jump to navigation

Some post-workshop irrelevant thoughts July 28, 2009

Posted by keithkchan in Whatever.
add a comment

I would like to collect some of my thoughts about the trip to the wild west of the USA. I am not going to talk about what I learn academically, for the obvious reason — I did not learn much. Fortunately I don’t need to report anything to my advisor, so it is OK. But usually in this type of workshop, we can only get a rough idea of what other people are doing in the adjacent fields. It is also a good chance to meet people working in the same field. Often it is a good approximation that this kind of trip can be transformed to a nice travel experience. I went to the Great Sand Dunes in Colorado and the Grand Canyon in Arizona. So, this time the approximation is excellent.

The exact location that I went to was Santa Fe, New Mexico. I realize that the life outside New York City is completely different. Without a car, you can do nothing. When I planed to go there, I thought I could buy soap, shampoo blah blah blah there easily, just as in Manhattan. Oh my beloved Nature, that was completely wrong. The undergraduates there told me that I could get to the nearest CVS by walk in one hour. Fortunately the undergraduates were so kind that they drove me to CVS by car. The pace of life there is slower than in New York. People there are generally nicer and more polite than New Yorkers.

One thing that annoys me pretty much is that people there are quite religious. One can find churches everywhere. The buildings, streets are usually named as Saint XYZ. OK, that is just some names. Who care? Near Santa Fe, it is the Los Alamos National Lab (LANL). A cosmologist working at LANL told me that people in that town were pretty religious. As you may know, at LANL, some people are working on weapons of mass destruction. Those people are particularly religious. That seems to me pretty understandable. For atheists, we at most make fun of those religious people. But religion can drive people insane. They can do crazy things if they believe themselves to be messengers, disciples or whatever of the god, and they think you don’t believe in his god, or even worse believe in the wrong god.

Even within the participants in the workshop, they are more religious than my colleagues at NYU. One guy believes that god can change the laws of physics at will. Moses could violate the laws of physics when he separated the water in Red Sea. But he worked on dark matter observation. What if he observes is just some tricks played by god? Then his work is totally meaningless. I thing he should add in the paper that the phenomenon he observed can be an artifact due to god. What they work on and they believe are fundamentally inconsistent.

There are two possibilities for the higher percentage of “religiousness”. First simply people outside New York are more religious. Secondly, astronomers are more religious than physicists. For physicists, the laws of physicists are fundamental, it can’t be changed arbitrarily. I will never ever identify myself as an astronomer.

A Comparison of the Plotting Softwares May 23, 2009

Posted by keithkchan in Whatever.

We need to plot some graphs from time to time. A good plotting tool is very important, in particular when you are going to put some graphs in the paper.

There are many plotting softwares available. However, I am a fan of the open software, and try my best not to use the commercial softwares, like Matlab and Origin, in plotting. Besides saving money, I admire the cause of open softwares. So far I can limit myself to using only open softwares happily (except Mathematica). So I will only talk about the open softwares only.

So far, I only need to plot not-so-fancy 2D graphs. I have tried several free plotting softwares, including xmgr, gnuplot and matplotlib.

xmgr is a graphical (GUI) plotting tool. Since it is GUI, it is relatively easy to start with. The quality is also good, as far as I remember. For those who used to use GUI, this is a good choice. But somehow I stop using it because it is not installed on the Linux system here at NYU. (Is it xmgr less popular than two other plotting tools? Is it not a routine part of the Linux distribution?)

Both gnuplot and matplotlib are script plotting tools. I think it is harder to begin with, in particular for people still live in the Windows world. If you use Linux or Mac, you probably are familiar with terminal and command-line approach already. I used to use Windows, and I found it pretty hard to accept the command-line softwares at that time. After quitting Windows, I am quite comfortable with scripts now. Of course, you don’t type the scripts every time. The first plot may be painful and time-consuming as you need to find out the appropriate commands to polish your graph. The second plot is going to be similar so you just need to copy from your old scripts. In the long run, the time required should be similar to, if not less than, the GUI plotting tool. I believe that the real master would use the scripts.

gnuplot is a sole plotting tool. The basic commands are easy to find online. I have been using it for some time. However, I (and some of my colleagues) find that the graphs from gnuplot still fall short of the publishing standard. Below I show the same data plotted using gnuplot and matplotlib respectively. Which one looks nicer?

