I noticed it several times that when I drink beer (even one bottle) after some heavy workout, the next day my muscles are more stiff than other times, and not the same way... Is it because that lactate and NADH levels are raised both due to the anaerobic training and alcohol metabolism?
Regarding the bacteria found in Mono Lake, CA that scientists believe uses or can use arsenic in its DNA backbone where life as we know it uses phosphorus (according to their experiments depriving the microbes of phosphorus and providing much arsenic), have researchers conjectured and tested whet...
My professor says
The principal nerve of the P.S. is n. vagus X, descends together with the
large cervical vessels into the thorax and abdomen where it gives off
visceromotor fibres to all visceral organs, forming visceral network -
plexus (down to flexura coli sinistra).
Nervus Vagus...
@Fx you need to look at the users' profiles, too. This guy Masi, spamming the hell out of the board with his wikipedia copy-pasting and self-answering, is med student.
Just looking at:
(Source: http://xkcd.com/327/)
What does this SQL do:
Robert'); DROP
TABLE STUDENTS; --
I know both ' and -- are for comments, but doesn't the word DROP get commented as well since it is part of the same line?
@yoda good grief. It's like every time a question pops into this guy's mind, his first thought isn't to have a bit of a think about it, but rather to post on SO.
Warning: mysql_connect() [function.mysql-connect]: Access denied for user 'nx000043'@'localhost' (using password: YES) in /home/nx000043/public_html/argentina/inserts/jofre/conexion.php on line 11 Error conectando al servidor
@CHM as I've said before, when you know Mathematica, Python's trivial. I picked it up in a few hours after a friend suggested to write something in it that I'd prototyped in Mathematica.
@OleksandrR This one hit home. A friend of mine is suffering from a severe case of something like Crohn's, and can chug the barium shakes. It makes the radiologists want to vomit. :)
@CHM A friend of mine in astronomy uses it almost exclusively.
@CHM many will agree that it's definitely worth it to learn. It's also a general purpose language, unlike mma et al., so you have a lot more flexibility in what you can do
All you'll ever need to do with installing python. No more worrying about glibc and crap errors and make files and breaking your ass over installing matplotlib (especially on a mac)
Free for academic use, so you just need to enter your univ id and they'll send you a link
@CHM my motivation was that nobody else in my university seems to use Mathematica, so it's not installed on our supercomputer, whereas Python's available everywhere. That said, I do very much like NumPy's advanced indexing. It's absolutely mind-bending at times, but highly flexible. Since questions came up here on the subject, I've been toying with implementing it properly in Mathematica if I get the time.
@yoda logical indexing isn't much different than Pick, IMO. Problem is, logical indexing returns a masked view onto the original array, not a copy like is usual in Mathematica. Not quite sure how to deal with that kind of issue in general.
@OleksandrR Sure, I know how to do an equivalent in mma, but there is something neat about being able to do a={1,2,3};a[[{0,1,1}]] and getting back {2,3}. Of course, this is just an old habit and I try my hardest to not transfer habits between languages. It's more of an will-this-be-possible-in-mma kind of thought
@yoda I had in more in mind to enable things that are tricky to do in Mathematica usually, like e.g. applying a set of Spans along an axis of an array. Main problem is, I find these even less intuitive than the (infamous?) second argument of Flatten...
@yoda for instance instead of Transpose@Apply[{Part[#1, ;; 200], Part[#2, ;; 190], Part[#3, ;; 175], Part[#4, ;; 155]} &, Transpose[array], {1}]; (which I need to use regularly), one might write array[[{1, 2, 3, 4}, {{;; 200}, {;; 190}, {;; 175}, {;; 155}}]] with advanced indexing as in NumPy.
@Heike yes, I use it quite regularly myself. But I usually find my initial idea of what it's going to do, is not what it really does. Thus an experimental approach to coding is sometimes called for.
Then again, perhaps I don't help myself by always choosing to work with data in the form of 5- or 6-dimensional arrays.
@CHM if that was meant for me, no; my arrays are full-rank. I just find it gets a bit non-intuitive to imagine what an argument of {{4}, {1,3}, {2}} will do.
