[texhax] Under-Sized Angle Brackets

Mc Kiernan Daniel Kian mc-kiernan at att.net
Thu Sep 3 01:56:51 CEST 2009


hh wrote:

 >> Mc Kiernan, Daniel Kian, wrote:
 >>
 >>>>> I'm attempting to produce formulae some of which have large
 >>>>> angle brackets as delimiters.  Some of these should be, say, 9
 >>>>>  lines tall.  But, when pdfeTeX renders them, none are a full
 >>>>> three lines tall.  Other delimiters (parentheses, brackets, and
 >>>>> braces), large and small, are rendering fine.

<snip>

 >>>> minimal example please
 >>>
 >>> Here's the first formula for which there is a problem:
 >>>
 >>> \begin{equation}
 >>> \left\langle \begin{matrix}\overline{{B_{1}\;\subseteq\;
 >>> C\left(B_{1}\;\cup\; B_{2}\;\cup\; B_{3}\right)}}\;\forall B_{3}\\
 >>> \vee\\
 >>> \left\{ \begin{matrix}\left[B_{1}\;\subseteq\;
 >>> C\left(B_{1}\;\cup\; B_{2}\;\cup\; B_{4}\right)\right]\\
 >>> \vee\\
 >>> \overline{{B_{2}\;\subseteq\; C\left(B_{1}\;\cup\; B_{2}\;\cup\;
 >>> B_{4}\right)}}\end{matrix}\right\} \forall
 >>> B_{4}\end{matrix}\right\rangle \forall\left(B_{1},B_{2}\right)
 >>> \end{equation}

> it seems that the command \left\lange and \right\rangle can produce bigger 
> brackets, but not indefinitely (as e.g. "\left\{" can). In the (old, from 1991) 
> book from Schwarz, Einfuehrung in TeX, only 
> "(, \{, \Vert, \lgroup and [" 
> are mentioned as growing. In your case this even does not allow the use of 
> \langle+\rangle, because they do not even cover 3 baselineskips.
> 
> Even the use of the \matrix command seems unreasonable.
> 
> I tried the building of the formula with "array" and the "rput" command of 
> pstricks. The outcome (below) looks a bit curious and might need some 
> better finish. See also the attachment (ps- and pdf-file). But it gives the idea.
> 
> {
> \thinmuskip=4mu                     % author wants wider spaces in math formula
> \medmuskip=6mu plus 2mu minus 2mu
> \thickmuskip=10mu plus 5mu
> %needs \usepackage{pstricks}
> 
> \noindent
> xxxxxxxxxxxxxxxxxxxx
> \begin{equation}
> \left\Vert 
> \begin{array}{@{\hspace{1mm}}l @{\hspace{1mm}} l @{\hspace{2mm}}l 
> @{}}
> &\hphantom{\mathopen [}\overline{B_1\subseteq C(B_1\cup B_2\cup 
> B_3)}{} \,\, \forall B_3\\
> &\hfil\vee\\
> &\mathopen [B_1\subseteq C(B_1\cup B_2\cup B_4) \mathclose]\\
> \rput[l](0,0){\left\{ \vrule height 8mm depth 0pt width0pt\right.}\,& \hfil\vee&
> \rput[l](-1,0){\left. \vrule height 8mm depth 0pt width0pt\right\} \forall B_4}\,\\
> &\hphantom{\mathopen [}\overline{B_2\subseteq C(B_1\cup B_2\cup B_4)}  
> \end{array}
> \right\Vert
> \forall(B_1,B_2)
> \end{equation}
> xxxxxxxxxxxxxxxxxxx

Thank you for the information and for the suggested response.  In the 
context of the former, I think that I should simply remove the angle 
brackets from the hierarchy of delimiters; use of the vrule or any 
other new delimiter is likely to confuse my readers.


More information about the texhax mailing list