I do not want to decrease the font size, nor increase the width of x-axis (I would prefer to set it with the default value instead of x=2cm)
Also, I do not want to increase the scale of the graph
Under this conditions, can you imagine an alternative? Thanks!!
P.S. Regarding the manual set of intersections points, I tried to use intersections library and name path within the two plots, but the only point of intersection that I get is the first one (marked in red). The other one I could not get it. I mention this if you are wondering why I am using approximate coordinates :P
P.S.2. I also do not want to delete the label -1 of x-axis
@TeXnician good idea! How can we do that? If I write something like node[above] {$\alpha_2$} then there is not the same space between the alpha-label and x-axis, and the tick labels and x-axis. It is a good idea, but I would like to get a proper space. I hope you understand me
@marmot hello marmot!! <3. Thanks for your answer! However, the problem is that the font size of the alhpa-labels. I want them to be the same as the tick labels font size i.e. font=\tiny, and if I add it as a parameter of node we get a non align alpha-labels wrt the axis labels
If I use \draw[gray,dashed] (2.16,0.84) -- (2.16,0) node[below,black,font=\tiny]{$\alpha_1$}; \draw[gray,dashed] (-1.09,-0.88) -- (-1.09,0) node[above,black,font=\tiny]{$\alpha_2$}; I get aligned alphas.
@marmot I do not get aligned alphas. I get the above output ^^^^^^ and the same with alpha_2 (if there would be axis labels above the x-axis line, not below like our graph)
@marmot sorry but not 3pt. If we use the default TikZ alignment i.e. we use xtick={-2,-1.09,-1,0,1,2,2.16,3} and xticklabels={$-2$,$\alpha_2$,$-1$,$0$,$1$,$2$,$\alpha_1$,$3$} we get alpha_1 aligned to the top of the number, not below, so there has to be something like 2pt (<3pt for sure)
@marmot also, the line that denotes the position of the number (here alpha_1 and alpha_2) is not drawn (take a look at the vertical line above of 2, 1, 3, etc.)
@manooooh You can use xticklabel cs but I do not know which output you are after. In any case, I do not know a magical way to avoid overlap, nor do I even personally think one should align them with the ticks. They should IMHO just sit where the dashed line ends.
@marmot why? I have calculated the values of alpha_1 and alpha_2 using approximation methods, so we can replace its labels by a number with a lot of decimal digits... Hence those labels are indeed part of the axis ticks labels, as 2, -1, etc.
@manooooh This is just my personal opinion. Also you could try \draw[gray,dashed] (2.16,0.84) -- (2.16,0); \draw[very thin] (2.16,2pt) -- (2.16,-2pt) node[below,black,font=\tiny]{$\alpha_1$}; \draw[gray,dashed] (-1.09,-0.88) -- (-1.09,0); \draw[very thin] (-1.09,-2pt) -- (-1.09,2pt) node[above,black,font=\tiny]{$\alpha_2$};
@marmot now it is perfect (at least for me), thanks!
Have you deduced the values mathematically or have you been trying by trial and error?
We are 5 animals in chat (counting the 2 ducklings that Ulrike has) <3
@marmot also, do you know how to not clip the graph but not using approximate values such as ymin=-1.55 (it must be -1.5 but it would cut the blue plot a little)? If I try clip=false I get a very large plot!!
I do not know why this happens, if we are restricting the plot using ymin and ymax
@marmot oh, awesome!! (Perhaps the new output is a bit bigger than the original but that's fine). Your idea is to delete the excess space and aplying it to ymin, right?
@manooooh The idea is to add the negative of the line width, \mydy to ymin. \pgfmathsetmacro{\mydy}{0.2/1cm} converts is to y units. If we do not know what the y unit is, this won't work. That's why I set y=2cm. If you set y smaller, you need to decrease 1cm in the above formula accordingly, e.g. for y=1.6cm you will need \pgfmathsetmacro{\mydy}{0.2/0.8cm}.
@HenriMenke how did that come up now, or has it being bothering you for years? (@egreg may have a view as well) of course if we do change I'd have to consider if I want to use luatex's direct control over cramped style if using luatex
@DavidCarlisle adding extra functionality to macros by making starred, plussed etc versions. Sometimes it makes sense, other times it can get confusing.
