A question regarding matrices

4 views (last 30 days)
jana
jana on 16 Oct 2013
Edited: Vivek Selvam on 16 Oct 2013
I have a data file that looks like this:
SECTION Graph
Nodes 10
Edges 5
E 1 2 10
E 2 4 2
E 4 3 1
E 9 10 9
E 4 10 6
E 1 2 10 means there is an edge between 1 and 2 and have a cost 10. I would like to read this data from my data file and create an adjacency matrix G using this data where the G(i,j) is the cost of edge(i,j) and it is -1 if there is no edge. Please Help, I am fairly new to matlab.
  4 Comments
Jos (10584)
Jos (10584) on 16 Oct 2013
Edited: Jos (10584) on 16 Oct 2013
There are some inconsistencies in your question and comment:
  • Does "E 1 2 10" mean that G(1,2) will be 10 but also that G(2,1) should be 10?
  • And do you want it to be 0 or -1 when there is no edge?
jana
jana on 16 Oct 2013
yes, "E 1 2 10" mean that G(1,2) will be 10 but also that G(2,1) should be 10
I am sorry, it is -1 when there is no edge.

Sign in to comment.

Accepted Answer

Yatin
Yatin on 16 Oct 2013
Edited: Yatin on 16 Oct 2013
Hi,
Below is a working code for reading your specific file.
function adjMatrix = adjMatrix()
fileID = fopen('adjMat.txt');
tline = fgets(fileID);
count = 1;
nodes = 0;
edges = 0;
adjMatrix = [];
while(ischar(tline))
count = count +1;
tline = fgets(fileID);
if(tline == -1)
break;
end
tline = strtrim(tline);
if(count == 2)
nodes = str2num(tline(end-1:end));
adjMatrix = zeros(nodes);
elseif(count == 3)
edges = str2num(tline(end));
else
splitStr = strsplit(tline);
node1 = str2num(splitStr{2});
node2 = str2num(splitStr{3});
cost = str2num(splitStr{4});
adjMatrix(node1,node2) = cost;
adjMatrix(node2,node1) = cost;
end
end
fclose(fileID);
adjMatrix(adjMatrix == 0) = -1;
end

More Answers (2)

Vivek Selvam
Vivek Selvam on 16 Oct 2013
Edited: Vivek Selvam on 16 Oct 2013
Using dlmread():
function G = adjacencyMatrix()
filename = 'myFile.txt';
delimiterIn = ' ';
startRow = 1;
startCol = 1;
M = dlmread(filename,delimiterIn,startRow,startCol);
numNodes = M(1,1);
G = -1*ones(numNodes);
node1 = M(3:end,1);
node2 = M(3:end,2);
cost = M(3:end,3);
for k = 1:numel(node1)
G(node1(k),node2(k)) = cost(k);
G(node2(k),node1(k)) = cost(k);
end
end

Yannick
Yannick on 16 Oct 2013
Edited: Yannick on 16 Oct 2013
Here's another option using TEXTSCAN. Note also the illustration of SPARSE and FULL at the end of the code. The main difficulty here may be to understand how to use cell arrays in MATLAB. You can find more information here: Cell arrays doc. At the very end, I also use logical indexing: Logical indexing doc page
% Open file
fileID = fopen('adjMat.txt');
% Read number of nodes and edges
H = textscan(fileID, '%s %f', 2, 'HeaderLines', 1);
disp(['Number of ' H{1}{1} ':'])
NumberOfNodes = H{2}(1)
disp(['Number of ' H{1}{2} ':'])
NumberOfEdges = H{2}(2)
% Read all remaining lines ("E x x x" lines)
C = textscan(fileID, '%c %f %f %f');
ArrayOfInitialNodes = C{2}'
ArrayOfFinalNodes = C{3}'
ArrayOfEdgeWeights = C{4}'
% Close file
fclose(fileID);
% Create adjacency matrix in sparse format with zeros instead of -1
% (memory efficient if there are lots of zeros)
AdjMatrixSparseWithZeros = ...
sparse(ArrayOfInitialNodes, ArrayOfFinalNodes, ArrayOfEdgeWeights, ...
NumberOfNodes, NumberOfNodes);
AdjMatrixSparseWithZeros = ...
AdjMatrixSparseWithZeros + AdjMatrixSparseWithZeros'
% Transform it to full format (no compact storage in memory)
% and replace zeros with -1
AdjMatrixFullWithNegOnes = full(AdjMatrixSparseWithZeros);
AdjMatrixFullWithNegOnes(AdjMatrixFullWithNegOnes==0) = -1
Hope this helps provide yet another perspective!

Categories

Find more on Historical Contests in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!