minimum_spanning_tree <- function(C1, C2) {
-# function A = minimum_spanning_tree(C1, C2)
+  #  Find the minimum spanning tree using Prim's algorithm.
-# %
+  #  C1(i, j) is the primary cost of connecting i to j.
-# % Find the minimum spanning tree using Prim's algorithm.
+  #  C2(i, j) is the (optional) secondary cost of connecting i to j, used to break ties.
-# % C1(i,j) is the primary cost of connecting i to j.
+  #  We assume that absent edges have 0 cost.
-# % C2(i,j) is the (optional) secondary cost of connecting i to j, used to break ties.
+  #  To find the maximum spanning tree, used - 1 * C.
-# % We assume that absent edges have 0 cost.
+  #  See Aho, Hopcroft & Ullman 1983, "Data structures and algorithms", p 237.
-# % To find the maximum spanning tree, used -1*C.
-# % See Aho, Hopcroft & Ullman 1983, "Data structures and algorithms", p 237.
 
-# % Prim's is O(V^2). Kruskal's algorithm is O(E log E) and hence is more efficient
+  #  Prim's is O(V^2). Kruskal's algorithm is O(E log E) and hence is more efficient
-# % for sparse graphs, but is implemented in terms of a priority queue.
+  #  for sparse graphs, but is implemented in terms of a priority queue.
 
-# % We partition the nodes into those in U and those not in U.
+  #  We partition the nodes into those in U and those not in U.
-# % closest(i) is the vertex in U that is closest to i in V-U.
+  #  closest(i) is the vertex in U that is closest to i in V - U.
-# % lowcost(i) is the cost of the edge (i, closest(i)), or infinity is i has been used.
+  #  lowcost(i) is the cost of the edge (i, closest(i)), or infinity is i has been used.
-# % In Aho, they say C(i,j) should be "some appropriate large value" if the edge is missing.
+  #  In Aho, they say C(i, j) should be "some appropriate large value" if (the edge is missing.) {
-# % We set it to infinity.
+  #  We set it to infinity.
-# % However, since lowcost is initialized from C, we must distinguish absent edges from used nodes.
+  #  However, since lowcost is initialized from C, we must distinguish absent edges from used nodes.
 
-# n = length(C1);
+  n <- length(C1)
-# if nargin==1, C2 = zeros(n); end
+  if (nargin() == 1) {
-# A = zeros(n);
+    C2 <- zeros(n)
+  }
+  A <- zeros(n)
 
-# closest = ones(1,n);
+  closest <- ones(1, n)
-# used = zeros(1,n); % contains the members of U
+  used <- zeros(1, n)# contains the members of U
-# used(1) = 1; % start with node 1
+  used[1] <- 1# start with node 1
-# C1(find(C1==0))=inf;
+  C1[find(C1 == 0)] <- Inf
-# C2(find(C2==0))=inf;
+  C2[find(C2 == 0)] <- Inf
-# lowcost1 = C1(1,:);
+  lowcost1 <- C1[1, ]
-# lowcost2 = C2(1,:);
+  lowcost2 <- C2[1, ]
 
-# for i=2:n
+  for (i in 2:n) {
-#   ks = find(lowcost1==min(lowcost1));
+    ks <- find(lowcost1 == min(lowcost1))
-#   k = ks(argmin(lowcost2(ks)));
+    k <- ks[argmin(lowcost2(ks))]
-#   A(k, closest(k)) = 1;
+    A[k, closest[k]] <- 1
-#   A(closest(k), k) = 1;
+    A[closest[k], k] <- 1
-#   lowcost1(k) = inf;
+    lowcost1[k] <- Inf
-#   lowcost2(k) = inf;
+    lowcost2[k] <- Inf
-#   used(k) = 1;
+    used[k] <- 1
-#   NU = find(used==0);
+    NU <- find(used == 0)
-#   for ji=1:length(NU)
+    for (ji in 1:length(NU)) {
-#     for j=NU(ji)
+      for (j in NU[ji]) {
-#       if C1(k,j) < lowcost1(j)
+        if (C1[k, j] < lowcost1[j]) {
-# 	lowcost1(j) = C1(k,j);
+          lowcost1[j] <- C1[k, j]
-# 	lowcost2(j) = C2(k,j);
+          lowcost2[j] <- C2[k, j]
-# 	closest(j) = k;
+          closest[j] <- k
-#       end
+        }
-#     end
+      }
-#   end
+    }
-# end
+  }
+  return(A)
+}

R/rBAPS-package.R

@@ -10,4 +10,5 @@
 #' size sortrows squeeze strcmp times zeros disp
 #' @importFrom stats runif
 #' @importFrom zeallot %<-%
+#' @importFrom matlab2r nargin
 NULL