This repository provides an interactive shiny dashboard for creating and evaluating written single / multiple-choice exams and creating reusable exercise pools. The underlying workhorse is R/exams, in particular the so-called NOPS exams functions, which support both randomized single-choice and multiple-choice exercises along with automatic evaluation.
Rex is developed by Sebastian Bachler and is still work-in-progress.
For Rex to function properly, make sure the following resources are available on your system.
- libsodium-dev
- libmagick++-dev
- libpoppler-cpp-dev
- ghostscript
- texlive-full
Install all the necessary R packages including their dependencies (see section "PACKAGES" in app.R).
Start up via shiny::runApp(appDir = '<base directory containing app.R>', host = '0.0.0.0', port = 3838); or, alternatively, via RStudio by
opening and running app.R.
Build your image with the supplied docker file: docker build --platform linux/x86_64 -t rex ..
Run your image: docker run -p 3838:3838 rex.
Caution
Do not host this version of Rex with uncontrolled / public access - Code from user input is parsed and evaluated without any security measures!
Once you got the application running, authenticate using "rex" as both login and password.
The entire management of exercises takes place in the "EXERCISES" tab of the application.
When managing exercise, the following input fields can be edited:
- Name
- Author
- Type
- Points
- Section
- Question (accepts LaTeX code when "TeX mode" is switched on)
- Figure (accepts a single PNG file per exercise)
- Answers (consisting of the solution, answer text, and a solution note)
On the top right of the exercises tab, summary information on all exercises that are marked as examinable is shown.
To create exams, navigate to the "EXAM" tab and further to "Create exam". There you will find a form with various options to set.
Before creating an exam make sure that exercises in the exercise tab are marked as examinable first.
When clicking the "Create exam" button the exam will be created. Once this process is finished, a popup will show and offer to save the exam. When saving the exam, a ZIP archive will be downloaded. In this archive you will find the following files:
- All the exam scramblings as PDF files
- All the exam scramblings as HTML files including the solutions and solution notes
- A RDS file to evaluate the exam
- All of the exercises used in the exam
- A TXT file "input.txt" containing all the input values used for creating the exam
To evaluate exams, navigate to the "EXAM" tab and further to "Evaluate exam". There you will find another form with various options to set.
Before the evaluation of an exam is possible, the following three files need to be prepared:
- "Solutions": the supplied RDS file in the ZIP archive downloaded when the exam was created
- "Registered participants": a CSV file containing all the information of participants
- "Evaluation scans": all the evaluation sheet scans as either PDF files (all pages in the same orientation) and/or correctly oriented PNG files
When clicking the "Evaluate exam" button the exam will be evaluated. At first all the supplied scans are processed. When this is finished, a popup will show and allow to inspect and manually edit any of the processed scans. When proceeding, the evaluation is finalized. When also this is finished, again a popup will show and offer to save the exam evaluation. When saving the exam evaluation, a ZIP archive will be downloaded. In this archive you will find the following files:
- Two ZIP archives, one (ending with "_nops_scan.zip") containing all the scans converted to PNG files plus the extracted raw data and another (ending with "_nops_eval.zip") containing the evaluation documents of each participant.
- Two CSV files, one contsining the registered participants and another (ending with "_nops_eval.csv") containing all the evaluation data.
- The RDS file used to evaluate the exam
- A TXT file "statistics.txt" containing some basic evaluation statistics of the exam
- A TXT file "input.txt" containing all the input values used for evaluating the exam
Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.