Perhaps my array-visualisation is not as good as yours, however.
commented: @CHM, just because a question does not have a clear-cut answer, it doesn't mean it isn't a valid question. Exploring such questions and explaining why or how is what scientific thinking is about.
@Verbeia good progress indeed. Thanks for posting. I must admit I'm a little surprised at the strong upward trend, but I suppose it shows that we're probably not too near to saturation point, which is good. Should get to 1500/day easily if this carries on.
@Verbeia that sounds reasonable, though TBH I'm not so sure if the "hot questions" we've had so far really represent the best that our community has to offer. It seems to be quite arbitrary which questions turn out "hot".
@Heike I think because most of the scifi/fantasy questions are totally subjective, so it's easy for five or six people to give their opinions without duplicating one another. Problem-solving questions of the kind we have here are obviously much more objective, so I suppose it's just the fact that things can be done in so many different ways in Mathematica that makes "hot" questions relatively common for our site.
@Heike for example this question seems to me quite ridiculous: the obvious answer is "because it's a fictional scenario" but the actual answers are various different quasi-plausible rationalizations of the obvious contradiction.
@yoda so basically it was not really a concrete question.
the situation is, I have a program which may be formulated in matrix form (construct large sparse matrix, find eingenvectors, construct other matrices from them etc)
and right now I have it in mma and scipy. they're roughly equally fast
I am really limited by the speed of this thing, so have to get it faster. I can't think of doing anything other than rewriting it in matlab, but that'll take me a few days as I am not used to writing nontrivial programs in it
so I am trying to decide if it's worth the effort
now, I remember you and JB saying that you find matlab faster or more convenient (I don't remember) for some things. could you indicate what kind of thing you find easier there? what faster? or do have you not used both for similar problems?
I just don't know how to find out short of writing the program twice, and I don't know anybody (in person) who can write decent mma code (in terms of speed for numerics), so I remembered that discussion and decided to ask
I have not used both for problems on the scale you're talking of, but I generally find matrix operations — constructing, indexing, slicing, reshaping, flattening etc., simpler in matlab.
If you can describe your problem or piece together an example that shows the core of what you're trying to do, I can try and port it to MATLAB and see if it makes a difference
That is, assuming it's not a huge project with many loc
so you have not noticed significant performance differences? (don't worry, I'll use your answer to decide, then forget you ever said anything about speed :) )
@yoda but from experience, if I do it in C it's marginally faster than mma (if matrices are the bulk of the work) and takes me much, much longer to write and debug
@yoda the size isn't a huge problem. they're up to 50000 by 50000 in extreme cases but sparse; but that is the same for both mma and matlab
and scipy is all good, but when I am forced to write actual code it slows right down. julia is nice but not available on our large machines
hmm... my gut tells me that it might be faster with MATLAB going by what I've noticed. However, I can't say for sure, without knowing the problem... since you say that there are other moving stuff and operations, it could be an area where matlab doesn't shine or needs excessively long code to do...
for example, say you have two vectors/matrices a and b and you wanted to do some f(a,b) and index the output (assume also a vector/matrix), you could do something like f[a,b][[index]] in mma, but with matlab, you can't do that. You'll have to do something like temp=f(a,b);temp(index)
If you vectorize your code, things are way faster in matlab than mma. Use linear indexing when you can. Sometimes flattening to a vector, doing all operations as a vector (where permissible) and then reshaping might be faster
that was the impression I got from playing with it some years ago, but I know more tricks in mma now. then again I notice that for large projects I tend not to use complicated speed-up tricks because I am already at my limit thinking about the physics and don't have the capacity to optimize much
I kind of do... attend 2 days, goof off 3 days. Can't argue with that. Add to that a week of post-conference laziness, and I've got a solid ~2week "holiday"
luckily, my institution is an extremely active conference centre; there's one every week or, at most, every two weeks. so I don't need to move much, people pass through at the rate of (I think) 1500 visitors a year... or so I read in some report, and I believe it. it's good for someone like me!
@acl oh, that's not what I meant. Here, there's a conference going on every time and they order excess food (yeah, wasteful...) and they send out an email to the gradstudents list to come grab it all (not participants at the conference)