iCFTL-Diagnostics dataset

This dataset accompanies the paper "Diagnosing Violations of Time-based Properties Captured in iCFTL", by Cristina Stratan, Joshua Heneage Dawes, and Domenico Bianculli, to appear in the proceedings of FormaliSE 2024. Together with the dataset, we provide the source code (iCFTL-Diagnostics-tool.zip) of the tool developed as part of the contribution of this paper. iCFTL-Diagnostics tool is also available at https://github.com/SNTSVV/icftl-diagnostics. The instructions on usage can be found in the README both in the zip file and GitHub).The dataset.zip has three folders:rq1 for research question RQ1rq2 for research question RQ2rq3 for research question RQ3In folder rq1 we have rq1.csv file that contains the data used to answer RQ1 and it has the following columns:project - for the project full nameexpected PNR, for the line number expected to be returned by the diagnostics approachdetected PNR, for the line number of the PNR detected by our diagnostics approachcorrect PNR, Boolean values to show if expected PNR coincides with detected PNRIn folder rq2 we have time-men-diagn.csv and time-men-diagn2.csv files that contain the data used to build Table 2: Average (over 5 runs) time and memory consumed by the diagnostics approach. Both time-men-diagn.csv and time-men-diagn2.csv have the columns:project - for the project full nametime-mean - for the mean time calculated over the 5 runs of the projecttime_std - for the standard deviation of the time takenmemory-mean - for the mean memory calculated over the 5 runs of the projectmemory_std - for the standard deviation of the memory consumedID - for the project IDbinding-count - for the number of binding the project hasavg-trace-length - for the average trace lengthFinally, in rq3 we have the rq3.csv file that contain data used to build Table 3: Mean of time and memory consumed by different instrumentation schemes. rq3.csv has the columns:project - for the project namemean-n-t - for mean time calculated over 5 runs of the non-instrumented programstd-n-t - standard deviation of the time taken over 5 runs of the non-instrumented programmean-v-t - for mean time calculated over 5 runs of the vanilla instrumented programstd-v-t - standard deviation of the time taken over 5 runs of the vanilla instrumented programmean-d-t - for mean time calculated over 5 runs of diagnostics instrumented programstd-d-t - standard deviation of the time taken over 5 runs of the diagnostics instrumented programID-t - for the project IDmean-n-m - for mean memory calculated over 5 runs of non-instrumented programstd-n-m - standard deviation of the memory consumed over 5 runs of the non-instrumented programmean-v-m - for mean memory calculated over 5 runs of vanilla instrumented programstd-v-m - standard deviation of the memory consumed over 5 runs of the vanilla instrumented programmean-d-m - for mean memory calculated over 5 runs of diagnostics instrumented programstd-d-m - standard deviation of the memory consumed over 5 runs of the diagnostics instrumented programID-m - for the project IDt-d-v - for approximate percentage of time taken by the diagnostics instrumented program over vanilla instrumented program (in percentage)m-d-v - for approximate percentage of memory consumed by the diagnostics instrumented program over vanilla instrumented program (in percentage)t-d-n-raw - for the time the diagnostics instrumented program took over non-instrumented program (in percentage)t-d-v-raw - for the time the diagnostics instrumented program took over vanilla instrumented program (in percentage)m-d-n-raw - for memory the diagnostics instrumented program took over non-instrumented program (in percentage)m-d-v-raw - for memory the diagnostics instrumented program took over vanilla instrumented program (in percentage)本数据集配套于Cristina Stratan、Joshua Heneage Dawes与Domenico Bianculli合著的论文《诊断iCFTL中捕获的基于时间属性的违规问题（Diagnosing Violations of Time-based Properties Captured in iCFTL）》，该论文将发表于FormaliSE 2024会议论文集。 伴随本数据集一同提供的还有本文研究贡献中开发的工具源代码（压缩包名为iCFTL-Diagnostics-tool.zip）。iCFTL诊断工具同时可通过https://github.com/SNTSVV/icftl-diagnostics获取。工具使用说明可在该压缩包与GitHub仓库的README文件中查阅。 数据集压缩包dataset.zip包含三个文件夹，分别为对应研究问题RQ1的rq1文件夹、对应RQ2的rq2文件夹，以及对应RQ3的rq3文件夹。 在rq1文件夹内，存有rq1.csv文件，其中存储了用于解答研究问题RQ1的相关数据，该文件包含以下列： - project：项目完整名称 - expected PNR：诊断方法预期返回的行号 - detected PNR：诊断方法检测到的行号 - correct PNR：布尔值，用于标识预期行号与检测到的行号是否一致 在rq2文件夹内，存有time-men-diagn.csv与time-men-diagn2.csv两个文件，存储了用于构建表2的数据：诊断方法的平均运行时间（基于5次运行取平均值）与内存消耗。两个CSV文件均包含以下列： - project：项目完整名称 - time-mean：该项目5次运行的平均耗时 - time_std：该项目5次运行耗时的标准差 - memory-mean：该项目5次运行的平均内存消耗 - memory_std：该项目5次运行内存消耗的标准差 - ID：项目ID - binding-count：项目的绑定数量 - avg-trace-length：平均跟踪长度 在rq3文件夹内，存有rq3.csv文件，存储了用于构建表3的数据：不同插桩方案下的平均耗时与内存消耗。rq3.csv文件包含以下列： - project：项目名称 - mean-n-t：未插桩程序5次运行的平均耗时 - std-n-t：未插桩程序5次运行耗时的标准差 - mean-v-t：标准插桩程序5次运行的平均耗时 - std-v-t：标准插桩程序5次运行耗时的标准差 - mean-d-t：诊断插桩程序5次运行的平均耗时 - std-d-t：诊断插桩程序5次运行耗时的标准差 - ID-t：项目ID（耗时相关） - mean-n-m：未插桩程序5次运行的平均内存消耗 - std-n-m：未插桩程序5次运行内存消耗的标准差 - mean-v-m：标准插桩程序5次运行的平均内存消耗 - std-v-m：标准插桩程序5次运行内存消耗的标准差 - mean-d-m：诊断插桩程序5次运行的平均内存消耗 - std-d-m：诊断插桩程序5次运行内存消耗的标准差 - ID-m：项目ID（内存相关） - t-d-v：诊断插桩程序相较于标准插桩程序的耗时占比（百分比形式） - m-d-v：诊断插桩程序相较于标准插桩程序的内存消耗占比（百分比形式） - t-d-n-raw：诊断插桩程序相较于未插桩程序的耗时占比（百分比形式） - t-d-v-raw：诊断插桩程序相较于标准插桩程序的耗时占比（原始百分比形式） - m-d-n-raw：诊断插桩程序相较于未插桩程序的内存消耗占比（百分比形式） - m-d-v-raw：诊断插桩程序相较于标准插桩程序的内存消耗占比（原始百分比形式）