TeX, LaTeX and Friends - 2013-08-30

« first day (1034 days earlier) ← previous day next day → last day (3893 days later) »

1:34 AM

I would suggest to close because op has an not working installation (see comment). Do you agree?

0

Q: compilation problem

Suddenly a document that TeXShop was compiling without problems, started to display some error I never saw before: **2 ! pdfTeX error (ext4): \pdfendlink ended up in different nesting level than \pd fstartlink. \@EveryShipout@Output ...@Org@Shipout \box \@cclv ** The first compilation pass goe...

bibtex pdftex compiling

3 hours later…

4:55 AM

@dustin What are you trying to do?

2 hours later…

7:17 AM

@Kurt Yup: I've pulled the trigger

4 hours later…

11:05 AM

@DavidCarlisle What an honor! No, I am not Paulo Cerada. I am new in LaTeX and thus dependent on well working tools like his. (I just saw you message about me using vim and arara and had to smile...)

@LaRiFaRi I don't talk to vim users:-)

hehe, if it pleases you, I use TC on my Win-machine at work. So just talk to me on working days.

@PauloCereda excuse my typo in your name. (I guess your tools has too many "a"s and I got stuck)

@LaRiFaRi you should get @PauloCereda to convince you to switch to emacs

11:20 AM

@TorbjørnT. I am trying to convert the solving of the linear KdV equation in Matlab code to Python. I have variables and vectors convert but for loop code block is posing the problem. I have an idea on it which could be wrong maybe on the right track.

@TorbjørnT. this is the code block I in Matlab: for nn=1:nmax % integration begins
du1=i*ifft(k3.*fft(u)); v=u+0.5*du1*dt;
du2=i*ifft(k3.*fft(v)); v=u+0.5*du2*dt;
du3=i*ifft(k3.*fft(v)); v=u+ du3*dt;
du4=i*ifft(k3.*fft(v));
u=u+(du1+2*du2+2*du3+du4)*dt/6;
if mod(nn,round(nmax/100)) == 0
udata=[udata u]; tdata=[tdata nn*dt];
end
end % integration ends

@TorbjørnT. this is what I wrote in Python but I don't know if it should be wrapped in return [ code...... ] for nn in range(1, nmax):
du1 = j * np.fft.ifft(k3 * np.fft(u))
du2 = j * np.fft.ifft(k3 * np.fft(v))
du3 = j * np.fft.ifft(k3 * np.fft(v))
du4 = j * np.fft.ifft(k3 * np.fft(v))
u = u + (du1 + 2 * du2 + 2 * du3 + du4) * dt / 6.0
v = u + 0.5 * du1 * dt
v = u + 0.5 * du2 * dt
v = u + 0.5 * du3 * dt
if np.mod(nn, np.floor(nmax / 100.0)) == 0:
udata = np.array([udata, u])

Hi @Paulo :)

@DavidCarlisle Hey! <3

@tohecz Hey Tom! :)

Wrong smiley, sorry. :)

@LaRiFaRi Nah don't worry. :) In all these years of TeX.sx, people called me Pablo, Paolo and Paul. :)

@PauloCereda Paul was me!

@kan Mr. Paul Kannappan? :)

@PauloCereda No, I called you Paul! :)

11:28 AM

@kan :)

We should call everyone Bruce, to simplify things.

@DavidCarlisle Monty Python? I never thought you were a fan as well. :) @JosephWright can confirm this. :)

Ladies and Bruces...

Interesting youtube video!

@PauloCereda vvv

@DavidCarlisle Thought so. :)

11:32 AM

@PauloCereda unlike you and Joseph I'm the right age to have watched the originals:-)

@DavidCarlisle or Bruce(vim) and Bruce(emacs)...

@DavidCarlisle :)

@DavidCarlisle Very true (re your comment). I wonder if there are any conference sites that still have LaTeX2.09 instructions.

@NicolaTalbot probably:-)

@NicolaTalbot If I'm not mistaken, IEEE still have L2.09 guidelines. :)

11:47 AM

@dustin For starters, you use v before you define it. Rearrange the lines so the statements are in the same order as in the Matlab code. Also, the range doesn't include the stop value, so the calculation isn't done for nn = nmax.

