Skip to content

Latest commit

 

History

History
62 lines (47 loc) · 2.11 KB

README.md

File metadata and controls

62 lines (47 loc) · 2.11 KB
Raw

A Parallel Gaussian Algorithm

How to use

Do something like this to run an example.

Read("read_hpc.g");
Read("main_full_par_trafo.g");
n := 4000;; chopSize := 8;; q := 5;;
A := RandomMat(n, n, GF(q));;
ChiefParallel(GF(q), A, chopSize, chopSize);

To get some benchmark data on

  • how the algorithm performs in relation to the Gauss pkg
  • wall and CPU time
  • lock contention

do Read("measure_contention.g");. The file will tell you how to proceed.

Output

At the moment the output looks different from what the Gauss pkg returns. The correspondence of record component names is (Gauss name = GaussPar name)

  • vectors = remnant
  • coeffs = transformation

Old info

The following information need to be updated and should go - in the form of comments - into the files they are about.

File contents

read_hpc.g

Loads all necessary functions into HPCGAP, assuming access to packages "IO" and "GAUSS".

Is this still true: My current versions of HPCGAP/ the GAUSS pkg are not compatible, hence the repo at the moment uses a rather obscure looking work-around..)

main_full_par_trafo.g

Contains a version of the elimination alg. for HPCGAP computing RREF and a transformation, running completely in parallel.

utils.g

Collection of small basic functions used in subfunctions of the algorithm

subfunctions.g

Collection of larger subfunctions used in the Gaussian elimination alg.

File contents of unused / unnecessary (?) files

main_seq_trafo.g

Contains a version of the elimination alg. for standard GAP computing RREF and a transformation

main_semi_par_trafo.g

Contains a version of the elimination alg. for HPCGAP computing RREF and a transformation, where the second step of the algorithm runs in parallel ( using HPCGAP's task arch. )

main_par_trafo.g

Contains a version of the elimination alg. for HPCGAP computing RREF and a transformation, where the first step of the algorithm runs in parallel ( using HPCGAP's task arch. )

read.g

Loads all necessary functions into standard GAP, assuming access to packages "IO" and "GAUSS".