Skip to content
📦 Replication Package for the ICST 2020 Paper Presenting the STICCER Method for Reducing Automated Schema Test Suites
Branch: master
Clone or download

Latest commit

Fetching latest commit…
Cannot retrieve the latest commit at this time.


Type Name Latest commit message Commit time
Failed to load latest commit information.

STICCER Data and R Analysis Scripts

This repository contains experimental results of the STICCER technique and its R scripts. The content of this repository are the following:

  • R functions for reading data sets, performing data manipulations, statistically analyzing the results, and generate tables.
  • Our experiment's data.
  • Running the scripts on new experimental results that equal to ours.
  • Tests


Please note that the following instructions have been tested on an Ubuntu 18.04 LTS workstation running the following version of R:

R version 3.4.4 (2018-03-15) -- "Someone to Lean On"
Copyright (C) 2018 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)

You would need the following packages:

  • dplyr 0.8.3
  • psych 1.8.1
  • kableExtra 1.1.0
  • knitr 1.23
  • stringr 1.4.0
  • purrr 0.3.2
  • tidyr 1.0.0
  • reshape2 1.4.3
  • ggplot2 3.2.0
  • readr 1.3.1
  • testthat

Running the scripts

In your terminal and the directory of the repo run the command R to enter the R session.

The scripts are ready to run the original experimental results.

To generate the tables in the paper execute the R/tables.R script (i.e., source("R/tables.R") in the R session). The will output latex tables for you to analyze in the texTables directory and the merges plots in the plots directory.

To get the data frames for further analysis, in the R session, execute the R/main.R script (i.e., source("R/main.R") in the R session) and you will the mutationanalysis data frame.

If you have new data that was generated by SchemaAnalyst reduction techniques copy the results directory that was created by SchemaAnalyst to this project. You would then need to change line results_path <- "experiment_results/" to results_path <- "results/" in the main.R file. Then re-run the tables.R script again to generate the tables and plots.

Running Tests

In the R session, run the source("R/main.R") first. Then, run this command testthat::test_file("tests/test_functions.R").

To create or modify tests, please place it in the tests directory.


You can’t perform that action at this time.