[pstricks] log plot

[Patrice Megret]

Dear Herbert,

I have just loaded the latest beta version on pstricks website (version 2.70 2005/10/18) and here is the code below. You can see that there is a test on \ifx\psk at Ox\@empty ... in the test loop for the y axis. This causes a crash if Oy={} without specifying Ox.

Best regards,

Patrice



\def\psaxes at iv(#1,#2)(#3,#4)(#5,#6){%
  \pst at killglue%
  \begingroup%
  \setbox\pst at hbox=\hbox\bgroup%
  \use at par%
  \pssetxlength\pst at dimg{#1}% o-x
  \pssetylength\pst at dimh{#2}% o-y
  \pssetxlength\pst at dima{#3}% bl-x
  \pssetylength\pst at dimb{#4}% bl-y
  \pssetxlength\pst at dimc{#5}% ur-x
  \pssetylength\pst at dimd{#6}% ur-y
    % If minimum values are negative in log mode, we modify Ox
    % (respectively Oy) if this was not done by the user
    % X axis labels (\psk at log = 0 or 1)
  \ifx\psk at xlogBase\@empty\else%
    \ifx\psk at Ox\@empty\else% we do nothing ... hv
      \ifdim\psk at Ox pt=\z@
        \ifdim#3pt<\z@
          \pssetxlength\pst at dimg{#1}% o-x
          \psset{Ox=#1}%
	\fi
      \fi%
    \fi%
  \fi%
% Y axis labels (\psk at log = 0 or 2)
  \ifx\psk at ylogBase\@empty\else%
    \ifx\psk at Ox\@empty\else
      \ifdim#4pt<\z@
        \ifdim\psk at Oy pt=\z@
          \pssetylength\pst at dimh{#2}% o-y
          \psset{Oy=#2}%
	\fi
      \fi
    \fi
  \fi
% Whole thing will be translated to origin:
  \advance\pst at dima-\pst at dimg% Dist. from bl-x to o-x
  \advance\pst at dimb-\pst at dimh% Dist. from bl-y to o-y
  \advance\pst at dimc-\pst at dimg% Dist. from ur-x to o-x
  \advance\pst at dimd-\pst at dimh% Dist. from ur-y to o-y
% Make lines/arrows or frame:
  \@nameuse{psxs@\psk at axesstyle}%  \psxs at axes or \psxs at frame
  \ifPst at plot@yAxis%
  \begingroup%
    \ifdim\pst at dima=\z@\else\showoriginfalse\fi%
    \ifnum\psk at dy=\z@%
      \pst at dimg=\psk at Dy\psyunit%
      \edef\psk at dy{\number\pst at dimg}%
    \fi%
    \ifPst at plot@xAxis\else\showorigintrue\fi%
    \pst at vlabels{\pst at dimd}{\psk at arrowB}{#3}{#5}%
    \ifPst at plot@xAxis\showoriginfalse\fi%
    \pst at vlabels{\pst at dimb}{\psk at arrowA}{#3}{#5}%
  \endgroup%
  \fi%
  \ifPst at plot@xAxis%
  \begingroup%
    \ifdim\pst at dimb=\z@\else\showoriginfalse\fi%
    \ifnum\psk at dx=\z@%
      \pst at dimg=\psk at Dx\psxunit%
      \edef\psk at dx{\number\pst at dimg}%
    \fi%
    \ifPst at plot@yAxis\else\showorigintrue\fi%
    \pst at hlabels{\pst at dimc}{\psk at arrowB}{#4}{#6}%
    \ifPst at plot@yAxis\showoriginfalse\fi%
    \pst at hlabels{\pst at dima}{\psk at arrowA}{#4}{#6}%
  \endgroup%
  \fi%
% Now close "\pst at hbox" (which is 0-dimensional), and put it at the origin.
  \egroup%
  \pssetxlength\pst at dimg{#1}%
  \pssetylength\pst at dimh{#2}%
  \leavevmode\psput at cartesian\pst at hbox%
  \endgroup
  \ignorespaces%
}


 
Prof Dr Ir Patrice Mégret
Head of Electromagnetism and Telecommunication Department
Faculté Polytechnique de Mons
Boulevard Dolez 31
B-7000 MONS 
 
-----Message d'origine-----
De : pstricks-bounces at tug.org [mailto:pstricks-bounces at tug.org] De la part de Herbert Voss
Envoyé : lundi 24 octobre 2005 16:25
À : pstricks at tug.org
Objet : Re: [pstricks] log plot

Patrice Megret wrote:

> By the way, if you prefer the way it is working now, could you just 
> change the test in line 1447 which is \ifx\psk at Ox\@empty ... instead of 
> \ifx\psk at y\@empty and causes thus a crash when Oy={} is specified and 
> check the order of the tests which are different for x and y.

Patrice,
I didn't found such a line in the latest version.

Herbert

_______________________________________________
pstricks mailing list
pstricks at tug.org
http://tug.org/mailman/listinfo/pstricks