Gsoc 2013 Mukul Priya - pgRouting/pgrouting GitHub Wiki

Report List

Biography

GSoC project

My Status Reports

OSGeo Wiki notes: http://wiki.osgeo.org/wiki/Google_Summer_of_Code_2013_Ideas#What_to_expect_during_the_summer

Report 14

Work done this week:

  • Completed integration with pgrouting.The algorithm however will need some more testing before it is available with the pgrouting package. The proposed goals have been met for now but i will always be available in future for its further testing , modification or bug fixing.The codes are available on the git repository I would like to thank my mentors Steve and Daniel for their support and guidance.

Next week plan :

  • upload code samples on melange.
  • Have the documentation ready.

Report 13

What did i do this week ??

  • Started integrating it with pgrouting and tested the code on a bigger scale.

What will I be working on next week ?

  • Will focus on integration and documentation.I will have discussions with my mentors regarding the same.In between further improve the code if required.

Did I meet with any stumbling blocks?

  • Not much .

Report 12

What did i do this week ??

  • Fixed some more errors in the code. Set up the whole testing environment .

What will I be working on next week ?

  • Testing and integration with pgrouting. A little bit of documentation if possible.

Did I meet with any stumbling blocks?

  • Got stuck with two or three segmentation faults ,that consumed time.

Report 11

What did i do this week ??

  • This week i fixed a lot of compile errors as i was trying to test my code.I generated my own test data.Overall the whole week was devoted to debugging.

What will I be working on next week ?

  • Fix all the compile errors , have my test cases formulated in the input file.Get the read me file ready so that the community can have it tested as well .

Did I meet with any stumbling blocks?

  • The debugging took some time , i hope i can finish the integration in time before the deadline .

Report 10

What did i do this week ??

  • Completed the solver class , will focus now on improving it a bit in some areas.

What will I be working on next week ?

*start testing it using input from files and make appropriate changes as required . update the c functions appropriately .Discuss with my mentor about any change if required .

Did I meet with any stumbling blocks?

  • i was having a health issue that elongated a bit and that derailed a lot of my plans.

Report 9

What did i do this week ??

  • The solver class is almost complete.There are still a few things that need to be fixed.

What will I be working on next week ?

  • Complete the solver class and start testing it.

Did I meet with any stumbling blocks?

  • The ethernet card of my laptop malfunctioned due to which i was not able to even connect to internet or update my code repository. Initially i thought it was some proxy issue . I have called the customer care people and they will change either the card or the motherboard by Monday or tuesday. I will be able to update the repository by then only.

Report 8

What did i do this week ??

  • Added some more lines of code to the solver class. It is yet not complete but i will try hard to complete it this week .

What will I be working on next week ?

  • Achieve the above objective .

Did I meet with any stumbling blocks?

  • I wasn't able to work much this week due to a health issue . I am still recovering from it and i will compensate the working hours that i lost in coming weeks. I deeply regret for the time that i lost due to my health issue this week .

Report 7

What did i do this week ??

  • Made significant progress towards a working solver class. Hopefully it will be ready by this week according to my original schedule.After that some testing process will be initiated.

What will I be working on next week ?

  • Achieve the above objective .

Did I meet with any stumbling blocks?

  • Nothing very significant. Steve did point out some mistakes in c++ code.

Report 6

What did i do this week ??

  • Reimplemented the partition retrieval mechanism so that regularly the partition can be fetched whenever demand arises and we don't have to make ant recursive SPI_calls ( inbuilt function which talks to database). Made few changes to the solver class which constructs the graph and computes the path. still in the process of making those changes .

What will I be working on next week ?

  • Try and complete implementing the solver class.

Did I meet with any stumbling blocks?

  • Basic implementation issue that was solved through discussion with Steve.

Any changes in schedule required ??

  • No i don't think any change is required . I devoted week 5-8 for solver class implementation. I will have my solver class completed by then.

Report 5

What did i do this week ??

  • started Implementing the main routing algorithm ( astar) , was finally able to implement partition retrieval mechanism using partition id. I hope that i will be able to complete a functional astar algorithm by next week , atleast a basic one that manually takes some input ( from file ).

What will I be working on next week ?

  • will try to complete the above mentioned task.

Did I meet with any stumbling blocks?

  • some c++ issues which i googled successfully.

Report 4

What did i do this week ??

  • Tried coding the partition retrieval and graph increment mechanism, the work is still in progress.I am having some difficulty but i hope i can do it this weekend and commit it by sunday.

What will I be working on next week ?

  • Complete the above mentioned work and try to start implementing astar.

Did I meet with any stumbling blocks?

  • Basically coding issues but learned a lot while tackling them.

Report 3

What did i do this week ??

  • Implemented the basic partitioning stuff with help from Steve's code on a sample self generated data set.
  • Tried to implement the same on an OSM data set .
  • Currently working on mechanism for retrieval of partitions so that the graph can be incremented.

What will I be working on next week ?

  • Basically get point 3 of the above done i.e code the partition retrieval and graph increment mechanism.

Did I meet with any stumbling blocks?

  • had some basic doubts that were solved through discussions.

Report 2

What did i do this week ??

  • Tried understanding the internal aspects of pgrouting i.e studied about the basic data flow using bd_astar source code .
  • Tested some basic pl/pgsql codes.
  • Read and deployed the basic partitioning code that Steve provided.Have made some changes to it, will try testing it this weekend.

What will I be working on next week ?

  • For the next two weeks the major focus will be to implement the partitioning stuff using quad tree approach as discussed with Steve and Daniel . Will have to discuss and tackle new issues that may pop up during this period.

Did I meet with any stumbling blocks?

  • It took a while for me to understand the basic data flow.
  • I am new to procedural languages but it wasn't so tricky to understand.

Report 1

What did i do this week ?

  • Installed postgresql , postgis and pgrouting successfully on a virtual machine that was set up using vagrant.
  • Built and tested pgrouting with some querries.
  • Got familiar with git, i was already provided with gsoc-partition branch by my mentors,so didn't have to do much regarding this.

What will I be working on next week?

  • Try to understand the internal aspects of pgrouting.
  • Get more familiar with PL/pgSQL Procedural Language.
  • Try and modify the code of partitioning that Steve provided.

Did I meet with any stumbling blocks?

  • Had to face a few problems because of proxy issue while i was trying to set up the development environment.Thanks to Steve and Daniel for the quick responses regarding the same.