Research software

This software was written while creating research paper Software metrics in Boa large-scale software mining infrastructure: challenges and solutions.

Authors: Agnieszka Patalas, Michał Malinka, Wojciech Cichowski, Piotr Maćkowiak

under guidance of associate professor Lech Madeyski

Models and Metrics in Software Engineering Project

Wrocław University of Technology

To allow easy management of Boa jobs and connecting job outputs with creating defect prediction models, a simple Java program has been written. The software uses [Boa Java API](https://github.com/boalang/api) to run jobs.

Few simple documents (input files) should be used to configurate query. They are stored in res/ directory:

• config.properties - stores the login data for Boa account. The account is used for running Boa jobs through API. Both user and pass attributes must be real and belong to registered Boa account. They are mandatory. Not providing those attributes will result in program failure.

• input.xml - keeps the configuration for the program:

• • <metrics>: a list of metrics as a list of <metric> attributes; each metric should be stored in a *.boa file with the same name and placed in metrics/ dictionary. *.boa files must have only one output named m
• • <dataset>: the ID of the picked dataset; the IDs of working datasets are defined in datasets_dictionary.xml file;
• • <output_parameter>: defines the evaluating value for prediction model; so far, only the fixing revision attribute is used in such manner. It can be imported in the same way as <metric>
• • <result_file_path>: prediction output file name, relative to main directory

Program builds simple prediction model which can be used to predcit number of fixing revision in no-classifed data set.

Result file has following structure:

Columns: List of input metrics, eg. WMC, CNOF, RFC

Rows: Values of each metic

Last column named "fixingRevisions" contains value calculated by predicition model.

Statistical tests

statistical_tests_with_MATLAB folder contains statistical tests used for calculating importance of cumulative metrics described in the paper.