Technical Debt in the Peer-Review Documentation of R Packages: a rOpenSci Case Study

<strong>Replication Package for the paper "Technical Debt in the Peer-Review Documentation of R Packages: a rOpenSci Case Study" (MSR '21).</strong> <strong># Scripts: Data Collection and Processing</strong> These are the scripts used to extract the data from _rOpenSci_. The following steps indicate how to use them. 1. Add all attached R files into an R project. 2. Install the following R packages. Moreover, the process also requires to have a working GitHub account, in order to obtain the corresponding token. ```{r} library(dplyr) library(stringr) library(stringi) library(jsonlite) library(httpuv) library(httr) library(ggplot2) library(tidyr) ```' 3. All the individual functions on the following files should be sourced into the R Environment: `getToken.R`, `comments.R`, `issues.R`, and `tagging.R`. 4. Run the script located on the file `process.R`. This will run all the previous functions in the corresponding order. <strong># Datasets</strong> The following files are included: -Dataset_1-100_Author1.xlsx contains the randomly selected 100 comments that were classified according to TD types by Author 1. -Dataset_1-100_Author2.xlsx contains the randomly selected 100 comments that were classified according to TD types by Author 2 and the combined classification (in blue) after discussion. -Dataset_Phrases_Both.xlsx contains the randomly selected 358 comments (resulting in 602 phrases) that were classified according to TD types by both authors 1 and 2. Their classification was incorporated into a single spreadsheet side by side for easy comparison. Disagreement was discussed and final classification is in the “Agreement” field. -UserRoles.csv contains the user roles associated with the 600 phrases. The “comment_id” is the unique identifier for the comment from which the phrase is extracted. The phrase is represented in the “statement” field. The “agreement” field shows the final technical debt label after the analysis by two of the authors. The user roles are shown in the “user_role” column.**论文《R包同行评审文档中的技术债务（Technical Debt）：以rOpenSci为例》（MSR '21）复现套件** **# 脚本：数据采集与处理** 本套件包含用于从rOpenSci中提取数据的脚本，以下为使用流程： 1. 将所有附带的R文件添加至R项目中。 2. 安装以下R包。此外，本流程需使用有效的GitHub账户以获取对应令牌（Token）： r library(dplyr) library(stringr) library(stringi) library(jsonlite) library(httpuv) library(httr) library(ggplot2) library(tidyr) 3. 将以下文件中的所有自定义函数导入R运行环境：`getToken.R`、`comments.R`、`issues.R`及`tagging.R`。 4. 运行`process.R`文件中的脚本，将按对应顺序执行前述所有函数。 **# 数据集** 本次套件包含以下数据集文件： - `Dataset_1-100_Author1.xlsx`：包含随机选取的100条评论，已由作者1按技术债务类型完成分类。 - `Dataset_1-100_Author2.xlsx`：包含随机选取的100条评论，由作者2按技术债务类型完成分类，并包含经讨论后以蓝色标注的合并分类结果。 - `Dataset_Phrases_Both.xlsx`：包含随机选取的358条评论（共衍生出602个短语），由两位作者分别按技术债务类型完成分类。为便于对比，二者的分类结果已并排录入同一张电子表格中。分歧项经讨论后达成一致，最终分类结果存储于"Agreement"字段内。 - `UserRoles.csv`：包含与600个短语相关的用户角色信息。其中，"comment_id"为提取该短语的评论的唯一标识符；"statement"字段表示该短语的具体内容；"agreement"字段展示经两位作者分析后的最终技术债务标签；"user_role"列则列明了对应用户的角色。