1

I hope this question actually is in order since it does not seem to have a "correct" answer:

I am about to get into drawing game-theoretic trees and while I am familiar with LaTeX in general, I am not so much with TikZ and particularly not with the drawing of trees.

From what I found, tikz-qtree and forest seem to be the two most common packages to be used (please correct me if I am mistaken). Since I will have to use the syntax of one either way, my question is: which one of the two to go for. Is one considered easier to use, more widespread/stable, more flexible, etc. - or are they not actually to be compared?

Bernd
  • 1,785
  • 1
    They can be compared, but I fear you have to do that by yourself. I would search for some posts and see, which syntax seems easier for you. And you could glance into their manuals. Often the most important factor regarding quality of a package. To answer your to points: tikz-qtree is more widespread, forest maybe more flexible (but they are both TikZ, so flexibility is not really limited...) – LaRiFaRi Aug 18 '15 at 10:18
  • If you find something interesting during your comparison, you could share that in this actual post: http://meta.tex.stackexchange.com/q/6248 – LaRiFaRi Aug 18 '15 at 10:27
  • 2
    Given that game-theoretic trees seem to require lots of labelled branches and may be constructable algorithmically, I think that forest would be a better choice than tikz-qtree, which was designed primarily for the needs associated to drawing linguistic trees. As @LaRiFaRi notes, forest is considered to be more flexible. This is true mainly because it provides simple ways to access the full flexibility of TiKZ within a tree; in tikz-qtree it's usually a little harder to access. – Alan Munn Aug 18 '15 at 12:57
  • @AlanMunn - thanks, that is quite a helpful comment! – Bernd Aug 18 '15 at 12:58
  • I'm not really familiar with tikz-qtree having mostly used qtree or forest, but I prefer forest's notation to qtree's as I find it easier to relate the notation to the structure of the tree. (Linguists may not find this but possibly because qtree's notation is based on a discipline standard? Not sure.) For game-theoretic trees and tree proofs in logic, I think forest can be better, at least if your trees will get complicated. However, I think qtree's manual is more accessible if you are starting this from scratch. – cfr Aug 18 '15 at 13:08
  • 5
    A brief introduction to forest's notation for trees and a few selected features is included in the second part of my answer to an earlier question. That might possibly be of some help in getting started with forest since the manual is a bit intimidating. Once you have the basic idea, the manual is really rather good, especially combined with examples from this site. – cfr Aug 18 '15 at 13:12
  • This question: What is a more efficient way to draw this tree? provides a nice comparison of the differences between tikz-qtree and forest on a tree of the sort you are asking about. You can see that the tikz-qtree version is much clunkier for this sort of tree. In fact, perhaps your question can be closed as a duplicate of this one? – Alan Munn Aug 18 '15 at 16:30
  • @AlanMunn Yes, I guess the question can be closed. I guess cfr's answer here is pretty close to what I was looking for. I just couldn't find a good source of information to get started with game trees but I guess that other thread should do for others in the future (the question being whether my question is a duplicate of the one you mention or rather this one). Thanks for the help anyway! – Bernd Aug 18 '15 at 16:49
  • 1
    The reason I suggested the other question is that @cfr 's question is specifically related to linguistic trees and the one I linked to seems closer to what many game theoretic tree requirements are like. (But maybe I'm mistaken.) What we can do is each chose one answer when we close and then both will appear linked to this question. – Alan Munn Aug 18 '15 at 17:02
  • @AlanMunn I did not mean the link I posted to be a candidate duplicate. I only posted it for the reason I explicitly stated! That is, if the OP wants to try forest, the notes there might be helpful to get started. It isn't a good question for help deciding which package to use to draw trees of this sort, and I didn't mean to suggest otherwise. – cfr Aug 18 '15 at 20:26
  • @cfr Yes, I know that. But the closing troops tend to act swiftly. I've got Joseph to change it to the other question. Your question is indeed a nice intro to using forest. But your comment here has 5 votes, so it will float to to top and others will see it I think. – Alan Munn Aug 18 '15 at 20:49

1 Answers1

0

For simple game trees Martin Osborne's egames is good enough. It is also flexible up to some degree.

Check the examples to see if you can modify them to your satisfaction: https://www.economics.utoronto.ca/osborne/latex/egameps.pdf

It really is a matter of taste.

Bayesian
  • 245
  • 3
    Could you perhaps expand this answer with an actual example document rather than just a link to the documentation? Also, this package is PSTricks based, which requires compilation with latex+dvips or xelatex rather than pdflatex. (That's not a criticism, but it would be helpful to note.) – Alan Munn Aug 18 '15 at 13:01
  • The documentation provides many examples that can easily be manipulated. What do you mean? Moreover, I recommended this package to other scolars who did not appear to have problems with pdflatex. But thanks for sharing this caveat - I was not aware of that! – Bayesian Aug 18 '15 at 15:30
  • 1
    One of the goals of the StackExchange sites is to be self-contained repositories of information. So answers that are just links are less useful than those that that provide content independently. And links can die easily, so just posting a link isn't usually the best practice. (On pdflatex issue: the documentation clearly states in the second paragraph "Because the style uses the PSTricks macros, it is incompatible with pdftex". – Alan Munn Aug 18 '15 at 15:58