How difficult is graduate statistical mechanics?

[AstroK]

Recently I have had a conversation with one of my professors, and he suggested me to take a graduate statistical mechanics course in the coming Spring semester. Although the various reasons my professor gave for his suggestion sound really appealing to me, I am a little bit worried about whether I can handle this graduate stat mech course along with my undergrad coursework. So I would like to ask how difficult a graduate course on statistical mechanics (mine is based on Huang) typically is? And on average how much time one has to put on studying for it per day / week?

I will also be taking classical mechanics (at the level of Talyor) and quantum mechanics (at the level of Griffiths) next Spring. I wonder if taking these two big physics subjects together with graduate stat mech is too much for someone who want a good academics / social life balance.

My background: I have just taken an intro thermo & stat mech course based on Schroeder this Fall. For the math, I have a solid background on calculus (multivariable), linear algebra, ODE's, a little bit of PDE's and perhaps slightly more.

I really want to make a careful decision on this. Thanks.

 

[ShayanJ]

In terms of having the prerequisites, you have all you need. But whether it will be difficult for you or not, is not something we here can say. It depends on your study habits and how sophisticated and well-versed you are at physics in general. But I should mention that solving graduate stat. mech. problems can be really time consuming.
Griffiths' is not a difficult text as far as I know it, but again its what I see and things may be different from your point of view. I'm not familiar with Taylor's though.
My suggestion is that you take some time to take a look at all the books you mentioned and see for yourself whether you can handle all of them in a semester or not.

 

[AstroK]

In terms of having the prerequisites, you have all you need. But whether it will be difficult for you or not, is not something we here can say. It depends on your study habits and how sophisticated and well-versed you are at physics in general. But I should mention that solving graduate stat. mech. problems can be really time consuming.
Griffiths' is not a difficult text as far as I know it, but again its what I see and things may be different from your point of view. I'm not familiar with Taylor's though.
My suggestion is that you take some time to take a look at all the books you mentioned and see for yourself whether you can handle all of them in a semester or not.

Thanks for the reply.

I understand. I did have a look at all the books I mentioned, but still I found it hard to judge whether I can handle them all in one semester. I of course enjoy reading all these subject while I am free from academic pressure, but from my own experience, things often just be completely different during the semester when all the quizzes, homework, exams and all kinds of things coming altogether at me, and it is often difficult for me to predict how bad it would actually be. That's why I try to ask for others' experience on the subject to see what I should expect.

Taylor's book is at the same level as Marion's, which is used typically for an upper division classical mechanics course.

 

[Dr. Courtney]

It's tough. When I did it (grad school) I only had one other physics course, and no other courses.

 

[mpresic]

Huang is a good book but it is tough. I think you will find Huang much harder than Taylor, Classical Mechanics, or Griffith, Quantum Mechanics.
I find in many grad schools, stat mech is taught last after grad mechanics, grad electrodynamics and grad quantum mechanics. I think this is sound placement.

Classical, electrodynamics, quantum are all subjects that treat few degrees of freedom. Stat mech is very different in that it treats many degrees of freedom. Many students who find the first three subjects approachable, can have a hard time when they take stat mech. Much of stat mech involves counting states quantum mechanically, this is quite different from classically (i.e. Maxwell Boltzmann). Even the math you mentioned while important does not overlap well with some of the math you will encounter (e.g. probability theory, statistics, steepest descents, which you can see in a good mathematical methods in physics course)

Maybe your professor knows the faculty who will teach the course and has some reason to believe it will be taught in a manner that a strong undergraduate can complete the course effectively.

I assume your professor thinks you can handle it or he would not have recommended it. Ordinarily, I do not think it is a good idea. If/When you go to graduate school, are they going to waive the requirement to take graduate stat mech. (Probably not).

 

[AstroK]

Thanks for the reply.

My professor is actually the one who teaches this graduate stat mech course. He knows me personally but not sure if he had overestimated my capability. For the graduate credit, I am not sure how it will work out if I go to another school for my graduate study, but if I stay in the same school, the stat mech requirement will indeed be waived, and this is actually one of the reason why he suggested me to take this challenge.

I had actually discussed with him about this difficulty / workload problem before, he was not as worried as I did, though he told me that I can simply audit the course once this year then it will be a piece of cake next year when I actually take it for credit. (And this will be what I would do if I decided not to take this course in the coming Spring)

 

[DrClaude]

It is hard for me to consider that you have the level for a graduate course in stat mech when you're not yet at the level of Griffiths in QM. I think you need a bit more "maturity" in physics before tackling Huang.

he told me that I can simply audit the course once this year then it will be a piece of cake next year when I actually take it for credit.

This sounds like the better option. Also, its sounds like you have a special relationship with that professor, the kind that can lead you to doing research work under him. In that case, it could be harmful to make a bad impression by getting a very low grade in a course he is teaching. I suggest you take your time and simply audit the course the first time around.

 

[dextercioby]

So what is the difference between an undergraduate and a graduate course in statistical mechanics? This American way of teaching is still confusing to me.

 

[ShayanJ]

So what is the difference between an undergraduate and a graduate course in statistical mechanics? This American way of teaching is still confusing to me.

The difference between Zemansky's and Huang's!
We have undergrad and grad statistical mechanics courses too!

 

[dextercioby]

OK, I'll try to look them up and see if I spot the difference between them on my own.

 

[ZapperZ]

Thanks for the reply.

