🤔 Frequently Asked Questions 🤔
Who are you?
I am Jonatan Rasmussen, a danish student studying Human-Centered AI (MSc) at DTU. This website is my hobby project. I am not paid by, or affiliated with, DTU's administration.
What is this site?
This website contains public course data for DTU's courses. It also offers more in-depth search filters than the official DTU websites as well as evaluation summaries and overviews.
How do I use this website?
In the 'Home' tab, use the filter to search for whichever courses you are interested in. Click on any Course Card to see in-depth data for that specific course.
Why did you make this?
I started the project back in 2019 because I liked the DTUCourseAnalyzer Chrome extension (I am NOT its author) but its data was out of date (yes, even back in 2019 it used outdated data). Back then I was also new to programming, so it was a fun personal project. Initially, I just wanted to collect all the data in a big spreadsheet. Over time however, I collected more and more data and I wanted to make it browsable via a website. It is my goal that people can use this site to find and discover high-quality courses.
How long did it take to make this project?
200 hours would be my rough estimate. I have written 9.000 lines of Python, 6.000 lines of HTML and 5.000 lines of C#, totalling 20.000 lines of code (LOC). This is all on my GitHub.
Is the data up to date?
The most recent data is from the Summer Exams 2024. So yes, it is quite up-to-date. Fetching new data from DTU's websites is something I do via scripts that I manually need to run. So expect new data to be added to this website 1-2 times per year.
How did you get the data?
I use this site to scrape all the course numbers. Then I use a Python script to go to https://kurser.dtu.dk/course/01001, https://kurser.dtu.dk/course/01002, and so on to scrape course data for all the 1700+ DTU courses. Note that I am not using an API, nor have I had any contact with DTU's administration.
Can I see your code / your GitHub?
Here you go: [GitHub repository link]. Please read the README on the GitHub page before diving into the code.
Can I get your raw data file(s)?
You can find the data on my GitHub. From here, it is possible to download all the data as one huge CSV: CSV Files.
You seem to be missing data for some courses?
That should generally not be the case, as I am confident in the completeness and correctness of my scraped data. I'm not aware of any missing data myself, but do keep in mind that only courses available during 2023-2024 (soon 2024-2025) are available. Old courses that are no longer offered at DTU are not included in my data. Brand new courses might also not be available yet. I also might be missing data for the most recent exams, as I only update my database a few times per year. If the course has changed its course number, my website is unable to detect data belonging to the old course number. In very rare cases, grades for a specific semester might be missing due to getting mistaken as a re-exam. This only happens if a semester has very few students compared to the other semesters.
How do you calculate the 5-star ratings?
Each course has public evaluation data for each semester. The evaluation includes the following questions: "I have learned a lot", "The learning activities motivate me", and "I've had the opportunity to get feedback on my performance". For each question, you can answer "Strongly disagree", "Disagree", "Neutral", "Agree", or "Strongly Agree". I map these answers to the numbers 1-5, and then calculate an average. These three evaluation questions each count for 1/3 (â…“) of the total rating for the course. This is a made-up ranking system that is not endorsed by DTU. Additionally, it should be noted that a given course rating is NOT affected by the rating of other courses. I do not calculate course percentiles. For example, 3 stars simply indicate that the average evaluation rating is in the interval [2.75, 3.25]. 3.5 stars indicate a rating in the interval [3.25, 3.75] and so on. The color of the rating is hardcoded to match the number of stars, with 4.0 being green, 3.5 being yellow and 3.0 being orange. In my opinion, courses that fail to average positive evaluation results (4.0 or higher) have room for improvement. And any course that fails to average neutral evaluation results (3.0 or higher) does not belong at an "Elite" university.
How do you calculate the 5-skull workload?
Grade statistics does NOT influence workload score. Instead I use the evaluation question "5 ECTS credits correspond to 9 hours per week for the 13-week period (45 hours per week for the 3-week period). The time I have spent on this course is: (Much less, Less, Equal, More, Much more)". I map these answers to the numbers 1-5, with 1 being low workload and 5 being high workload. "More" and "Much more" are considered "Overworked student" and "Severely overworked student", respectively. Just like the 5-star ratings, this is a made-up ranking system that is not endorsed by DTU.
What are your future plans for this website?
I don't really know honestly. If this website starts to become popular I want to host it somewhere better than pythonanywhere.com. I don't have a lot of time these days, so don't expect a lot of new features. I will however continue to update it with new data for as long as possible.
Can you make a similar website for KU / AAU / SDU / RUC / (insert other university)?
Sadly this is not going to happen. The course data is completely different for non-DTU courses. I would have to re-write all of my code.
I have found a bug!
That is not a question. Jokes aside, feel free to leave a comment on my announcement post on Reddit or report it on my GitHub (I'm kinda new to running an open-source project however, so don't expect too much).
How can I contact you?
Feel free to leave a comment on my announcement post on Reddit. Alternatively, contact me on my personal email jonatan.j.k.rasmussen@gmail.dk for business-related stuff.
Give me a fun fact!
I'm the creator of the YouTube video Best of Mat 1, currently sitting at 12,000+ views.
Show me something cool!
Sure, take a look at the subpage /wip for a behind-the-scenes sneak peek of the website design process!