UoI

Distributed-UoI_Lasso

   Updated with BOOST CommandLine and rectified lasso_admm
  tested with real dataset and validated with sklearn Lasso results
                Runs tested on Cori NERSC supercomputer.

Requirements

languages: C, C++

computing resources: The model is created for Cori KNL.

API and Libraries: MPI, HDF5-parallel, eigen3, gsl

Installation

1. Clone the module into your directory
2. There are two subdirectories namely UoI_Lasso and UoI_VAR.
3. Please follow the README.md in each of these directories.

Credits

Author: Mahesh Balasubramanian (guidance from Kris Bouchard, Prabhat, Brandon Cook)

Publication

Mahesh Balasubramanian; Trevor Ruiz; Brandon Cook; Mr Prabhat; Sharmodeep Bhattacharyya; Aviral Shrivastava; Kristofer Bouchard: "Scaling of Union of Intersections for Inference of Granger Causal Networks from Observational Data." In: Proceeding of the 34th IEEE International Parallel & Distributed Processing Symposium (IPDPS)

Citation

  @inproceedings{BalasubramanianIPDPS2020,
  title = {Scaling of Union of Intersections for Inference of Granger Causal Networks from Observational Data},
  author = {Mahesh Balasubramanian and Trevor Ruiz and Brandon Cook and Mr Prabhat and Sharmodeep Bhattacharyya and Aviral Shrivastava and Kristofer Bouchard},
  year = {2020},
  date = {2020-05-01},
  booktitle = {Proceeding of the 34th IEEE International Parallel & Distributed Processing Symposium (IPDPS)},
  keywords = {Data Analysis},
  pubstate = {forthcoming},
  tppubtype = {inproceedings}
  }

Notes

The version of MPICH is reverted back to 7.4.4 because of an open ticket regarding the new mpich.