I often break long proofs into smaller pieces, in order to make it easier to read and understand. I'm doing that by adding a fbox
with "Part X: …". However, this doesn't work out very well all the time: the box and the according part of the proof can be split by a pagebreak, as illustrated in the code underneath.
\documentclass{article}
\usepackage{amsthm}
\usepackage{lipsum}
\parskip 12pt
\parindent 0pt
\newtheorem{theorem}{Theorem}
\begin{document}
\lipsum[1-4]
\begin{theorem}
This is my theorem.
\end{theorem}
\begin{proof}
Let's cut this long proof into little pieces.
\fbox{Part 1: firstly}
This is part 1.
\fbox{Part 2: secondly}
This is part 2.
\end{proof}
\end{document}
Is there a way to prevent this in general? Is it possible to create a new environment that prevents splitting and that makes adding such a subproof easier (I'm sick of having to type \fbox{Part X: ...}
all over again)? I'm looking for something that allows me to type for example
\begin{proof}
Let's cut this long proof into little pieces.
\begin{subproof}{Title of subproof 1}
(subproof 1)
\end{subproof}
\begin{subproof}{Title of subproof 2}
(subproof 2)
\end{subproof}
\end{proof}
and that would result in
Maybe Parts numbering in Proofs can help to number the subproofs automatically.
Best Answer
You could use the same mechanism as section titles:
Sorry, but I can't stand seeing nonzero parskip documents, that I find a waste of paper that adds nothing to clarity and readability.
A frame around “Part 1: firstly” is a punch in the eye, I suggest just italics like for “Proof”. In all honesty, I wouldn't add a line break after the proof part heading, but it's personal taste.
A variant that automatically numbers proof parts.