How to locally change counter for theorem-environment with referencing

[fjcmurphy]

Hello!

I'm writing on my thesis and would like to have the following: In some
chapter I use a theorem with
\begin{theorem} \label{thm:A}
...
this is the Theorem 7.1
...
\end{theorem}
Lets say it is now called "Theorem 7.1".

Then in a later chapter (lets say Chapter 10) I use the same theorem
again. I repeat it with
\begin{theorem}
...
this is the Theorem 7.1
...
\end{theorem}
so the reader does not have to change back and forth all the time. How
do I tell LaTeX, to print again "Theorem 7.1" instead of "Theorem
10.1"?

I hope my question is clear ;)


Cheers and thanks alot in advance,
frank

[Ulrich Diez]

fjcmurphy wrote:

...

In some chapter I use a theorem

...

Then in a later chapter ...
I repeat it with

...

so the reader does not have to change back and forth all the time.

My suggestion:

The number of the <theorem> comes from printing the <theorem>-
count-register's value via the macro \the<theorem>.
- Don't reset the counter-register but locally redefine the
associated \the<theorem>-macro.
- Save the expansion of \the<theorem> right after writing the
theorem the first time.
- When writing the theorem the second time, locally redefine
\the<theorem> to expand to that saved expansion.

In case you don't have changed referencing for your <theorem>-
counter, you can use \label and \ref for saving and regaining
\the<theorem>'s expansion. Otherwise you can define another
macro to expand to the desired values:

\documentclass{article}

\newtheorem{theorem}{Theorem}[section]%

\begin{document}
\section{test 1}

\begin{theorem}
This is the first theorem that shall be repeated.
\end{theorem}
\xdef\savedtheoremnumber{\thetheorem}%
% You may wish to first check if \savedtheoremnumber is already
% defined:
% \makeatletter\@ifdefinable\savedtheoremnumber{%
% \xdef\savedtheoremnumber{\thetheorem}%
% }\makeatother
% or:
% \newcommand\savedtheoremnumber{}%
% \xdef\savedtheoremnumber{\thetheorem}%

\section{test 2}

\begin{theorem}\label{thm:B}
This is the second theorem that shall be repeated.
\end{theorem}

% repeat the first theorem:
{%
\renewcommand\thetheorem{\savedtheoremnumber}%
\begin{theorem}
This is the first theorem that shall be repeated.
\end{theorem}
\addtocounter{theorem}{-1}%
}%

\begin{theorem}\label{thm:C}
This is another theorem.
\end{theorem}

% repeat the second theorem:
{%
\renewcommand\thetheorem{\ref{thm:B}}% Use \ref* instead with hyperref
\renewcommand\label[1]{\relax}%
\begin{theorem}\label{thm:B}
This is the second theorem that shall be repeated.
\end{theorem}
\addtocounter{theorem}{-1}%
}%

\begin{theorem}\label{thm:D}
This is yet another theorem.
\end{theorem}


\end{document}


Ulrich

[fjcmurphy]

Hello Ulrich!

Thanks alot!

The first version with
\xdef\savedtheoremnumber{\thetheorem}%

works for me. However, when I use
\usepackage{amsmath}
\usepackage{hyperref}
\usepackage[amsmath, thref,hyperref ]{ntheorem}

I can't use your second proposed way. I always get
--- TeX said ---
\***@ifnextchar ***@d = #1\def \***@a {
#2}\def \***@b
{#3}\f...

while compiling. Moreover, what does the opional "[1]" in
\label[1]{\relax} do?

Thanks alot.

All the best,
frank

[Ulrich Diez]

fjcmurphy wrote;

Post by fjcmurphy
Hello Ulrich!
Thanks alot!
The first version with
\xdef\savedtheoremnumber{\thetheorem}%
works for me. However, when I use
\usepackage{amsmath}
\usepackage{hyperref}
\usepackage[amsmath, thref,hyperref ]{ntheorem}
I can't use your second proposed way. I always get

With this preamble referencing is changed.

Post by fjcmurphy

Post by Ulrich Diez
In case you don't have changed referencing for your <theorem>-
counter, you can use \label and \ref for saving and regaining
\the<theorem>'s expansion.

As referencing is changed, the inverse-case is not given
any more. So stuff of course might not work as expected
any more when going this route.
You can use another \ref-like macro instead which extracts
only the reference-number (no hyperlinks and stuff).
There was plenty of discussion about this recently/yesterday
at comp.text.tex.
I recommend using Heiko Oberdiek's refcount-package.

...

Post by fjcmurphy
Moreover, what does the opional "[1]" in
\label[1]{\relax} do?

It's not optional.
It's: \renewcommand\label[1]{\relax}
In case missing braces confuse you:
It's equal to: \renewcommand{\labe}l[1]{\relax}

This means: \label is redefined to "eat" it's argument and to
spit out \relax. For writing your theorem the second time, you
might use the "edit-copy-paste"-feature of your editor.
You might easily omit to remove \label-s placed into the theorem.
This might lead to erroneously getting multiple labels
with equal name. This can be prevented by locally redefining
label to do nothing or to \relax.

Ulrich

[Ulrich Diez]

Post by Ulrich Diez
It's not optional.
It's: \renewcommand\label[1]{\relax}
It's equal to: \renewcommand{\labe}l[1]{\relax}

Correction:

It's equal to: \renewcommand{\label}[1]{\relax}

Sorry for the typo.

Ulrich

[fjcmurphy]

Thanks alot for all the clarification.

Cheers,
frank