Mini projects#
Introduction#
This mini-project aims to allow the students to practice their programming skills in Python in a project of their choice. A practice session will be dedicated to these projects but you can already think about it and even start to work.
Description#
The choice of the topic is free as long as the student hands a personal job which relies on scientific programming. If possible, this mini-project can be coupled with projects from other classes followed by the student in this semester.
Discuss about your topic!
Students are advised to discuss together and with one of the teachers about their topic before starting. You can open a dedicated issue in https://gricad-gitlab.univ-grenoble-alpes.fr/meige-legi/scientific-computing-m2-efm.
Workload
An indicative personal workload for the mini-project is about 6 hours in addition of the dedicated practice session. This information should help the student to choose a topic which will not require an unreasonable time investment.
Programming languages
We will study Python during this course so it will be easier for you to use Python. However, you are free to use any open-source tools. The tool has to be open-source because it has to be possible for the teacher to try your code. Note that if you choose another language than Python, we won’t be able to help you about language related problems.
Evaluation#
The students have to put the code in a Gitlab project hosted on
https://gricad-gitlab.univ-grenoble-alpes.fr. To ease the evaluation, we ask you
to use a particular name: miniproject-scientific-computing2022 (so that the full
address is
https://gricad-gitlab.univ-grenoble-alpes.fr/your-username/miniproject-scientific-computing2022).
You will have to post the address of the Gitlab project of your mini-project in a message in the issue #27.
Then, the evaluation of these mini-projects will be done through a short oral sessions (~ 10 minutes) where the students will be asked to present their work. It will corresponds to half of the practical session grade.
Examples of mini-projects#
Please do not take a machine learning problem! There is a specific course in your Master 2 on this subject and there are many other things to do in scientific computing.
Remember that interesting and graphical representations are nice to show. Of course, if you find something useful (for you or for others), it would be appreciated. You can also look at examples of nice software projects to help you to choose a subject and a set of technologies. Here is a small list of Python softare projects that can be interesting to use: matplotlib, pandas, ipyleaflet/voila, pyscript, BeautifulSoup, textual…
You can also focus on performance and try to use tools for high performance with Python like PyPy (an alternative Python interpreter that can be installed with mamba) or Transonic/Pythran/Numba. In this case, you should present some benchmarks with nice figures.
Important
You can try to look for subjects and set of technologies by searching good keywords on the web. If you want to work on a subject, try to find a good set of software projects to work on it. Use your favorite web search engine!
Important
If you work on a program, it’s much better if it can be installed by other people. Read about Python packaging!
Monte Carlo Integration#
Read about the Monte Carlo integration numerical method (e.g. https://en.wikipedia.org/wiki/Monte_Carlo_integration). In a similar way to what will be done in the practice session, investigate and illustrate pros and cons of this method. How does it compare to other methods that you know about ? In which situation can it be useful ?
Animation of simulations of particles in a box#
Use Matplotlib to create an (interactive) animation of colliding particles in a box. An extension to this project could be to link this animation to the ideal gas law.