For example I use \For at lot. It is defined as \qquad\text{#1}, but the starred version only used \quad. Here the star makes sense. But not sure a mathtools user would associate \substack* with a cramped version (especially because most users does not know what cramped is)
@egreg could you explain what this means? "Pronto prontissimo son come il fulmine: sono il factotum della città." I tried google translate, but it gave a me word-by-word translation (which to me is unclear).
I hyperrefed PDF files, can we build the bookmark index "manually" that is have one set of data in the ToC and a different set in the PDF bookmark?
I'm playing with a book of abstracts. I have the talk titles as main items in the ToC, but might want presenter names in the bookmarks instead. (haven't decided yet)
@Slereah it'll depend a lot on which back end you are using, native mathml, svg, common html, html+css, .... but the mathjax developers don't chat here you'd be better asking on the mathjax list (or a question on stack overflow mathjax tag)
@UlrikeFischer I think the requirement is for the bookmark structure to be different to the headings/tableofcontents structure so switching individual strings via texorpdfstring is probably not the right level
@daleif It wouldn't describe this as a hack - if the structure is right, and you only want to change the words, using \texorpdfstring is imho quite appropriate.
@UlrikeFischer Seems to work just fine. Now I just need to get my custom toc entry types to have the proper bookmark level. But that shouldn't be too hard (or I should just make them non-custom...., they came from a custom toc list, that I've now changed into a "normal" toc list)
@PauloCereda so close to actually having snow! You missed only 3 or 4 degrees and a bit of percipation! (Btw, did I tell you that I slept outside last week with the night temperatures down to 7C?)
In the meantime, thank you to all of you for having learned a bit of spoken English since in the seaside resort where I live there are many English tourists. At least she understood me a little :-)
@DavidCarlisle Hi, there. Can I disturb, please, you a little bit for some clarification? I am currently using the translator.
@egreg -- Among people I've met, most seem to prefer meat either "well done", or cooked only just enough to be safe, but not both. I infer, though, regarding this aspect of cookery, you're omnivorous.
@barbarabeeton In Portuguese, 'rare' is mal passado 'badly done' as opposed to bem passado 'well done'. To me they are exactly the opposite of what they should be. :)
@AlanMunn -- Yes. In Portuguese restaurants, if my husband orders a steak, he will specify "muy mal passado", and always answers "yes" when asked to confirm. Even then, in an establishment where we're not known, it often is served barely pink. (Philosophy: the poor beast has already been killed. No need to kill it a second time.)
LaTeX Warning: Reference `enci' on page 4 undefined on input line 303.
LaTeX Warning: Reference `fig:subim7' on page 4 undefined on input line 306. Overfull \hbox (8.39914pt too wide) in paragraph at lines 312--312
[][]
[] LaTeX Warning: Reference `raggivet' on page 4 undefined on input line 328.LaTeX Warning: Reference `fig:subim4' on page 5 undefined on input line 354.
LaTeX Warning: Reference `Galileo' on page 5 undefined on input line 356.
LaTeX Warning: Reference `derivata' on page 5 undefined on input line 359.
@egreg Many of my mistakes in the book are mostly these. When I enter a \label{name} and have to refer to an equation I enter \eqref{name} or a table or figure. Some have a label but do not have a reference because I will probably quote them later following the revision of the book with my prof.
@egreg I have not compiled at the same time never twice but at subsequent times. If I don't bother you, I can send you the complete log code by email. For the figure I use \ref.
@AlanMunn well, "bleu" is yet less than "saignant". And the only reason why I don't like "bleu" is that if I wanted a cold steak, I would have ordered a "tartare" :)
@Sebastiano you don't need to run latex twice every time, if you have ever run latex previously then any labels that were defined then should be defined, so it is only new label that you have just added that are undefined until the next run.
Is there a canonical way to explicitly (not via \bgroup, \egroup which is implicit) refer to the current catcode 1,2 tokens? For example, \def\z{\mytoks{\fbox{A}}} works if defined when braces are the catcode 1,2 tokens. It works even after the 1,2 catcodes are redefined BUT, \mytoks still will contain catcode 1,2 BRACES rather than the newly defined catcode 1,2 delimiters.