I find the second one looks better, and it is plotted by matplotlib. If you think otherwise, I have nothing to say. matplotlib starts its life as a mimick of matlab, and it is a library in python. As you may know, python is a rather popular scripting language. To use it you have to import the matplotlib library in python. However to install the library may not be so easy, depending on your system. If you use Windows or Mac, you can install it easily using the Enthought Python Distribution, which is free for educational purpose. Certainly I satisfy this condition. I heard that installing it on Linux is pretty tricky. On one hand since it is just a library in python, you need to know the language python a little bit also, the potential barrier to overcome is higher than that for gnuplot. For example, I need to plot some graph using the columns of data in an ASCII file, I find it ridiculously complicated to do. It took me quite some time to google how to do it. On the other hand, as it is just a library in python, learning how to plot a graph using matplotlib you in fact have also started learning the python language. Isn’t it one stone two birds?

So what’s the conclusion? Which one is better? There is no conclusion. It is up to you to decide. I will use both gnuplot and matplotlib whenver convenient.

The Art of Asking “Stupid” Questions March 21, 2009

Posted by keithkchan in Philosophy, Whatever.
add a comment

These few days we tried to verified the approximation used in some guys’ paper, who claimed to have done what we have done. In their paper they used the spherical approximation, instead of solving the equations exactly. When I showed my advisor the results, he immediately said that it must be wrong since in the large scales the nonlinear effects should not matter. I then spent a couple of days to check if there were stupid mistakes that I have made. Indeed I did, but it was not related to the major problem. Although the conventional wisdom was that the linear regime should not be affected by the nonlinear effects, the equation we solved were highly nonlinear, this was not obvious to me. I went to press him, and we finally realized that in the spherical approximation the nonlinear effect in the high k can indeed influence the scale physics. Incidentally, this partially suggests the answer to my previous question if quantity is more important than quality. Those guys write many papers, quality may fluctuate a bit.

I often ask questions most of them are probably stupid. I think the real good physicists certainly know the importance of asking questions. Sometimes, some concepts may be so “obvious” to them they did not even think about it. Like in the case of my advisor who is regarded as an expert in this field, but he could still miss something sometimes.

I believe that if I keep on asking questions I will be able to good questions sometime. The most important thing is not to stop asking questions, even if they are are stupid. After all, it is not clear if it is stupid a priori.

Happy Birthday Mr Darwin February 12, 2009

Posted by keithkchan in fun stuffs, Philosophy, Whatever.
1 comment so far

I heard that this year is some anniversary of Charles Darwin for some time. When I look around other blogs, I learn that today is Mr Darwin’s 200 birthday. Although I am not a biologist, I love science, and Darwin’s evolution is a beautiful theory.

From my previous criticism on the stupid creationist, you may know that I believe in evolution. But I believe in evolution is not because Mr Darwin told us so, not just because we look like chimpanzee. This theory has been found to be in agreement with a lot of evidences. There are thousands of fossil records, which enable us trace out the evolution tracks of species. We have sequenced the genes of many species. For example, our genes are more than 90% similar to the chimpanzees. In medicine, we have seen the evolution at work many times, although this time it is undesirable. When we attack the virus or bacteria with some medicine, the medicine kill all those strains that are not resistant to the drug, and only the resistant strain survives. Technically it is said that a strong selection pressure is applied. Before other non-resistant strains are killed, they compete with resistant strains for food and mate. Since selection pressure killed the non-resistant strains, the resistant strain take over the whole species. The species becomes resistant to the drug. That is one of the reason why curing AIDS is so difficult.

When people see the complexity of the species in our world, which is beyond their imagination, so they cook up the stupid creationism. This stupid idea (note that it is qualified to be called a theory since it has no predictive power at all) has no foundations, well except those people believe in the Book written one thousand years ago by some people. If the universe if created by some intelligent stuff, which is usually called God, according to those creationists’ logic, God has to be created by some super-God, which may be created by Super-Super-God, blah, blah, blah. On the other hand, evolution is a very slow process, the species evolves adaptively based on the environment.

Those religious people often attacks evolution saying some of the links are not clear, for example from fish to reptiles. There are more and more fossils discovered. I think the complete picture will be possible soon. But if you think about those religious people’s attack, you will find it silly. Suppose there are two jigsaw puzzles, the one representing evolution has been filled say 70%, the one representing creationism is empty. The religion people then tell you to believe the Holy Book and the jigsaw has been filled completely although you can’t see it. They then say the evolution one is 30% not filled, it must be wrong. Which one do you believe? Surprisingly, they are quite a lot of people believe in religion one.