@TorbjørnT. should it be for nn in range(1, nmax + 1)?

@dustin Yeah, I guess so. range(1,5) gives you [1,2,3,4].

@TorbjørnT. after for, should I have return [then that code block with the lines separated by commas and then ]?

@PauloCereda Ooh, cutting-edge!

@dustin It's not a function is it? so you don't have to return anything. And brackets are just used for indexing I think, so not needed.

11:55 AM

@TorbjørnT. is there anyway to force nmax to be an integer since the code returned TypeError: range() integer end argument expected, got numpy.float64.

@TorbjørnT. I have nmax defined earlier is nmax = np.floor(tmax / dt)

@dustin nmax = int(np.floor(tmax/dt))

@TorbjørnT. can I email you the matlab code and the python code files? It may be easier to understand what I have. They are only both about 30 lines max.

@dustin ok, sure. totaskj at gmail dot com

@TorbjørnT. ok thanks, sent.

12:17 PM

@dustin I sent a reply.

@TorbjørnT. thanks. Would I end up plotting udata v tdata?

@dustin Well, that imagesc creates a contour plot of sorts, but plt.contourf(np.real(udata)) doesn't give anything like the plot the Matlab-script gives.

... which may be because the definition of k is wrong.

@TorbjørnT. it is?

12:32 PM

@dustin Heh, yeah. It's just 0. Should be a vector. just a sec ...

@dustin k = np.hstack((np.arange(0,N / 2.0 - 1.0),np.arange(-N/2.,0))).T * 2.0 * np.pi / L

@TorbjørnT. thanks. Is there a way to make the image look crisper?

@TorbjørnT. always I had a question on SO if you want to answer it. I had to resort to getting help here because personally they suck over there. stackoverflow.com/questions/18519523/…

1 hour later…

1:50 PM

Meow.

2:17 PM

@PauloCereda Er, quack. :-)

@NicolaTalbot :)

2:31 PM

I'm just hoping that someone will post a question to which this is the answer

 \def\deff\end
{\def\deff\end}
 \deff\end{\def}
 \deff\end\def\def
{\end}\def\def

4

@DavidCarlisle What does it do? I can provide a question. :)

@PauloCereda nothing

@DavidCarlisle ooh.

@PauloCereda but it's error free

@DavidCarlisle :)

3:00 PM

@DavidCarlisle: provide an answer: tex.stackexchange.com/questions/130586/…

:)

@PauloCereda the one above? (vim users would downvote me to record levels:(

@DavidCarlisle :)

3:24 PM

@PauloCereda I bought a fotocamera. But don't expect anything. ;-)

@egreg Yay! :) Yay! :)

3:41 PM

@dustin Sorry, life called. Not sure what you mean by crisper, but pcolor created a nice plot (I sent another email).

@dustin Nah, you seem to have answered both of your questions yourself.

4:07 PM

Kerming problems can bite these words:

ahern, airn, alburn, arn, burn, carn, carne, carnel, churn, corner, cornet, cornic, corning, curn, curnin, darn, doorn, ern, fornent, frorn, garnes, gurn, gurns, harn, harns, hern, herne, horne, horner, horney, hornily, horniness, larn, larne, marna, modern, morn, rna, scarn, sharn, spurne, stearn, stern, tarn, tarne, torn, turnor, urn, yarn

Just to see the difference:

ahern, airn, alburn, arn, burn, carn, carne, carnel, churn, corner, cornet, cornic, corning, curn, curnin, darn, doorn, ern, fornent, frorn, garnes, gurn, gurns, harn, harns, hern, herne, horne, horner, horney, hornily, horniness, larn, larne, marna, modern, morn, rna, scarn, sharn, spurne, stearn, stern, tarn, tarne, torn, turnor, urn, yarn

(Via Christian Perfect on Google+)

2 hours later…

6:00 PM

@kan You forgot pom.

@Qrrbrbirlbel :-)

Whee, today I received plain-tex badge. /me is so proud. And in good company, too, I might add. :-D

@morbusg Congrats! :) And beware of David, he might steal your green ticks. :)

@Paulo: Thanks! :-) Well, usually it's egreg who steals the ticks. The man is inhumanely fast. :-D

Hey has anyone here done any self-publishing?

