https://github.com/mitya57/jordan-system

A program to solve linear systems using the Gauss–Jordan method.
https://github.com/mitya57/jordan-system

Last synced: 21 days ago
JSON representation

A program to solve linear systems using the Gauss–Jordan method.

• Host: GitHub
• URL: https://github.com/mitya57/jordan-system
• Owner: mitya57
• Created: 2012-10-27T12:47:12.000Z (about 12 years ago)
• Default Branch: master
• Last Pushed: 2014-03-27T14:40:37.000Z (over 10 years ago)
• Last Synced: 2024-10-13T17:05:39.713Z (24 days ago)
• Language: C++
• Size: 199 KB
• Stars: 0
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.mkd

Awesome Lists containing this project

README

        
This program is able to solve linear systems using the Jordan method.

## Usage

### Reading matrix from the file

    matrix MatrixSize BlockSize FileName

will read the matrix from file `FileName`, which should have the following format:

    N

    Element00 Element01 ... Element0N   RightColumn0

    Element10 Element11 ... Element1N   RightColumn1

    . . . . .

    ElementN0 ElementN1 ... ElementNN   RightColumnN

### Automatically generating the matrix

    matrix MatrixSize BlockSize

will automatically generate the matrix using a formula given in `function.hh` file.

### Compiling

To reach the maximum performance, we recommend using the [profile-guided optimization][1].

First you need to generate the profile. To do that, run:

    g++ -O3 -lpthread -fprofile-generate matrix.cc -o matrix

Then run the `matrix` executable as described below and wait for the profile to be generated.

After that is done, run:

    g++ -O3 -lpthread -fprofile-use matrix.cc -o matrix

Now your program will be optimized based on its profile. You can also use `make` targets

(`make pg` and `make pu`, respectively).

[1]: https://en.wikipedia.org/wiki/Profile-guided_optimization

## Running the tests

To ensure that the functions work well, a test program `matrix_tests.cc` is included.