[pgrouting-users] suspected bug in TSP distance matrix pgRouting RC1 released

[Dave Potts]

I have been trying to do some work which requires the tsp, some of my 
values have values of 0.1 in the dataset, I keep getting odd results.

So I tried a well known example from the  pgr _tsp distance page and it 
work as expected.

I then repeated the same example but reduced all of the distance by a 
factor off 10 and got an error.  Unless I have make a mistake in my 
understanding of the manual page, I think we might a problem where

-- version used
select pgr_version();
                    pgr_version
-------------------------------------------------
  (2.0.0-dev,v2.0.0-rc1,0,bf13fd7,develop,1.48.0)
-- Use a example get a result set
SELECT seq, id FROM 
pgr_tsp('{{0,1,3,3},{1,0,2,2},{3,2,0,2},{3,2,2,0}}'::float8[],1);
  seq | id
-----+----
    0 |  1
    1 |  2
    2 |  3
    3 |  0
(4 rows)
-- repeat for reduce everything by  scale factor of 10  ie 1.0 becomes 
0.1 etc

SELECT seq, id FROM 
pgr_tsp('{{0,0.1,0.3,0.3},{0.1,0,0.2,0.2},{0.3,0.2,0,0.2},{0.3,0.2,0.2,0}}'::float8[],1);

ERROR:  Error TSP fail to findEulerianPath, check your distance matrix 
is valid.

As far as I understand it my matrix has 0's on the leading diangonal  
and [a,b] == [b,a]

Dave.