[pgrouting-dev] Having problems under standing the distance matrix of TSP in respect to route network pgroute version 2

[Daniel Kastl]

Hi Dave,

Somehow we already discussed this in May:
http://pgrouting.974090.n3.nabble.com/pgrouting-dev-New-TSP-API-td4025094.html

I think the current function pgr_tsp function is good as it is, but some
alternative with SQL query is a good idea.
Then there would be two ways how to call TSP, for example

pgr_tsp(matrix float[][], start integer)
pgr_tsp(sql text, start integer)

But instead of rewriting it's better to write a wrapper function that
accepts SQL and transforms it into the matrix (and ensures that the matrix
is valid).

Daniel






On Fri, Jun 14, 2013 at 2:17 PM, Dave Potts <dave.potts at pinan.co.uk> wrote:

>  On 14/06/13 02:59, Daniel Kastl wrote:
> Hi Daniel, List
>
> I must admit I found the entire process of creating a postgress array in a
> suitable form for the pgr_tsp function to be a exercise in slow torture!
>
> Would anybody have any major objections if I attempted to rewrite the
> interface so it took a normal sql querry?
>
> Something like
>
> pgr_costResult[] pgr_tsp('SELECT source,target distance FROM vertex_table',2);
>
> That way the problem with creating a square array in postgres would disappear.
>
> Dave
>
>
>
> <https://www.google.nl/search?client=ubuntu&hs=uQi&channel=fs&gl=uk&q=exercise+in+slow+torture&spell=1&sa=X&ei=BKa6UYnAF5G20QXc_4HQCg&ved=0CCcQvwUoAA>
>
> Hi Dave,
>
>  Thanks for reporting. Maybe I just mixed up something with copy & paste.
> I remember that I needed to add a vertex table to the sample data for the
> new TSP with distance matrix and got the ID's a bit wrong first.
> Will take a look.
>
>  Daniel
>
>
> On Fri, Jun 14, 2013 at 7:40 AM, Stephen Woodbridge <
> woodbri at swoodbridge.com> wrote:
>
>> Dave - moving this over to the pgrouting list.
>>
>>
>> Hi Dave,
>>
>> I'm not sure the Daniel used the example data network for this example.
>> The data looks like one of the matrix I used in my test cases.
>>
>> src/tsp/test/*.test
>>
>> I just made up numbers for a symmetric matrix. It was not taken from an
>> actual network.
>>
>> So the sly part is that its NOT from a graph ;)
>>
>> Daniel - please correct me it you did something else.
>>
>> -Steve
>>
>> On 6/13/2013 6:11 PM, Dave Potts wrote:
>>
>>>
>>> In the travel sales person problem an example is provided,  I having
>>> problems understanding it
>>>
>>> The example route network is posted as
>>>
>>> http://docs.pgrouting.org/dev/doc/src/developer/sampledata.html#sampledata
>>>
>>> I have assume the distance between red node 1 and red node 3 is 2, ie
>>> the unit distance 1 +1  which I  make to be 2
>>>
>>> given that
>>>
>>> I think the route matrix should be
>>>
>>>      1 2 3 4
>>> ======
>>> 1 | 0 1 2 3
>>> 2 | 1 0 1 2
>>> 3 | 2 1 0 1
>>> 4 | 3 2 1 0
>>>
>>> distance 1 to 4 is 1+1 +1 ==3
>>> distance 2 to 4 is 1+1       ==2
>>> distance 3 to 4 is 1
>>> etc
>>>
>>> But the give value is (please see
>>> http://docs.pgrouting.org/dev/src/tsp/doc/index.html)
>>>      1 2 3 4
>>>
>>> 1  {0,1,2,3},
>>> 2  {1,0,3,2},
>>> 3  {2,3,0,4},
>>> 4  {3,2,4,0}
>>>
>>> I just don't understand how the distance between 3 and 4 can be the
>>> value 4, ditto he values between 2 and 3
>>>
>>> Have I missed something sly ?
>>>
>>> Dave.
>>>
>>>
>>>
>>> _______________________________________________
>>> postgis-devel mailing list
>>> postgis-devel at lists.osgeo.org
>>> http://lists.osgeo.org/cgi-bin/mailman/listinfo/postgis-devel
>>>
>>
>> _______________________________________________
>> postgis-devel mailing list
>> postgis-devel at lists.osgeo.org
>> http://lists.osgeo.org/cgi-bin/mailman/listinfo/postgis-devel
>>  _______________________________________________
>> pgrouting-dev mailing list
>> pgrouting-dev at lists.osgeo.org
>> http://lists.osgeo.org/mailman/listinfo/pgrouting-dev
>>
>
>
>
>  --
> Georepublic UG & Georepublic Japan
> eMail: daniel.kastl at georepublic.de
> Web: http://georepublic.de
>
>
> _______________________________________________
> pgrouting-dev mailing listpgrouting-dev at lists.osgeo.orghttp://lists.osgeo.org/mailman/listinfo/pgrouting-dev
>
>
>
> _______________________________________________
> pgrouting-dev mailing list
> pgrouting-dev at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/pgrouting-dev
>
>


-- 
Georepublic UG & Georepublic Japan
eMail: daniel.kastl at georepublic.de
Web: http://georepublic.de
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/pgrouting-dev/attachments/20130614/6d4a5a80/attachment-0001.html>