[pgrouting-dev] Boost graph algorithm

Hi all,

prior to routing, i'd like to extract connected components from my network.
Reading the mailing list i've learned this algorithm is implemented in
the boost graph library.

Could you tell me what are the main steps to make this algorithm
available from my database?
Or, could you advise me an existing tutorial about this?

Thanks in advance,

regards,

Adrien

On 2/26/2014 9:22 AM, Adrien ANDRÉ wrote:

Hi all,

prior to routing, i'd like to extract connected components from my network.
Reading the mailing list i've learned this algorithm is implemented in
the boost graph library.

Could you tell me what are the main steps to make this algorithm
available from my database?
Or, could you advise me an existing tutorial about this?

Hi Adrien,

On the wiki are some pages on development and getting started.
Basically the process we use is to have people look at pgr_dijkstra() function as this is pretty simple and it uses boost.

src/dijkstra and clone this into your own directory

Update the CMakeLists.txt file here and below and change the name of the files to your algorithm

src/dijkstra/doc -- is for your doc files
src/dijkstra/sql -- is the code the links the sql function to the C/C++ code
src/dijkstra/src -- is you C/C++ code
src/dijkstra/test -- are your test files for the automated test system
src/dijkstra/tester -- can be used to write a command lin test tool that test your code outside of the database

Jump right in and ask questions when you are confused or need help.

Oh, I recommend that you go to github and fork the pgrouting project into your account, then clone you fork, and create a branch off of master or develop and do you development in that branch. Then when you have code to submit back you can generate a pull request that is easy for us or other to merge back into the main development branch.

Thanks,
   -Steve

Hi,

Stephen, thank you for your answer.

Despite not being at ease with tools as Git and CMake or C/C++
development and compilation, i began following steps you described.

Here is the new branch:
https://github.com/adrienandrem/pgrouting/tree/components

I've noticed the
https://github.com/pgRouting/pgrouting/wiki/Developer---Getting-Started
page, and will keep you informed of my progress.

Regards,

Adrien

On 26/02/2014 11:52, Stephen Woodbridge wrote:

On 2/26/2014 9:22 AM, Adrien ANDRÉ wrote:

Hi all,

prior to routing, i'd like to extract connected components from my
network.
Reading the mailing list i've learned this algorithm is implemented in
the boost graph library.

Could you tell me what are the main steps to make this algorithm
available from my database?
Or, could you advise me an existing tutorial about this?

Hi Adrien,

On the wiki are some pages on development and getting started.
Basically the process we use is to have people look at pgr_dijkstra()
function as this is pretty simple and it uses boost.

src/dijkstra and clone this into your own directory

Update the CMakeLists.txt file here and below and change the name of the
files to your algorithm

src/dijkstra/doc -- is for your doc files
src/dijkstra/sql -- is the code the links the sql function to the C/C++
code
src/dijkstra/src -- is you C/C++ code
src/dijkstra/test -- are your test files for the automated test system
src/dijkstra/tester -- can be used to write a command lin test tool that
test your code outside of the database

Jump right in and ask questions when you are confused or need help.

Oh, I recommend that you go to github and fork the pgrouting project
into your account, then clone you fork, and create a branch off of
master or develop and do you development in that branch. Then when you
have code to submit back you can generate a pull request that is easy
for us or other to merge back into the main development branch.

Thanks,
  -Steve
_______________________________________________
pgrouting-dev mailing list
pgrouting-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/pgrouting-dev

On 3/10/2014 6:52 AM, Adrien ANDRÉ wrote:

Hi,

Stephen, thank you for your answer.

Despite not being at ease with tools as Git and CMake or C/C++
development and compilation, i began following steps you described.

Yes, there are a lot of moving pieces to get your head around. They are supposed to make things easier in the long run but the upfront learning curve can be pretty steep at times. Ask if you get stuck and there are a bunch of us here that will jump in with ideas and help.

Here is the new branch:
https://github.com/adrienandrem/pgrouting/tree/components

This is the right place to start.

I've noticed the
https://github.com/pgRouting/pgrouting/wiki/Developer---Getting-Started
page, and will keep you informed of my progress.

Great! I look forward to more reports and/or questions as you progress.

Thanks,
   -Steve

Regards,

Adrien

On 26/02/2014 11:52, Stephen Woodbridge wrote:

On 2/26/2014 9:22 AM, Adrien ANDRÉ wrote:

Hi all,

prior to routing, i'd like to extract connected components from my
network.
Reading the mailing list i've learned this algorithm is implemented in
the boost graph library.

Could you tell me what are the main steps to make this algorithm
available from my database?
Or, could you advise me an existing tutorial about this?

Hi Adrien,

On the wiki are some pages on development and getting started.
Basically the process we use is to have people look at pgr_dijkstra()
function as this is pretty simple and it uses boost.

src/dijkstra and clone this into your own directory

Update the CMakeLists.txt file here and below and change the name of the
files to your algorithm

src/dijkstra/doc -- is for your doc files
src/dijkstra/sql -- is the code the links the sql function to the C/C++
code
src/dijkstra/src -- is you C/C++ code
src/dijkstra/test -- are your test files for the automated test system
src/dijkstra/tester -- can be used to write a command lin test tool that
test your code outside of the database

Jump right in and ask questions when you are confused or need help.

Oh, I recommend that you go to github and fork the pgrouting project
into your account, then clone you fork, and create a branch off of
master or develop and do you development in that branch. Then when you
have code to submit back you can generate a pull request that is easy
for us or other to merge back into the main development branch.

Thanks,
   -Steve
_______________________________________________
pgrouting-dev mailing list
pgrouting-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/pgrouting-dev

_______________________________________________
pgrouting-dev mailing list
pgrouting-dev@lists.osgeo.org
http://lists.osgeo.org/mailman/listinfo/pgrouting-dev