matrix is a command line tool, allowing one to easily manipulate a matrix using several tools, invert, transpose & multiply. The tool is built ontop of my matrix.h library and this allows inversion and multiplication to utilise POSIX threads to increase performance.
The tool reads matricies from a file, and can output an inverted matrix to either a file, or the terminal.
Tool: Invert
This tool inverts an nxn matrix and outputs the resulting matrix either to stdout or a specified file (--output flag)
$ matrix -d | --dimension -i | --input [-o | --output] [-p | --parallel] [-h | --help] [-t invert | --tool invert]
Dimension: The dimension flag must be accompanied by an integer representing the dimension of the input nxn matrix
Input: The input flag, must be accompanied by a path to an inputfile holding a matrix
Output: The output flag must be accompanied by a file name to where the output will be written, if it is not provided, the matrix will be printed to the screen
Parallel: The parallel flag must must be accompanied by yes or no, informing matrix if it should use the POSIX threaded version of it's invert function
Help: This prints a help message for the invert tool\
Tool: The tool flag is optional if you wish to use the invert tool since invert is the default tool
Tool: Transpose
This tool transposes an nxm matrix and outputs the resulting matrix either to stdout or a specified file (--output flag)
$ matrix -x | --width -y | --height -i | --input -t transpose | --tool transpose [-o | --output] [-h | --help]
Width: The width flag must be accompanied by an integer representing the width of the input nxm matrix
Height: The height flag must be accompanied by an integer representing the height of the input nxm matrix
Input: The input flag, must be accompanied by a path to an inputfile holding a matrix
Output: The output flag must be accompanied by a file name to where the output will be written, if it is not provided, the matrix will be printed to the screen
Help: This prints a help message for the invert tool
Tool: The tool flag must specify that you wish to use the transpose tool, the default tool is invert
Tool: Multiply
This tool transposes an nxm matrix and outputs the resulting matrix either to stdout or a specified file (--output flag)
$ matrix -x | --width -y | --height -i1 | --input-one -i2 | --input-two -t multiply | --tool multiply [-p | --parallel] [-o | --output] [-h | --help]
Width: The width flag must be accompanied by an integer representing the width of the first, leftmost, input nxm matrix
Height: The height flag must be accompanied by an integer representing the height of the first, leftmost, input nxm matrix
Input-One: The input-one flag, must be accompanied by a path to an input file holding the first matrix to be multiplied
Input-Two: The input-two flag, must be accompanied by a path to an input file holding the second matrix to be multiplied
Parallel: The parallel flag must must be accompanied by yes or no, informing matrix if it should use the POSIX threaded version of it's multiply function
Output: The output flag must be accompanied by a file name to where the output will be written, if it is not provided, the matrix will be printed to the screen
Help: This prints a help message for the invert tool
Tool: The tool flag must specify that you wish to use the transpose tool, the default tool is invert
Tool: Generate
This tool generates an nxm matrix of numbers with in a given range of a certain datatype. The output can be written to a file or the stdout.
$ matrix -x | --width -y | --height [-d | --datatype] [-l | --lower] [-u | --upper] [-o | --output] [-h | --help] -t generate | --tool generate
-x, --width: The width, x, of the output x by y matrix. x must be an integer
-y, --height: The height, y, of the output x by y matrix. y must be an integer
-d, --datatype: Define if only integral or floating values should be returned (integer or float) (default integer)
-l, --lower: The lowest value with which to fill the matrix (default -100)
-u, --upper: The greatest value with which to fill the matrix (default 100)
-o, --output: The name of a file which the matrix will be written to (default stdout)
-h, --help: Display help message
-t, --tool: choose which tool to utilise, invert, transpose or multiply (default invert)
The input file represents a stream of numbers, which will be read, left to right, top to bottom into the matrix of given dimension (remembering that only square matricies are invertable). This means that the input file can be a list of space seperated numbers, tab seperated with newlines or any mixture.
e.g. space seperated 1 8 -9 7 5 0 1 0 4 4 0 0 1 2 5 0 0 0 1 -5 0 0 0 0 1
The tool uses the method of cofactor expansion to invert a matrix, this gives exponential time and limits the size of the input matrix to realistically less than 14x14. Internally, the numbers are represenetd as double precision, this leads to the all too common limitations when working with high precisions.
There are two ways to compile the project, both use the g++ compiler. There is included a Makefile with the project, a simple call to
$ make
should be enough to compile the project. However, for those who wish to edit the source code, a codebocks file is included as well, and can be used to compile, debug and edit the project. To test the projecct, a test 5x5 matrix is provied in the file "matrix.txt". Run:
$ ./matrix -d 5 -i matrix.txt --took invert
$ ./matrix -d 5 -i matrix.txt --parallel no -o matrix-output.txt
Remember, if your using code::blocks, copy matrix.txt into the directory of the binary. The appropriate output from these tests should be printeed to the terminal and written to the file matrix-output.txt:
| 1 | -8 | 9 | 7 | 17 |
| 0 | 1 | 0 | -4 | -24 |
| 0 | 0 | 1 | -2 | -15 |
| 0 | 0 | 0 | 1 | 5 |
| 0 | 0 | 0 | 0 | 1 |
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent