Anyone know how to put long proofs in Beamer onto two or more frames (since it won't fit onto one)? allowframebreaks along with \framebreak or \break doesn't seem to work, and I can't find anything online to help.
What do you mean by "long proofs"? Is the problem that you want to use the {proof} environment? I just title the appropriate frames "Proof of Theorem 3". The {proof} environment is not the only way to declare that some block of text is a proof, and I'm pretty sure that some set of AMS guidelines (can't remember which one) recommend confining it to short proofs.
Yes I want to use the AMS environment, but to have the proof extend to two frames. I did read a rather 'halfassed' way to solve this by just ending the proof on the first frame (getting rid of the square) and then in the next frame start it off with \begin{proof}{proof continued}. That's fine for me, but now out of pure curiosity I'm wondering how to genuinely extend proofs beyond one frame.
If you're putting long proofs in a slide talk then you're going to give a bad talk. Don't prove things. Describe your theorem clearly, then give the most basic sketch of the proof using the most basic ideas.
The ones who could follow your long proof will follow your proof sketch, and the ones who couldn't will thank you for it.
Tom, this is a not a traditional talk. It's a 3 lecture series, and I was told by the one who invited me to first discuss the basic properties of what I study since this is aimed towards grad students. My proofs are totally elementary and easy to follow, but some don't fit in one frame.
Anyway normally I totally agree with you and usually only throw in one or two almost trivial proofs in a talk. Talk and chalk might I know is more appropriate if I'm proving things, but I'm terrible at that, and this won't change before the talk.
Thanks so much for the clarification, David.1 I'm sorry for my snappy response. You're absolutely right that in a short lecture series, proofs are appropriate.
What's the lecture series on?
1. I couldn't find your name so I'll just assume you're the reincarnation of Hilbert himself
It might actually make more sense to forget the proof environment altogether and just manually insert the QED symbol at the end of your proof (which is now going to just be a bunch of text rather than something in a proper environment).
The thing is that proofs in Beamer get inserted a block environment and that's why allowframebreaks doesn't work: as far as the frame is concerned, there is just one large object on it (the block), so there's nothing to break.
If your proof is longer than one frame, this means that you want to end up with a block environment that spans an entire frame, and that doesn't really make much sense in terms of layout: blocks are designed to single out a part of the content -- if the content of the block is the entire frame, what are you singling out exactly?
Taking a look at the documentation, it looks like the beamer proof environment (yes, beamer defines its own proof environment) is set up to play nicely with overlays. The play-nicely-with-overlays code packages proofs into unbreakable blocks, because this makes it easier to figure out how much space to leave on the overlay slides where the proof doesn't show up.
allowframebreaks is not set up to play nicely with overlays (for good reason). Unfortunately, allowframebreaks does not turn off the play-nicely-with-overlays code, so proofs are still unbreakable.
So there's no "real" way to do what you want; I just thought you might want to know why.
Pedagogically, I think it's probably worthwhile to have a "Proof (continued)" or "Proof of Theorem 3" on every frame of the proof, so you don't actually want the raw proof environment. You want a separate breakable proof environment, which apparently the beamer team has not written. The same probably goes for other unbreakable things.
Comments 7
Reply
Reply
The ones who could follow your long proof will follow your proof sketch, and the ones who couldn't will thank you for it.
Reply
Anyway normally I totally agree with you and usually only throw in one or two almost trivial proofs in a talk. Talk and chalk might I know is more appropriate if I'm proving things, but I'm terrible at that, and this won't change before the talk.
Reply
What's the lecture series on?
1. I couldn't find your name so I'll just assume you're the reincarnation of Hilbert himself
Reply
The thing is that proofs in Beamer get inserted a block environment and that's why allowframebreaks doesn't work: as far as the frame is concerned, there is just one large object on it (the block), so there's nothing to break.
If your proof is longer than one frame, this means that you want to end up with a block environment that spans an entire frame, and that doesn't really make much sense in terms of layout: blocks are designed to single out a part of the content -- if the content of the block is the entire frame, what are you singling out exactly?
Anyway, just my 2c.
Reply
allowframebreaks is not set up to play nicely with overlays (for good reason). Unfortunately, allowframebreaks does not turn off the play-nicely-with-overlays code, so proofs are still unbreakable.
So there's no "real" way to do what you want; I just thought you might want to know why.
Pedagogically, I think it's probably worthwhile to have a "Proof (continued)" or "Proof of Theorem 3" on every frame of the proof, so you don't actually want the raw proof environment. You want a separate breakable proof environment, which apparently the beamer team has not written. The same probably goes for other unbreakable things.
Reply
Leave a comment