Planned intervention: On Wednesday April 3rd 05:30 UTC Zenodo will be unavailable for up to 2-10 minutes to perform a storage cluster upgrade.
Published November 4, 2019 | Version 1
Software Open

Diffusion tensor imaging codes from "Primal-dual block-proximal splitting for a class of non-convex problems"

  • 1. Loschmidt Laboratories, Masaryk University
  • 2. University of Eastern Finland
  • 3. University of Helsinki & Escuela Politécnica Nacional

Description

These are the Julia codes for the diffusion tensor imaging experiments of the manuscript “Primal-dual block-proximal splitting for a class of non-convex problems” by S. Mazurenko, J. Jauhiainen, and T. Valkonen (arXiv:1911.06284). The codes were written by T. Valkonen.

Prerequisites

These codes we written for Julia 1.1 but are known to work with Julia 1.2. The Julia package prequisites are from May 2019 when our experiments were run, and have not been updated to maintain the same environment we used to do the experiments in the manuscript. You may get Julia from https://julialang.org/.

To generate the Helix test data (the data we have generated are included), the Teem toolkit is needed. You may download it from https://sourceforge.net/projects/teem/. Please note that the version included in Homebrew is missing the emap binary that is required for the visualisation; indeed as of May 2019 the building of this binary is disabled in Teem. You will therefore need to build Teem manually from source, applying the included teem-1.11.0-emap_build_fix.patch. In your unix shell, in the top-level directory of the Teem toolkit source codes, run:

$ patch -p1 < path_to_these_codes/teem-1.11.0-emap_build_fix.patch

Then build and install Teem according to instructions.

The visualisation further requires the open -g to open the generated PDF file in a PDF viewer that will correctly refresh the file on further launches. This is the case on MacOS with Preview or Skim.

Using

Navigate your unix shell to the directory containing this README.md and then run:

$ julia --project=BlockPDPS

Afterwards in the Julia shell, type:

> using BlockPDPS

This may take a while as Julia builds any missing dependencies. Then, to run the default experiments, run:

> test_dti_helix(visualise=true)

This will use any cached data in the data/ subdirectory of the current working directory. If no cached data is found, new data will be generated using Teem. If you use provided data and set visualise=false, the Teem dependency will be removed. The results are saved under the img/ subdirectory of the current working directory.

The test_dti_helix function has further parameters; please see the source code for details.

Files

nlpdps_block_dti_codes-832b9910dd9a.zip

Files (6.0 MB)

Name Size Download all
md5:00ad08cd28a5a9a583cc25635258f135
6.0 MB Preview Download

Additional details

Related works

Is cited by
Journal article: 10.1553/etna_vol51s15 (DOI)
Preprint: arXiv:1609.07373 (arXiv)

Funding

PARTIAL Analysis of Relations in Tasks of Inversion for Algorithmic Leverage EP/P021298/1
UK Research and Innovation