GAP package EDIM

Author: Frank Lübeck
Needs: GAP in version at least 4.5.;
  a C-compiler to compile a new kernel function;
  GAPDoc in version at least 1.5 for accessing the documentation
Operating systems: Any, on which GAP is running. (For the provided kernel function the operating system needs to support dynamic loading of kernel code.)
Current version: 1.3.2 (released 12/06/2013)
Contact: Frank.Luebeck@Math.RWTH-Aachen.De
Download: See below for archives in several formats.
History: This text file CHANGES describes the most important differences to former versions of the package.


The main purpose of the EDIM package is to publish an implementation of an algorithm (found by the package author) for computing prime parts of the elementary divisors of integer matrices (i.e., the diagonal entries of the Smith normal form).

The programs are developed and already successfully used for large matrices (up to rank >12000) with moderate entries and many non-trivial elementary divisors which are products of some small primes. But they should be useful for other types of matrices as well.

Among the other functions of the package are: An inversion algorithm for large rational matrices (using a p-adic method), a program for finding the largest elementary divisor of an integral matrix (particularly interesting when this is much smaller than the determinant) and implementations of some normal form algorithms described by Havas, Majewski, Matthews, Sterling (using LLL- or modular techniques).


Detailed installation instructions are contained in the package README file.

Online Documentation

Here is the documentation of the EDIM package in several output formats. If you have installed the package as described above you can also access all of these documents from the GAP online help.

Package Archives for Download

The EDIM package is availabe in several formats:

Frank Lübeck / Last modified: Wed, 12 Jun 2013 16:45:48 +0200