DroidFax Usage

Characterization results analysis

After installing the toolkit and performing your characterization study using it by following the DroidFax Installation and Usage guide, you can get both the raw data and the final results in graphical and tabular format. In addition, you can use the following scripts to do detailed analysis and/or inspection. These R scripts are used for computing statistics and producing figures and tabular results of raw characterization.

1 general metrics statistics

• callercalleeRanking.py - rank callers and callees by their out/in degress in the dynamic call graph
• compcov.py - compute component level coverage
• compdist.R - compute component distribution
• covstat.py - compute coverages at class and method levels
• edgefreqRanking-cdf.R - call frequency ranking plotted using CDF (cumulative distribution function)
• edgefreqRanking-scatter.R - call frequency ranking plotted using stacked scatter plots
• gdistcov-combine.R - compute execution composition in the unique call view, combining method and class level in one figure
• gdistcovIns-combine.R - compute execution composition in the call instance view, combining method and class level in one figure
• gdistcovIns.R - compute execution composition in the unique call view, producing method and class level in separate figures
• gdistcov.R - compute execution composition in the call instance view, producing method and class level in separate figures
• callback.R - calculate callback usage
• eventHandler.R - plot event handler categorization with percentage distribution
• eventHandler-tab.R - tabulate event handler categorization with percentage distribution
• lifecycleMethod.R - plot lifecyle method categorization with percentage distribution
• lifecycleMethod-tab.R - tabulate lifecyle method categorization with percentage distribution

2 ICC metrics statistics

for ICCs from single-app traces:

• gicc.R - compute ICC categorization producing separate figures
• gicc-combine.R - compute ICC categorization combining all plots in one figure
• iccdataextras-combine.R - compute ICC data carriage combining all plots in one figure
• iccdataextras.R - compute ICC data carriage producing separate figures

for ICCs from inter-app traces:

• ginterIcc.R - compute ICC categorization producing separate figures
• interIccDataExtras.R - compute ICC data carriage producing separate figures

3 security metrics and statistics

• srcsink.R - calculate source and sink usage and reachable taint flow (at method level) on the dynamic call graph
• src.R - plot source categorization with percentage distribution
• src-tab.R - tabulate source categorization with percentage distribution
• sink.R - plot sink categorization with percentage distribution
• sink-tab.R - tabulate sink categorization with percentage distribution