Release Notes
- based on the artefacts of this release, you can:
- execute the experiment on your own machine with our input data
- use your own data as input
- explore our results either manually or with IncrementalAnalysesHelpers (use Version 4.0)
File | Description |
---|---|
diffs-ordered-list-of-commits.txt | ordered commit hashes line by line containing all commits that were used for the evaluation |
diffs.zip | Diff files used as input for the analysis executions |
data-CONFIG.zip | data generated during execution with configuration CONFIG(=configurationblockchange/blockchange/artifactchange/baseline). During the baseline execution for commit 410, the wall clock jumped back an hour. We therefore include an adjusted version of the log-file for commit 410 where the timestamps do not jump an hour. |
investigation-cached-models.zip | ten executions of the non-incremental dead code analysis with cached models |
investigation-reextracted-models.zip | ten executions of the non-incremental dead code analysis with reextracted models |
evaluator-logs.zip | Logs from IncrementalAnalysesHelpers that include performance results as well as a consistency checks for the identified dead code blocks. For the performance, we provide two sets of results for each incremental configuration: one with (*-incl1.log ) and one without (*-excl1.log ) the first commit. |
linux-vm.7z | Virtual machine with prerequisites for running the evaluation installed. User: sse Password: sse |
evaluation-results.zip | Evaluation results containing the intensity of variability changes in the commits sequence (directory “comani-check”), the performance results and related calculations (directory “performance-check”), the correlation between execution times and number of blocks (directory “performance-commits-check”), and a general summary in “performance-evaluation-summary.xlsx” |