So if you want a satisfactory explanation of the Nature, evolution is only theory that you can trust. Of course evolution tells us that if somebody proposes another theory with better agreement with the evidences, we will accept that theory as the working theory. This is spirit of evolution, survival of the fittest. Happy birthday to Mr Darwin.

Alternating Group S_3 January 24, 2009

Posted by keithkchan in fun stuffs, Trivia, Whatever.
add a comment

As expected, after the start of the spring semester, there are many more chores or even crap to deal with, I don’t have much thoughts to post. Therefore I probably will not update this blog very often. If you check this blog regularly, feel free to reduce your frequency of visiting this blog by a factor of 2. I also find that my research productivity has dropped by at least a factor of 2 when the semester starts.

OK. In this post I would like to talk about the group, the alternating group S_3, another name is dihedral group D_3. Well, everybody in QFT II class knows it well, why do I bother talking about it on the Blackboard? First, I  want to say something, you can’t stop me from doing this. Second, I realize it may be a good analogue for the Lorentz group. This connection may turn out to be another stupid idea, as most of my “great ideas”. That’s why I want to write it down before I realize it is really stupid.

The name alternating group originates from the fact that the group elements in this group are the permutation operations of 3 (non-identical) objects. There are altogether 3!=6 ways to permute them. The other name dihedral group is from its  geometric interpretation.  We can regard the elements as rotations and reflections of an equilateral triangle. The rotations by 0 deg,  60 deg and 120 deg, and reflections about an axis passing through one of the vertex and perpendicular to the opposite side of the triangle.  Therefore there are 3 reflections.

Take a subgroup group H, which contains the 3 rotations. We denote the other 3 reflections elements by Hr.  One can check that H^2=H (rotation twice is still a rotation),   H \cdot Hr = H (rotation followed by a reflection is equivalent to a reflection about a rotated axis, so it is still one of the reflection), (Hr)^2 =H (it is well-known that two reflections can be cast as a rotation about the intersection of axes of reflections). So H and Hr from the group Z_2!

Now lo and behold. I think S_3 is an analogue of the Lorentz group in some vague sense. The Lorentz group consists of the rotations and boosts, while the alternating group consists of rotations and reflections. Obviously I want to link the boosts to the reflections.  If you remember the commutation relations between the generators of Lorentz group J (rotations) and K (boosts). Schematically, [J,J]~J, [K,K]~J, [J,K]~K.  This looks similar to relation between H and Hr.

Of course, the alternating group is discrete and the Lorentz group is a Lie group. They are rather different. The claimed  connection may be just some bullshits. But there is carbon in bullshit, and we know that diamonds is just some crystal structure of carbon. So there exists non-vanishing chance that bullshits may become diamonds.  I will keep it in mind, we will see…

Effective Temperature of the Papers January 20, 2009

Posted by keithkchan in fun stuffs, Statistial Physics, Whatever.
add a comment

Today when I listened to my colleague tidying up the papers from the printer so that he could staple them neatly, I came up with an idea that we can probably define some effective temperature to characterize how neat the system is.

OK, let me state the problem more clearly. For example when you print an article, the papers from the printer do not completely line up. So usually we will take them up a little bit and release it. (It is hard to describe it, but I suppose you know what I mean.) Usually we do it a few times, until we are satisfied with the neatness. What I am interested in is that how many times is “good enough”. Clearly after doing it 100 times will not be better than 10 times since the system is in equilibrium, and subsequently there will be fluctuations about the equilibrium. I think we can define an effective temperature for the system, and from which we can extract the fluctuation. I expect how fast the equilibrium can be achieved depends on the textures of the papers, and the number of pages.

I was not sure if somebody had come up with similar ideas before. I was so excited (maybe crazy) that I even talked to Paul Chaikin about my idea. You know Chaikin likes to pack M & M chocolates. He was busy and did not seem to be very enthusiastic. In particular, he cautioned that there were many ways to define the effective temperature, which may or may not be meaningful in this kind of non-thermodynamic system. One way to define the effective temperature of the system of papers is to calculate the root-mean-square (RMS) of the papers from the centroid of the system.

I agree that the definition of the effective temperature is not unique, may even be arbitrary to some extent. Besides that, there are some arbitrariness in the each padding process if it is done by hands. If it is done by machines, machines probably can do it “perfectly”, and it does not address my original motivation, i.e. tidying up papers by hands.

It does not seem to me this problem will yield anything more than satisfying my curiosity, and I have many other things to do, thus I don’t think I will pursue it further unless somebody is also interested in and want to do something crazy. But at least this problem manages to become a post on the Blackboard.

