Mixer
=====

This tool should serve as generator of random permutations of training and testing data sets. It takes a *.raw file and corresponding *.txt annotation file as input and produces arbitrary number of pairs of training and testing data sets. 

Use:

mixer.exe -i inputFileName [-ratio 0.5] [-t count] [-a] [-x]

-i    		Specifies the files to process (without extension).
      		The program will add extensions as needed (.raw, .txt).
-a    		Will process annotation.
-x   		Will create dataset XML file.
-t    		Specifies how many file pairs to create.
-ratio    	Specifies the size ratio between training and testing sets.

Common use:
mixer.exe -i newDataset -ratio 0.75 -t 10 -a -x


Comments:
	Dev-C++ project included.
	Uses libxml2 which is also included