Skip to contents

RUBer

Overview

RUBer is an R package created for parameterized reporting according to the requirements at the Ruhr-Universität Bochum (RUB). So far, the package was primarily used for the eigth (2018-19) and ninth (2019-21) reporting cycle for teaching (Lehrberichterstattung). The package provides a large example dataset containing fake values, an R Markdown template in the RUB corporate design, preconfigured ggplot2 plotting functions, custom themes for ggplot2 and flextable, as well as functions making the RUB corporate design colors available.

Main features

  • An R Markdown template for use with rmarkdown::draft. The template uses knitr::knit_expand to dynamically create code chunks containing figures, captions and subcaptions based on the data frame passed on as a parameter.
  • The packages officedown and officer enable the automatic numbering of figures and tables, the insertion of corresponding table of contents, as well as various other helpers and post-processing features.
  • Vectorized figures in Microsoft’s Enhanced Metafile format using the devEMF package.
  • Five preconfigured ggplot2 plotting functions.
  • A custom ggplot2 theme, theme_rub.
  • A custom flextable theme.
  • Word reference document in the RUB corporate design for use with the R Markdown template.
  • Functions to access the RUB color palette.
  • Scale functions to make the RUB color palettes available in ggplot2.
  • An example dataset, df_example, that illustrates how we used this package in 2022 to create 67 reports with a total of over 5,000 unique figures. All values in the dataset are fake and algorithmically generated, the other columns, though, are mostly identical to the confidential production dataset.

Installation

Install the development version from GitHub with:

# If package "remotes" is not installed, install it first:
# install.packages("remotes")

# Install RUBer from Github
remotes::install_github(
  repo = "RichardMeyer-Eppler/RUBer",
  build_vignettes = TRUE
  )

Usage 

library(RUBer)

# RUBer comes with two example data sets with generated data

# df_report contains meta data about each of the 68 reports
df_report
#> # A tibble: 68 x 6
#>   report_nr report_type_id report_title        report_author file_name subfolder
#>       <int> <chr>          <glue>              <chr>         <glue>    <chr>    
#> 1         1 STG            RUBer Example Repo~ Example 01    RUBer_Ex~ Geistesw~
#> 2         2 STG            RUBer Example Repo~ Example 02, ~ RUBer_Ex~ Geistesw~
#> 3         3 STG            RUBer Example Repo~ Example 04, ~ RUBer_Ex~ Geistesw~
#> 4         4 STG            RUBer Example Repo~ Example 08, ~ RUBer_Ex~ Geistesw~
#> 5         5 STG            RUBer Example Repo~ Example 14, ~ RUBer_Ex~ Geistesw~
#> # ... with 63 more rows

# df_example contains the data required to dynamically create all figures
df_example
#> # A tibble: 164,794 x 24
#>   report_nr figure_nr report_type_id x     x_label    y     y_axis_label    fill
#>       <int>     <int> <chr>          <chr> <chr>      <chr> <chr>          <dbl>
#> 1         1         1 STG            20152 WiSe 15/16 526   Studienfälle ~     2
#> 2         1         1 STG            20152 WiSe 15/16 616   Studienfälle ~    20
#> 3         1         1 STG            20152 WiSe 15/16 170   Studienfälle ~    21
#> 4         1         1 STG            20162 WiSe 16/17 520   Studienfälle ~     2
#> 5         1         1 STG            20162 WiSe 16/17 576   Studienfälle ~    20
#> # ... with 164,789 more rows, and 16 more variables: fill_label <chr>,
#> #   facet <chr>, group <dbl>, group_label <chr>, source_caption <chr>,
#> #   question_txt <chr>, figure_type_id <int>, figure_caption <glue>,
#> #   heading <chr>, subheading <chr>, is_heading <lgl>, is_subheading <lgl>,
#> #   report_author <chr>, report_title <glue>, file_name <glue>,
#> #   figure_height <dbl>

# This generates a Word report using df_report and df_example
render_report(quiet = TRUE)
#> i Report "C:/Users/Richard/AppData/Local/Temp/Rtmpa4Jpt6/
#> RUBer_report_1c9429401b2b.docx" was written successfully.

Guiding design principles

  1. Output format for the paramterized reporting is Microsoft Word, so that the reports can be shared and edited as widely as possible.
  2. The generated reports adhere to the RUB corporate design and are ready for printing as is, with no further adjustments to layout or content required.
  3. As far as possible, the data manipulation is done independently of all the reporting steps. The idea is that the data passed to the R Markdown template can come from any source, not just R. While the data wrangling for the eigth (2018-19) and ninth (2019-21) reporting cycle for teaching was done in R, this is not a requirement. If there is a need to create or alter datasets using Microsoft Excel, it is possible to use these datasets as basis for all reporting functions of RUBer.

Project background

Every three years, the teaching and program quality of all degree programs at the University of Bochum is assessed (Lehrberichterstattung). The university administration assists this process by providing data reports that combine descriptive higher education statistics with survey data. The University of Bochum regularly collects survey data in the study entry phase (Studieneingangsbefragung), mid-study (Studienverlaufsbefragung) and for its graduates (AbsolventInnenbefragung, collected through the nationwide Graduate Survey Cooperation Project (KOAB)). Previously, for the data reports, these rich data sources were analyzed and visualized by hand. RUBer automates this process, creating large numbers of print-ready reports in Microsoft Word.