:50949126 luatex/extex primitive but can be made a macro for pdftex, but might be tricky to make work for {}. Why do you want this? what's the use case where {} don't work?
@StevenB.Segletes for pdftex you can use \uppercase to make character tokens with any character code and catcode 1
@DavidCarlisle Well it works in the sense that it typesets correctly, but it is part of a parsing macro, and I have the case where, if someone changes the catcodes and calls on the parsing routine, it will convert his grouping tokens into catcode 1,2 braces, as part of the parse
@DavidCarlisle Your answer is helpful in that it sounds like, as long as you get the proper catcode 1,2 behavior, the code should be considered as "working".
@DavidCarlisle I have interrupt the complete compilation because the .aux file I don't know why it is lost or deleted from TeXworks when the compilation is complete.
@Sebastiano oh you said you did interrupt the compilation? well that will break the aux fiel so you will need to do two latex runs to get the cross refernces back.
@Sebastiano "same labels" ??? you can not have two things with the same label.
@DavidCarlisle Yes, of course. I thought of printing the code and then correcting it by hand. There is no order for the labels and you can imagine after 2 and a half years the concentration decreases. I promised myself to change them one by one.
Hi folks. I have a sort of general opinion type question. Over the last few years, I've but together an ad-hoc custom mini-referencing system, similar in spirit to BibTeX. But this just displays the integers where the references should appear. This is for correspondence, not papers/books. I've been using dates as a natural key.
@UlrikeFischer I'm sure I agree with you very much. Busy=occupied during compilation. But I blocked it. The .aux file is generated only when the compilation is either interrupted or there are serious errors that block the compilation.
@Sebastiano I do not see how you can get in that position, as soon as you add a label that clashes with an existing one, then you get a latex error. So fix that before writing any more document.
I was wondering if that key would be a better choice than integers. The difference is in "see [1]" vs "see [2019.06.01.foo]". While the former is shorter, the latter is more informative.
And it would save me having to guess what [1] meant every time I saw it.
@Sebastiano if your document is producing errors there is no point asking about why some output looks wrong, the output is nonsense and should not be used for anything. But you know that. we have said it enough times.
3
@Sebastiano No. the aux file is generated on every run, and required so it can be input on the next run.
Has anyone ever had occasion to dump Lua structures in a serialised format (whatever that might be) to the aux file?
I guess I should have written write, not dump.
And I don't see why Lua can't echo the freaking line where the freaking error is. Seriously, what is your problem? All the other languages can manage that.
@AlanMunn -- Ah! That's almost the equivalent of "black and blue", where the "black" refers to the outside and the "blue" to the inside. "Pittsburgh" is another term for that quality.
@FaheemMitha -- Not unpacked yet. All the boxes are in the basement. Unfortunately, the stairs to the basement are a bit shaky, and there's no handrail, and since I'm still shaky myself (and using a cane), getting there isn't safe. So it's a one-by-one operation.
@yo' -- I was making a quality comment. You're correct that "medium" is a proper culinary term, but for my taste, I can't think of anything that is appropriately prepared to "medium". ("Over medium heat", however, is a useful instruction.) On a different comment of yours, "bleu" implies that the meat is still in one piece; "tartare" implies that it has been minced, and no flame/heat has been applied. A solid uncooked chunk is just "raw".
@barbarabeeton I know the difference between bleu and tartare of course (it would be difficult not to after living in Paris). I just say that cold piece of meat is not something I like, and bleu is cold (well, certainly not warm). So while I like my steaks as low as possible, I want them warm.
@FaheemMitha -- That's why I'm still using a cane on uneven surfaces. I got a new titanium hip joint for a Christmas "present" in 2017 after an ice storm on December 23.
@barbarabeeton Steak tartare is though quite a different story: I love these, and the best is tartare de saumon, ideally with pommes frites maison, i.e. home-made fries, and the larger the better.
@barbarabeeton On the other hand, a “fiorentina” requires quite a long time next to the fire, but it's just to grill the exterior part. Since its height is at least 5–6 cm (2–2.4 in), the interior is still rare.
@DavidCarlisle -- It looks like at least two people are consistently downvoting. Of the last 15 most recently posted new questions, there are none that haven't been downvoted, and almost all have 2 downvotes. Is this something a moderator can investigate? (@JosephWright)
