or the left unsolved problems in General Relativity are so hard that only mathematicians have sufficient math knowledge to work on them while physicists should resort to numerical methods for the solutions?
The boundary between "being a mathematician" and "being a physicist" isn't that black and white if you're talking about theoretical physics. Presumably if you're working at a research level in mathematical physics you need to be a pretty proficient mathematician. Maybe not with the same volume of knowledge in certain topics but certainly in the topics that are relevant to your field you would need to know enough mathematics to develop new things
usually the kind of questions that can be answered by numerical methods are very different from the kind of questions that can be answered by more abstract methods
but there are "mathematicians" that research numerical methods and there are "physicists" that do "abstract" work, math/physics is not a very useful distinction in this area and often is more about the rigor of the approaches than the kind of questions asked
(im new to this) Basically they are coupling a hamiltonian system with a finite number of degrees of freedom ( so say in $R^6$ 3 spacial positions and 3 velocity positions) with what they call a heat resevoir
while one might consider it poor pedagogy, many texts simply don't justify their ansatz for something, but just show that this ansatz does what they want
I don't know anything about this topic in particular, so I can't say whether there might be a better motivation or not
where (I suppose) we think of the heat reservoir as infinitely number of smaller particles and since we cant describe all of their positions we use a distribution. Which they start with as the solution to wave equation
Im very confused.
no worries @ACuriousMind
Hold on is it just that (12) above is the density of the velocity of a brownian particle + position = velocity dt .....
@ACuriousMind a professor in general relativity told me the proof of Penrose's inequality is very difficult and only very few special cases have been done.
he said a lot of physics problems are too difficult to solve by analytical methods and he used to do research by analytical methods but has changed to numerical approaches some years ago.
but he said:"if you are a mathematician, you can do analytical research."
@DanielSank assuming you mean the fermilab results, some people are saying we just need to apply a bit of duct tape, see physics.stackexchange.com/q/627849/50583
the other standard solution of turning it off and back on again seems risky in the case of the Standard Model