###Description of the scripts and diagrams used in the MASCOTS2013's paper###

Scripts and data for running experiments:

	* folder experimentScripts contains scripts which are used to run experiments
		* runExperiment.py: python script for running expermint.py script with following configurations:
			- choosing between zero, one or more leak in one run of the experiments 
			(single link vs multiple leaks) and strength of the leaks
			- different number of iterations of map-reduce jobs in k-mean for controlling length of tests
			- instrumented or not-instrumented run
			- profiling or noProfiling run
		* experiment.py: main script for running experiments which calls other scripts when it is needed
		* constants.py: define constants for using in other codes
	* folder leakConf includes different leak properties with different strength used in the experiments

Which library is needed to re-produce the results?

	* Mahout library: k-means clustering (integration tests)


What data is used to produce which diagram?

	Figure 1. absolute rank vs. defect identifier: absolute rank for root causes of class a based on integration testing
		input: ranking-result-type-a.csv
		output: leak-type-a-leaks-identifier.pdf
	Figure 2. absolute rank vs. strength difference factor: absolute rank for root causes of class b based on integration testing
		input: ranking-result-type-b.csv
		output: leak-type-b-strength.pdf
	Figure 3. absolute rank vs. different versions: absolute rank for root causes of class b based on integration testing
		input: ranking-result-type-b.csv
		output: leak-type-b-span.pdf