6:47 PM

Can I test somehow whether something is \csname safe or not (in my case possible dimension registers)?

Hm, I'm trying \detokenize but it seems that there is yet another problem...

@morbusg @NicolaTalbot

@PauloCereda If I have any left after egreg...

@DavidCarlisle egreg has now a photocamera. :) You need to encourage him to take lots of photos, so you can keep your ticks. :)

Keep him busy. :)

@morbusg Yes, I have but I have to go out now. (In fact, I should have left 10 minutes ago!)

@PauloCereda I'm going for my silver tikz-pgf just 5 more answers required I think:-)

@DavidCarlisle ooh! :)

7:03 PM

user image

@PauloCereda It's a bit modern for egreg, but he'll get the hang of it

@DavidCarlisle Where's the microSD slot? :)

@PauloCereda not sure it'll fit:

user image

@DavidCarlisle LOL

Hey, where's @JosephWright? We miss him. :)

@DavidCarlisle It's a bit smaller than that.

@egreg David is jealous because he wants the new Samsung smartphone. :)

7:10 PM

@PauloCereda actually we got one of these at start of the summer

user image

@DavidCarlisle ooh! :)

If I'm not mistaken, we have two photography experts in the house: @AndrewStacey and @percusse. :)

@PauloCereda I don't need the samsung smartphone as I have one of these

user image

@DavidCarlisle :)

I have one of these:

user image

7:30 PM

oh man pictures everywhere....

photo expert me? I'm an expert of looking at them

I'm no expert either. I've asked a couple of questions at Photography-SX but that's about it.

@Qrrbrbirlbel You mean things like \the\csname pgf@x\endcsname?

@percusse No, I meant \csname something \pgflinewidth something\endcsname. (\pgfgetarrowoptions is used inside \csname which disallows \pgflinewidth in the options, yes this could be fixed but inside PGF code).

@Qrrbrbirlbel \pgflinewidth will expand so trouble ahead.

@percusse Hm, no, even not. That's the problem.

7:40 PM

@Qrrbrbirlbel Not to its value but as a value expecting macro

have you tried protecting it or chain of \noexpands ?

@percusse I can do \string but than it doesn't get recognized by PGF math of course. A \protect is followed by "The definition of an arrow may not use \pgfusepath. Interesting.

I would need to \detokenize in \pgf@arrow@full@name while also \expandafter \pgfgetarrowoptionscorrectly. Meh.

I found a photo of my new gadget; as you can see, it's easily portable because it's pliable

user image

@percusse It's about this answer. You can't use \pgflinewidth (or any register) in dashed arrow pattern:

1

A: tikz: segments with 5pt dotted ends

Here is an arrow tip implementation much like How do I draw lines that end with an ellipsis? but with options implemented with the help of \pgfsetarrowoptions and \pgfgetarrowoptions, though you don’t need to use them as I have added three TikZ keys: dotted arrow length: the length of the “arro...

@Qrrbrbirlbel Aah lots of @s :P

Arrows are protocoled so they are pretty anti-social about dynamical value setting.

@percusse I'm sorry?

7:53 PM

That's why many things are not allowed explicitly since they are defined when they are read and that's fixed for good.

@percusse Yes, that's \pgf@arrow@full@name at work.

@Qrrbrbirlbel So probably you are defining an arrow head at each issue of the command. Not that I've digested your code in 2 seconds.

Do we have a question-and-answer that will tell me all I want to know about how TeX handles vertical spaces?

@Qrrbrbirlbel My guess is that you have to expand the value explicitly. Otherwise protocol thingy will complain

In particular, I want to define some environments/commands in such a way that there's always "enough" space before or after them, but if two follow each other then I don't want the spaces to add, and space at the top of the page should be ignored.

7:56 PM

@percusse Exactly. But you would only do this in this example. I simply want to show the possibilities since I'm actually using three arrow options (length, pattern, phase).

@percusse Yes, yes. As I said \pgflinewidth inside of \csname. Last two paragraphs of the answer explain that. Either \the\pgflinewidth or hiding in another style (whose name will be used in \csname instead).

@Qrrbrbirlbel I'm not sure if you can fool via expasion blanket but if it works it works :)

@percusse A simple idea is to use a substitute for \pgflinewidth which expands to a simple value (possible nothing) and only is inside the arrow's code it is set to \pgflinewidth. But why the hassle?

