MP graph auto axes

MP graph auto axes

[Adam Lindsay]

Hi Folks.

This is not strictly a ConTeXt problem, but I was hoping there were
enough MetaPost experts on the list that I might get an idea of what's
going on, or at the least, where to go for more information.

I think the MetaPost graphing features are fantastic. It produces very
nice output, and automatically handles a lot of axis labelling that other
packages don't bother with.

However, it auto-labels the y-axis on the following three graphs very
differently. I was wondering if this is a strict limitation in the axis
ranges that MetaPost deals with gracefully, or if it might be a bug. I
haven't seen any reference to the correct range for graph axes. I've
tried many ranges, and some work, some don't.

Would the MetaFont list be able to help?

\usemodule[graph]
%%  This one works as it should.
\startMPcode
draw begingraph(130mm,35mm);
  setrange(0,0,10,22000);
  glabel.lft(btex  {correct 0--22000} etex rotated 90, OUT);
  autogrid(itick.bot,grid.lft) withcolor .75white ;
  endgraph;
\stopMPcode

%% This one gives output, but yields incorrect labels
\startMPcode
draw begingraph(130mm,35mm);
  setrange(0,50,10,22000);
  glabel.lft(btex  {wrong 50--22000} etex rotated 90, OUT);
   autogrid(itick.bot,grid.lft) withcolor .75white ;
  endgraph;
\stopMPcode

%% This simply gives up.
\startMPcode
draw begingraph(130mm,35mm);
  setrange(0,110,10,22000);
  glabel.lft(btex  {failed 110--22000} etex rotated 90, OUT);
   autogrid(itick.bot,grid.lft) withcolor .75white ;
  endgraph;
\stopMPcode



Thanks in advance,
adam

--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
 Adam Lindsay                  +44(0)1524 594 537
 atl@comp.lancs.ac.uk
 http://www.comp.lancs.ac.uk/computing/users/atl/
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

Re: MP graph auto axes

[Hans Hagen]

At 07:13 PM 11/6/2002 +0000, you wrote:

>This is not strictly a ConTeXt problem, but I was hoping there were
>enough MetaPost experts on the list that I might get an idea of what's
>going on, or at the least, where to go for more information.
>
>I think the MetaPost graphing features are fantastic. It produces very
>nice output, and automatically handles a lot of axis labelling that other
>packages don't bother with.
>
>However, it auto-labels the y-axis on the following three graphs very
>differently. I was wondering if this is a strict limitation in the axis
>ranges that MetaPost deals with gracefully, or if it might be a bug. I
>haven't seen any reference to the correct range for graph axes. I've
>tried many ranges, and some work, some don't.

looks like some problem in interpreting the big numbers, (mp can handle 
only numbers <4000 and some string trickery is used in the graph package); 
does this also happens in stand alone graphics or is it context specific?

Hans
-------------------------------------------------------------------------
                                   Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                       Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
  tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------
                        information: http://www.pragma-ade.com/roadmap.pdf
                     documentation: http://www.pragma-ade.com/showcase.pdf
-------------------------------------------------------------------------

Re: MP graph auto axes

[Adam T. Lindsay]

Hans Hagen said this at Thu, 7 Nov 2002 09:55:29 +0100:

>At 07:13 PM 11/6/2002 +0000, you wrote:
>
>>However, it auto-labels the y-axis on the following three graphs very
>>differently. I was wondering if this is a strict limitation in the axis
>>ranges that MetaPost deals with gracefully, or if it might be a bug. I
>>haven't seen any reference to the correct range for graph axes. I've
>>tried many ranges, and some work, some don't.
>
>looks like some problem in interpreting the big numbers, (mp can handle 
>only numbers <4000 and some string trickery is used in the graph package); 
>does this also happens in stand alone graphics or is it context specific?

I'm not able to mpost things alone at the moment because of my earlier-
described texnum.mpx weirdness. So I run all my MP graphics from within
ConTeXt right now. When I get home, I'll try to investigate this side
further, though.

I'm aware of the MP number limit, but MP-graph has its stated (and
empirical) limits at 32768. Anyway, I tried variants with all numbers in
quotes. It made a difference with numbers over 2^15, but not in the test
cases that I sent.

It could well have to do with the big number limit, but I'm very
confident it's not that alone.