My professor is actually the one who teaches this graduate stat mech course. He knows me personally but not sure if he had overestimated my capability. For the graduate credit, I am not sure how it will work out if I go to another school for my graduate study, but if I stay in the same school, the stat mech requirement will indeed be waived, and this is actually one of the reason why he suggested me to take this challenge.

I had actually discussed with him about this difficulty / workload problem before, he was not as worried as I did, though he told me that I can simply audit the course once this year then it will be a piece of cake next year when I actually take it for credit. (And this will be what I would do if I decided not to take this course in the coming Spring)

Don't you have an undergraduate advisor you can present this to? He/she should know more about you, more than we do. For example, nowhere in here did you describe how well you are doing in your studies. Are you doing well? Do you have high GPAs or equivalent? Are you struggling in your courses? Are you taking other classes as well as your physics classes? Have you ever taken more than 2 physics classes per semester/quarter? Can this course be delayed until after you've completed all your physics undergraduate requirements? Etc... etc. There's a lot of information that needs to be known before anyone can responsibly offer even a suggestion.

So if you want my advice, go to your academic advisor, sit down with him/her, and present the exact same thing that you're asking here. He/she should know more about you than we do, and may even be privvy to information that we do not have access to, such as the nature of the course and the workload when compared to what you will be taking.

Discussion of this nature (and we get A LOT of questions in this forum of this type) requires a 2-way communication. It isn't just one person saying I want to do this-and-that, and wait for answers. It often requires a back-and-forth discussion on what is best based on the situation. I've also often relied on body-language of a student to determine that what he/she said may not be what he/she meant, especially when there's hesitation in the responses. Something like this requires that you see someone in person.

Zz.

 

[Dr. Courtney]

It is hard for me to consider that you have the level for a graduate course in stat mech when you're not yet at the level of Griffiths in QM. I think you need a bit more "maturity" in physics before tackling Huang.

Unless you follow Zapperz's advice and are advised to take the course for credit by your undergrad adviser, I think the best idea above is to audit the course and work as many of the assigned problems as you can along the way. It's not sitting in the class that will make the material slow down when you later take it for credit (at your current school or another), it's working as many problems as you can. Sitting in class and reading the book are not where most learning occurs. Working the problems is where most learning occurs.

You've gotten very good advice.

 

[gleem]

Recently I have had a conversation with one of my professors, and he suggested me to take a graduate statistical mechanics course in the coming Spring semester

What was this conversation about that he would suggest that you take this course? By asking your question in this forum you seem to me to not have a particularly large amount of confidence which you should appreciate. We do not know your full story to really evaluate your situation goals, past performance. How does taking this course fit into your goals as a physicist? Is it worth risking excellent grades in CM and QM (which should be challenging in their own right) as you stretch your ability and time by including a course that is not required at this time, depends on CM and QM which you will be taking concurrently, is math intensive, and will have fellow students with more academic experience and maturity which raises the performance bar. Audit if you want.

 

[AstroK]

Thanks everyone. These are exactly the kinds of things I am looking for.

I definitely agree that I should give more details about the whole story / background behind all these. I was just trying to start off my question without giving too much at once, to avoid a "unreadably long post" and to get as much different opinion / viewpoint as possible. Because I believed that any kind of decision should not be made just based on one or two specific pieces of advice from others. That's why I try to keep a small conversation with people in here and through it hopefully I can figure out what I really want to do for myself.

Regarding my adviser, it is unfortunate that I afraid he won't know me much better than all of you would, except that he may have access to some numbers that I don't know. I had only talk with my adviser once and I afraid that he won't even remember who I am. But thank you for reminding me and I do agree that I should have someone to talk to face-to-face, I would certainly approach him ASAP regrading my question.

It is actually quite ironic that the one faculty / stuff who know me the most is exactly this graduate stat mech professor. I have been working with him for quite some time on a small undergrad research project (so formally I should call him my supervisor), and so far he seem to be very satisfied with my overall performance on this project. It is quite difficult to describe the nature of this research project, but in short it is about some "non-standard physics" (forgive me if this is not the right word to say it) like nonlinear dynamics, information thermodynamics, complex system etc, and I basically do some simple numeric works on the particular topic given by my supervisor and from there learn the subject, at least qualitatively. My supervisor is a very experienced one, both as a physics professor and an educator, and he always try to give some good advice for his student regarding their future studies and career. The conversation which he suggested me to take his graduate course is just one of the usual career advising he gives to let me know what options / opportunities are out there for me. But this kind of challenge he usually reserves for the best students he got (according to him). I know I did not sounds very confident, because I don't know whether I am really those who are capable of taking these challenges or I am just being overestimated. I do really want to give it a try but at the same time I can't afford to have these "unnecessary" challenges ruin my GPA, that's why I am so hesitated and hesitated enough to post my concerns on an internet forum.

About my performance in physics in general: Last semester I have taken 3 physics course (modern physics, E&M (first half of Griffiths), intro thermo & stat mech),a lab on modern physics, and also an intro CS course. I got A for all of them. It is worth to mention that, the intro thermo course is usually taken by seniors who have already done QM, CM, EM (I had skipped the prerequisite), and taking this course early is basically another suggestion my supervisor gave. Although it work out OK at the end, but the process is quite tiring. For the coming Spring, apart from the physics courses, I will also have to take a humanity and a language course.

Once again, thank you very much.