/examgen

Primary LanguagePythonMIT LicenseMIT

Exam Generator

Generate exams using Markdown and Python.

Features

  • Write questions in Markdown
  • Generate PDF exams
  • Generate answer keys
  • Randomize questions and answers
  • Generate multiple randomized exams

Install

Using pip:

$ pip install examgen

Using the repository code:

$ python setup.py install

Creating an exam

First, you need to create an exam file. It is a python module. Then, create an exam.py file for example.

You need to set up five variables in the module: institution name, course name, professor name, exam title, and questions. Below there is an example of an exam file:

institution = "Federal University of Paraiba"
course = "Artificial Intelligence"
professor = "Yuri Malheiros"
exam = "2nd Exam"

questions = ["question1", "question2", "question3", "question4", "question5"]

The first four items are straightforward. You only need to put the text you want. For the questions, you must provide a list with the names of the question files. In the case above, the exam has five questions. The first question was created in the file question1.md, the second in the file question2.md and so on.

Let's check how to create a question file.

Creating a question

A question file follows a markdown-ish syntax. Below there is an example of a question file:

Which algorithm is used for unsupervised machine learning?

---

- KNN
- Linear regression
x K-means
- Decision tree
- Logistic regression

The file has two parts divided by the "---" line. The first part (above the ---) is the question text. You can write any markdown in that part. In the second part (below the ---), you put the answers. They must be created as a list, and the correct one has an "x" instead of "-".

Comments

Questions may have a comment section. You need to insert an extra "---" line after the answers. Below the "---", you can write any comment you want. Check the example below:

Which algorithm is used for unsupervised machine learning:

---

- KNN
- Linear regression
x K-means
- Decision tree
- Logistic regression

---

This is a comment.

Generating an exam

To generate an exam, use the Command Line Interface (CLI).

examgen -e EXAM_FILE

EXAM_FILE is the python file with the information about the exam.

The CLI supports other options:

  • -rq, --random-questions: randomize the questions' positions in the exam.
  • -ra, --random-answers: randomize the answers' positions in each question.
  • -b, --batch: generate multiple exams.
  • -q, --quantity INTEGER: set the number of exams in batch generation.
  • -m, --merge: merge all exams in one file in batch generation.
  • -fb, --front-and-back: use this option if you want to print the merged exams using the front and back of the sheets.

Note: ExamGen uses pandoc to generate PDF files from Markdown.

Question bank

Randomize question order sometimes is not the behavior we want. So, we provide an alternative way to mix questions in exams.

Using a question bank, you can create many questions and set which questions can appear as the first question, the second question, the third, etc.

To do this, you only need to set the questions in the exam file as follows:

institution = "Federal University of Paraiba"
course = "Artificial Intelligence"
professor = "Yuri Malheiros"
exam = "2nd Exam"

questions = [("q1v1", "q1v2"), ("q2v1", "q2v2, q2v3"), "q3", ("q4v1", "q4v2"), "q5"]

In the example above, the first question will be picked randomly between the questions in the files q1v1.md and q1v2.md. The second question will be q2v1.md, q2v2.md or q2v3.md. Notice that not all questions need multiple possibilities. In this example, the third question will always be the question in the file q3.md. The fourth question also has two possibilities, and the last question will always be q5.md. In summary, use a tuple to define the possible questions, and the exam generator will pick one randomly.

This feature is very useful for generating multiple random exams.