Teoretisk fysiks kursutväderingar


Resultat av: Computational physics, SI2530, ht2008

Status: Avslutad
Publicerad under: 2008-12-04 - 2009-01-07
Antal svar: 10
Procent av kursdeltagarna som svarat: 50%
Kontaktperson: Olle Edholm

What is your overall impression of the course?

10 svarande

Very positive1 10%
Positive6 60%
Neutral2 20%
Negative1 10%
Very negative0 0%

- It was more a lecture in statistical mechanics than in computational physics. We did the theory, than we skipped writing programmes (the computation part!) and just ran some pre-written simulations, and then we did the evaluation of the simulation. The most interesting part, i. e. how to turn a physical problem into a computer programme, was not really discussed. (Neutral)
- The lectures are not at all usefull, emphasis should be put on more practical things. (Negative)

The suggested course literature (Allen and Tildesley) was

10 svarande

Very good1 10%
Good3 30%
Average6 60%
Bad0 0%
Very bad0 0%

- quite a complete book (Good)
- Good book for MD- and "ok" for MC-simulations. Expensive though, it would have been good to provide some exemplars for the students. (Good)
- Did not use course literature since the ordering process took too long (and the book is still to be delivered..) (Average)
- I used it more as a reference book. (Average)

How were the lectures (Anatoly) ?

10 svarande

Very good0 0%
Good3 30%
Average2 20%
Poor5 50%
Very poor0 0%

- the lectures have not clear ooutline (Poor)
- The goal of lectures was not clear. No structured notes on the board. (Poor)
- No structure in the lectures at all, i had no idea about how to start a MD simulation. Bad Fortran teaching, very poor(!) notes on the board. (Poor)

How were the lectures (Mats) ?

10 svarande

Very good6 60%
Good4 40%
Average0 0%
Poor0 0%
Very poor0 0%

- Just well done. (Very good)
- See "Anatoly" but opposite. (Very good)

How were the lectures (Olle) ?

10 svarande

Very good0 0%
Good8 80%
Average2 20%
Poor0 0%
Very poor0 0%

- Important connections pointed out. Sometimes a bit unstructured notes on the board. (Good)
- A bit more teaching on how to implement a physical problem in Fortran would have been useful. Also practical things like why and how do we use reduced or different units were missing. Rest OK. (Good)

To follow the course, my prerequisites in physics have been ..

10 svarande

More than enough1 10%
Enough6 60%
Not enough3 30%

- I have a backgound in engineering, thats why I lack of knowledge in statistical mechanics, and modern physics (as necessary for simulating spin systems) (Not enough)

To follow the course, my prerequisites in computer science have been ..

10 svarande

More than enough3 30%
Enough4 40%
Not enough3 30%

- From a programming point of view, the course is not very demanding... (More than enough)
- I knew C-programming before and also worked with Matlab before. I never used Fortran though. (Enough)
- I did not study Fortran and Unix before (Not enough)

The 4 computer labs were

10 svarande

Very good0 0%
Good7 70%
Average1 10%
Bad2 20%
Very Bad0 0%

- The labs them selves were good, and helped develop skills in the subject but the computer resources available for them is not very satisfying. (Good)
- The topics covered a broad range (at least in my eyes) and gave a good possibility to get to know some computational physics. (Good)
- The first lab was the hardest one, which might not have been the intention? I suggest using the argon lab first. (Good)
- Veeeeeeery time consuming. Due to overlaps, I could never attend the full class. It took me lots of time to complete the tasks without a teaching assistant. Even though they were very helpful whenever I asked them in their offices. (Average)
- Ratchet and ising was good. Argon and espicially water/ice was very confusing. Not technically or theoreticaly difficult, just really unclear what were to do/what you expected on the reports. (Bad)
- The goal of these computer labs are not clear enough and it deals too much with physics and not at all with computation. (Bad)

Should we have graded marks (A,B,C,D,E) on the computer labs ?

10 svarande

Yes1 10%
No9 90%

- A mark cannot harm and gives a feeling how good the report is. Sometimes the comments were a bit short. (Yes)
- Sure you could have graded marks, but then you'd have to be much better at explaining what you expect from the reports. (No)
- they are normally here in order to give experience , pass or fail is sufficient. (No)
- They seemed hard to grade in a good way. (No)
- No way!!! (No)
- Hard to tell, because of everyones different background. Might have been unfair. I guess I would have suffered from that... Not sure... (No)

Was the examination project work good ? Comments ?

10 svarande

Yes8 80%
No2 20%

- That was the best part, since it included a little more programming etc. Also, projects is a good way to learn physics. (Yes)
- This was the best part of the course, and where I learned the most. (Yes)
- This was the most interesting part of this course. (Yes)
- I did my best, please give me A. (Yes)
- Hmmm, I am still working on it. That is definitely where I learned most about Fortran, for sure! Therefore: good and helpful. But: also very time consuming, more than any other lecture. I probably spent one fourth of all my time in the full semester(!) on computational physics. That is too much, and I am still not done. (Yes)
- Perhaps introduce two ("mandatory") meetings, in the beginning and nearing the end to discuss problems and check that the problem itself and difficulties are understood. (Yes)
- I am still on it and to my point of view it is too much time consuming, especially when you are not that good in program writing. (No)
- Too difficult for me (No)

The computer facilities at Filialen are

10 svarande

Very good0 0%
Good3 30%
Adequate5 50%
Poor0 0%
Very poor2 20%

- Although I personally did the labs on my own equipment. (Good)
- More extended simulations are difficult to perform, I think even my four year old laptop has more computing power. But there are always free PCs, so I could spent a lot of time. and fortunately it?s open 24-7. (Adequate)
- Personally I feel more attention should be taken for new students (Adequate)
- Terribly slow. During my project, I have been unable to conduct several computations that would greatly improve the quality of my data. One can of course remote log in and do things via ssh, but the operations are for some reason shut down after some time (despite using 'limit cputime xxx') This means that some computations are not possible to make at all at nadas/filialens computers. (Very poor)
- Doing the water/ice on these was ridiculous. (otherwise fine) (Very poor)

The amount of work that has to be done i this course is

10 svarande

To high2 20%
Reasonable8 80%
To low0 0%

- For reasons: see 10. (To high)
- Not easy work but one learns a lot. (Reasonable)

If one would reduce the number of computer labs or replace one of them eith a new one which should then be taken out?

10 svarande

1. Brownian ratchet2 20%
2. MD of Argon0 0%
3. Ising model0 0%
4. Melting of ice8 80%

- Learning how a specific computer package works without even looking at what it does or how it does it will not help me learn or understand anything about physics. (4. Melting of ice)
- Melting of ice was a bit boring compared to the others since one just used GROMACS (this is my choice if I was forced to pick one). (4. Melting of ice)
- Where?s the benefit from using a huge MD-package in a introduction course to computational physics? It is just "click here", than "type in that", "do this" and at the end you have a result which means this. Aha, but how did we get there? I have no idea! The labs should be structured, so that one starts by writing (or changing) a simple programme in the beginning and the go over to more complicated things which end in a bigger project such as our homework. (4. Melting of ice)
- Due to the complexity of the lab, I got the feeling many people didn't really understand what was happening and why. (4. Melting of ice)

Kursutvärderingssystem från

[Theoretical physics home page] [KTH home page]   webmaster