Replication package for the paper: "Exploring the Suitability of Source Code Metrics for Indicating Architectural Inconsistencies"
The following folders and files are part of this replication package:
data\
: Contains the raw metric data used in the study in the form of comma-separated-value files. There is one file for each case study system, e.g.jabref-data.csv
.models\
: Contains textual version and images of the modules of the systems and the mapping of the source code to the indended module structure, e.g.jabref-archmodel.txt
andjabref-model.png
.scripts\
: Contains R scripts that have been used to execute the statistical computations discussed in the paper. There is one file for each case study system, e.g.jabref-analysis.R
.
To reproduce the calculations, simply start your favourite R development environment and open the R scripts. The scripts assume that the repository is located on your R working path. More instructions can be found inside the scripts.
For convenience, the repository contains a Dockerfile which sets up RStudio in a docker container. The advantage of that is that you do not have to install the environment locally, but can access everthing through your brower. To run this repository inside a container, you need to have Docker installed on your system. If you do, cd
to the directory of the repository and do the following:
- build an image:
docker build -t arch-metrics-replication .
- run the image (RStudio will be available on port 8787):
docker run --rm -p 8787:8787 arch-metrics-replication
- open your browser at http://localhost:8787/ (Linux) or http://YOUR-DOCKER-MACHINE-IP:8787/
- RStudio's default credentials are username:
rstudio
, password:rstudio
- Open one of the R scripts in folder
arch-metrics-replication
, read the code and execute it