Skip to content

Latest commit

 

History

History
92 lines (58 loc) · 3.9 KB

README.md

File metadata and controls

92 lines (58 loc) · 3.9 KB
Raw

Network extraction by routing optimization

NextRout - Network Extraction from Routing Optimization - is a tool created to extract Network topologies from dynamical equations related to routing optimization problems. It is divided into three steps: extraction of an optimal trajectory from a dynamical system of equations in a continuous framework; pre-extraction of the network; and filtering of possible redundancies.

You can use in input pre-build functions, graphs or images. The filtering allows you to have a less redundant network. You can choose whether only have a graph from a continuous framework on routing optimization, or go beyond and apply a similar dynamics that allows to execute a filtering without losing important information stored on the network.

NextRout is based on the theory described in this paper:

  • Network extraction by routing optimization, D. Baptista, D. Leite, E. Facca, M. Putti and C. D. Bacco, arxiv 2005.02805, 2020. Preprint can be found here.
    If you use this software please cite this work.

Prerequisites

The required dependencies are

  • For the DMK-Continuous

    • A Fortran compiler (any gfortran version>4 but version 5);
    • Blas and Lapack libraries;
    • Python 3;
    • Meshpy;
    • PyVTK;

  • For the DMK-discrete

    • Python 3 is enough.

Installation

Linux/OSX

Just clone this repository:

git clone https://github.com/Danielaleite/Nextrout
cd Nextrout
python setup.py

This is going to download the "DMK-solver" and all required python dependencies. The execution of "setup.py" takes around 9 minutes to be finished (so dont't panic if you notice it's taking a while). In the end, you should have the following folders:

  • otp_utilities
  • python_scripts
  • results

Inside otp_utilities there is a bash script named "init_otp_muffe.sh". This is a piece previously executed in "setup.py" that takes all the repositories at https://gitlab.com/opt-muffe/otp_utilities (within this link you can find a more detailed description to the DMK-solver). In python_scripts, as the name suggests, you will be able to find the main python scripts and subfolders where the entire procedure takes place. The folder results stores the final graphs and solutions from DMK-continuous. You can find in each folder another README.md with a more detailed description and instructions for each subfolder/script.

How to perform a simulation?

Inside "python_scripts" look for "network_extraction.py" ,

cd python_scripts
python network_extraction.py

For the input parameters, you should reply the following questions:

  1. Do you have an input.ctrl file? [y/N]:

If your answer is yes, skip to the question 8. If not,

  1. What is the input flag?

This is where you give the initial transport density. There are three options: predefined funcions for DMK-continuous (tdens0), images or your own graph; for more details, see https://github.com/Danielaleite/Nextrout/blob/master/python_scripts/default_inputs.md.

  1. What is the number of divisions of the mesh?
  2. Which beta should be used in the DMK-solver?
  3. What is are) the flag(s) for source function?
  4. What is(are) the flag(s) for sink function?
  5. Which beta should be used in the filtering?
  6. Should the DMK-solver be executed?
  7. Do you want the graph pre-extraction to be done?
  8. Do you need the graph filtering/reduction to be done?

After this, depending of your input parameters, it might take from a few seconds to minutes untill the entire procedure is completely done. To see the results, just do

cd results

Contributing

Nextrout is being developed continuously. We appreciate any help or suggestions. Do not hesitate to contact us in case you want to contribute or have any questions.

Authors

  • Daniela Leite, Diego Theuerkauf

See also the list of contributors who participated in this project.