Big Crunch and Academic Job Prospective January 17, 2009

Posted by keithkchan in Philosophy, Whatever.
add a comment

In the past I thought that the job opportunities were insensitive to the general economical environment. Now it seems to me that this is not true. The credit crunch in the financial sector has non-minimal coupling to the academic job market.

Now it is the season of post-doc hunting. Around this month, many institutes are selecting their prospective post-docs. It seems the competition for post-doc position is veeeery intense. Let me give you an example. There are 500 applicants for a post-doc position at NYU CCPP. A colleague here at Room 538 is also looking for post-doc. He told me that he applied for more than 10 post-docs. He have several papers already but so far he only got negative reply. He seems not so optimistic about his job hunting.

There are several reasons for the strong coupling between the financial sector and the academic market. First the money in the academic sector shrinks. Many positions are in the academic sector are funded by some endowments, foundations, or government agencies. The returns from the sources is dramatically reduced after the the economy turns bad. Many institutes have started to reduce costs. For example, MIT is making 5% cut in expenditure. One of the easy ways to cut cost is to reduce the number of new openings.
A common path for the graduate students from theoretical physics is the financial sector since it offers much higher salary and the numbers of opening are much more. However, ever since the big crunch, the competitors from MBA, laid-offs from the other financial companies have increased dramatically. The financial market path also becomes very difficult. Some graduates may then think it would be better to stay in the academia in this difficult financial situation. The increases the number of applicants for the post-docs. Since many people have anticipated the bad situation, they try to maximize the chance of success by applying as many as possible. These cause the huge number of graduates fighting for a job.

I still have several years to go, and by the time I graduate the economic situation will probably have recovered. If possible, it is better to stay in the current, for example delaying the graduation. For the friends who are applying for post-docs or whatever jobs, good luck, Nature bless you.

A Nice Introduction to Quantum Gravity January 2, 2009

Posted by keithkchan in Whatever.
add a comment

For those who are excited by my two-cents review of modified gravity and want to learn something about quantum gravity, I recommended this talk by Daniele Oriti . I found this nice talk from Dmitry Podosky’s blog. I hope you enjoy it as I do. These talks were held at Perimeter Institute. You can check out other talks here.

By the way, our colleague Mr Eyal will join our party, and write something smart on the Blackboard to keep you guys entertained.

Pros and Cons of Mac November 19, 2008

Posted by keithkchan in Whatever.

As many people in our CCPP center, we like a Mac. In fact, in our office, 73% use Mac laptop, 18% Windows, and 9% Linux.

I think most of us like Mac since it is very sleeky and stylish. Besides its cool appearance, it is very stable as OS X is modified from Linux, with some added features. We can easily install the Unix package without much difficulty. We have the stability and safety of Linux. From my own experience, my laptop never (almost) hangs, and if some application goes wrong, we can always kill and we never need to restart boot the computer. We don’t need the expensive and resource-consuming anti-virus program at all! On the other hand, it is easier to use than Linux. There are also many commercial package have been adapted to Mac OS.

I have been using for my MacBook for one and a half year. I thought Mac was perfect until my laptop went to the hell. On that day, my laptop suddenly hangs and every thing freezed. I realized something serious going on because it never happened. I turned off my computer, and turned on it again. Oh, my beloved Nature, it could not started. I later learned many people who used MacBook suffered similar problem as I within one or two years after using it. It was hard drive failure. Those people who have that problem within one year warranty can have hard disk replaced free, but they lost everything in the hard disk. Fortunately all my research stuffs are in the department server, otherwise I don’t know what to do. Since my warranty has passed a few months ago, I asked them how much would be charged. They told me it varied from $300 to $800. Oh, f*ck you Apple! Since I was pretty sure that it was hard disk failure, I ordered one and replaced it myself. It costed me $140 for a 160GB hard drive. Now my laptop has returned from the hell now. I just need to install everything again. It is kind of pain ass.

This reminds me that Apple stuffs are usually much more expensive than its competitors. One of the reason is that many companies manufacture computer with Windows, but almost every Mac products from Apple. For example, Apple has its own VGA port adapter, but no usual non-Apple displays such as projectors and monitors, use it. Hence, the speakers, who use Mac, always need to carry an extra adapter with them. If something goes wrong, and the warranty has expired, it will cost you an arm and a leg to repair it.

With those pros and cons in mind, it is up to you to decide which OS to use.