HWMX. Level 1: matrix

This project is a task on course "Uses and applications of C++ language" by K.I.Vladimirov. The task was to implement a class representing matrix that provides a method to compute determinant.

Requirements

The following application has to be installed:

• CMake of version 3.10 (or higher)

How to install

git clone [email protected]:KetchuppOfficial/Matrix.git
cd Matrix

How to build

0) Make sure you are in the root directory of the project (i.e. Matrix/)

1) Build the project

cmake -B build
cmake --build build [--target <tgt>]

tgt can be unit_tests, test_generator, gauss or bareiss. The name of the first target is pretty much straightforward. The second target is a program that generates a matrix of a certain size with a certain determinant. Last two targets are drivers which input is a number N (the size of a matrix) and then N * N numbers (matrix itself) (on stdin). The output is the determinant (on stdout).

Last three targets are recommended to be built simultaneously as they are a part of the end-to-end testing system.

If --target option is omitted, all targets will be built.

2) Install executable files (optional)

You can install executable files in any directory you wish:

cmake --install build [--prefix <prefix>]

prefix can be any directory you want to install in.

If --prefix option is omitted, executable files will be installed in Matrix/bin/ directory.

How to generate tests

If you want to run some tests on the determinant calculation algorithm, go to Matrix/tests/end_to_end directory. There you will find a special script det.sh provided for such purpose.

Let N be the number of tests you want to run, S - the size of matrices, mD - maximal absolute value of the determinant, then command sequence:

./det.sh mode N S mD

generates N random matrices of size S with determinants in range [-mD; mD]. The determinants are saved in files answers_i (i in range [1; N]). After that this script runs determinant calculation algorithm on those tests and saves results in files result_i. Finally, both sets of values are compared. If values in any pair answer_i/result_i are different, then this test is considered "failed". It is considered "passed" otherwise.

mode argument has to be of one of 2 values only: gauss or bareiss. It specifies what algorithm will be tested.