Introduction

This is the repo to reproduce the results in Fu et al. ``Revisit Unsupervised Effort-Aware Just-in-Time Defect prediction'' . The package is modified basd on the code provoided by Yang et al.

`` Effort-aware just-in-time defect prediction: Simple unsupervised models could be better than supervised models.''

Dependence

• R Environment
• RWeka, car, usdm, effsize, ScottKnott packages are required. Please run the following commands in R.

	install.packages("RWeka")
	install.packages("car")
	install.packages("usdm")
	install.packages("effsize")
	install.packages("ScottKnott")

Run

 source("exeMain.r")
 

Reference

@inproceedings{fu2017unsupervised,
 author = {Fu, Wei and Menzies, Tim},
 title = {Revisiting Unsupervised Learning for Defect Prediction},
 booktitle = {Proceedings of the 2017 11th Joint Meeting on Foundations of Software Engineering},
 series = {ESEC/FSE 2017},
 year = {2017},
 pages = {72--83},
 numpages = {12},
 publisher = {ACM}
}

Folder Structure

.
├── LICENSE
├── README.md
└── jit
    ├── input   ## input data
    │   ├── bugzilla.csv
    │   ├── columba.csv
    │   ├── jdt.csv
    │   ├── mozilla.csv
    │   ├── platform.csv
    │   └── postgres.csv
    ├── output  ## reuslts are saved here accordingly
    │   ├── cross-project
    │   ├── cross-validation
    │   └── cross-validation-timewise
    └── script_r
        ├── ReportResults.r
        ├── core.r
        ├── evaluate.r 
        ├── exeMain.r  ## the start point of R code
        ├── learner.r  ## learners are called here
        ├── packages
        │   ├── EMImputation1.0.1.zip
        │   ├── RBFNetwork1.0.8.zip
        │   ├── citationKNN1.0.1.zip
        │   ├── conjunctiveRule1.0.4.zip
        │   ├── gaussianProcesses1.0.1.zip
        │   ├── grading1.0.1.zip
        │   ├── gridSearch1.0.7.zip
        │   ├── gridSearch1.0.8.zip
        │   ├── gridSearch1.0.9.zip
        │   ├── isotonicRegression1.0.1.zip
        │   ├── kernelLogisticRegression1.0.0.zip
        │   ├── lazyBayesianRules1.0.1.zip
        │   ├── leastMedSquared1.0.1.zip
        │   ├── linearForwardSelection1.0.1.zip
        │   ├── multiBoostAB1.0.1.zip
        │   ├── paceRegression1.0.1.zip
        │   ├── partialLeastSquares1.0.4.zip
        │   ├── racedIncrementalLogitBoost1.0.1.zip
        │   ├── ridor1.0.1.zip
        │   ├── rotationForest1.0.2.zip
        │   ├── stackingC1.0.1.zip
        │   └── userClassifier1.0.3.zip
        ├── results
        └── utils.R