Can an moderator please merge the "answer" here with the question?

6

Q: How to lock the back reference at the top left corner of the citation?

anybody knows how to lock the back reference at the top left corner of the citation within bibtex? for example It works to me when using \bibitem (yes, it goes there automatically) but not through a .bib file... Thank you in advance

hyperref back-referencing

@Qrrbrbirlbel I got that but I don't know if when expanded the arrow declaration will be happy about it.

though you say it works so problem solved :)

@percusse As the line width is "protocolled" (as you say) either way that's not a problem.

Well, I haven't actually tried it ...

8:06 PM

@Qrrbrbirlbel System layer is not my thing, when I think I fooled TeX, the PDF stops working. It's Heiko, Herbert stuff. I'm not even good at those things. So I rather hope for the best about those things instead of understanding them.

When I retire I plan to read that damn PDF 1.8 spec document

There is also Paul Isambert (i guess he is still around on comp.text.tex under zappathustra nick) these guys are doing unbelievable stuff :)

@percusse I recently browsed through the PDF specification to find the \special things PGF inserts when a path is stroked or filled, very interesting how the filling rule (even odd vs non zero) is communicated through PDF.

@Qrrbrbirlbel yea those pdfliterals are crazy

Awesome, from now on I'll use this: asciiflow.com

Just a simple verbatim and voilà. :)

@PauloCereda But it's WYSIWYG! :O

@Qrrbrbirlbel check out yax package it's awesome. In most the cases better than pgfkeys

8:16 PM

@Qrrbrbirlbel :P

@percusse Haha, from CTAN's package description: "Finally, YaX’s syntax is a quite peculiar (as few braces as possible)"

@Qrrbrbirlbel Hahah, it's the reason why i gave up text based key system. It's too good.

1 hour later…

9:27 PM

What is exponent in latex?

$e^{7}$? Or $\exp{7}$?

-

Err solved it, sorry noise.

@hhh Depends on what notation you want to use ...

2.71... for e, I have always used $e^{...}$, not knowing whether correct one.

@hhh e^x and exp(x) are the same thing.

@hhh This really depends on the notation you want to use. But for the sake of your readers use exp if you have a rather big exponent. Nobody wants to read e^{\frac{1}{n^x}} or stuff. Sometimes you also will see \mathrm{e} for a roman upright e.

I have what I hope is a simple tex question, perhaps too simple to ask on the main site. Google hasn't helped though. Is this the right place to ask?

9:38 PM

@Y.P. Sure.

Great! What is the way to typeset a long equation in pnastwo document class?

with revtex4 (and revtex4.1) the way to do it is \begin{widetext}\end{widetext}

pnastwo doesn't support that.

@Qrrbrbirlbel You mean $exp\left(...\right)$ or is there some special exp with long exponent?

including the widetext package doesn't work either, as it seems something in pnastwo forces the tex compiler to ignore the directives in the package

@hhh I'm sure a \exp is defined, otherwise use \DeclareMathOperator\exp{exp} so that "exp" is typeset in the correct font, then you can use either \exp(<something>) or \exp\left(<something bigger>\right) or \exp\biggl(<something>\biggr).

@Qrrbrbirlbel Yes, thank you. Readability, priority 1 :)

9:45 PM

@Y.P. Never heard of that document class. You could use a multiline math environment from the amsmath package (for example multline) to break the equation over several lines. If widetext just increases the text width locally, you could try adjustwidth from the package of the same name (see e.g. tex.stackexchange.com/a/36124/586, use negative lengths).

@Y.P. If it doesn't work, I'd ask a question on the main site with a small compilable example. But please add a link to the pnastwo class, it isn't on CTAN.

@TorbjørnT. ok. I will try what you suggested, and if that doesn't work I will ask on the main site. The pnastwo class is a the custom style class for the Proceedings of the National Academy of Sciences

« first day (1034 days earlier) ← previous day next day → last day (3893 days later) »

Transcript for

Aug '1330

$TeX - LaTeX$ TeX, LaTeX and Friends

\begin{chat}

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

$TeX - LaTeX$

site design / logo © 2024 Stack Exchange Inc; legal

mobile