Thanks,
adam

-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 Adam T. Lindsay                      atl@comp.lancs.ac.uk
 Computing Dept, Lancaster University   +44(0)1524/594.537
 Lancaster, LA1 4YR, UK             Fax:+44(0)1524/593.608
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Re: MP graph auto axes

[Jens-Uwe Morawski]

On Thu, 07 Nov 2002 09:55:29 +0100
Hans Hagen <pragma@wxs.nl> wrote:

 
> looks like some problem in interpreting the big numbers, (mp can handle 
> only numbers <4000 and some string trickery is used in the graph package); 
> does this also happens in stand alone graphics or is it context specific?
> 

no, the problem is not specific to ConTeXt. The problem is in the
Gipick_ macro. There a logarithm is calculated:
n = 1 for i=1 upto
	mlog(xpart Gmhh_-xpart Gmll_)/Mten - mlog m/(Mten-epsilon):
      *10 endfor;

The problem is that the argument of mlog becomes <0, thus it is not
defined. Here is the error message for the 3rd graph:

! Logarithm of -1385.99995 has been replaced by 0.
Gipick_->...i=1upto.mlog(xpart.Gmhh_-xpart.Gmll_)/
                                                  Mten-mlog.m/(Mten-epsilon)...

auto->...fi.else:hide(Gme_:=Gesame_)Gigen_(Gipick_
                                                  ,Gcma_.Gpack_(m,Gme_))fi
<to be read again>

I spend some time yesterday, but the package is a little tricky
and uses some not intuitive variable names. so...

Jens

Re: MP graph auto axes

[Adam T. Lindsay]

Hi Jens. 

Thanks for those insights. Perhaps ironically (perhaps not), I don't
encounter the problem when I put the axes on a log-scale. 

Is MP-graph being actively maintained? Is this something that I could
push to get fixed?

(It *is* something I can work around in this particular application, I
think, but it would be nice if it worked in a predictable manner.)

Cheers,
adam



Jens-Uwe Morawski said this at Thu, 7 Nov 2002 14:36:50 +0100:

>On Thu, 07 Nov 2002 09:55:29 +0100
>Hans Hagen <pragma@wxs.nl> wrote:
>
> 
>> looks like some problem in interpreting the big numbers, (mp can handle 
>> only numbers <4000 and some string trickery is used in the graph package); 
>> does this also happens in stand alone graphics or is it context specific?
>> 
>
>no, the problem is not specific to ConTeXt. The problem is in the
>Gipick_ macro. There a logarithm is calculated:
>n = 1 for i=1 upto
>	mlog(xpart Gmhh_-xpart Gmll_)/Mten - mlog m/(Mten-epsilon):
>      *10 endfor;
>
>The problem is that the argument of mlog becomes <0, thus it is not
>defined. Here is the error message for the 3rd graph:
>
>! Logarithm of -1385.99995 has been replaced by 0.
>Gipick_->...i=1upto.mlog(xpart.Gmhh_-xpart.Gmll_)/
>                                                  Mten-mlog.m/(Mten-
>epsilon)...
>
>auto->...fi.else:hide(Gme_:=Gesame_)Gigen_(Gipick_
>                                                  ,Gcma_.Gpack_(m,Gme_))fi
><to be read again>
>
>I spend some time yesterday, but the package is a little tricky
>and uses some not intuitive variable names. so...
>
>Jens
>_______________________________________________
>ntg-context mailing list
>ntg-context@ref.ntg.nl
>http://ref.ntg.nl/mailman/listinfo/ntg-context

-- 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
 Adam T. Lindsay                      atl@comp.lancs.ac.uk
 Computing Dept, Lancaster University   +44(0)1524/594.537
 Lancaster, LA1 4YR, UK             Fax:+44(0)1524/593.608
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Re: MP graph auto axes

[Hans Hagen]

At 03:03 PM 11/7/2002 +0000, you wrote:
>Hi Jens.
>
>Thanks for those insights. Perhaps ironically (perhaps not), I don't
>encounter the problem when I put the axes on a log-scale.
>
>Is MP-graph being actively maintained? Is this something that I could
>push to get fixed?

since hobby published the manual in a recent tugboat ... i suggest that you 
send him a mail with a test.mp file demoing the problem; afaik normally 
he's responsive to real bugs (you can cc to me)

"John Hobby" <hobby@research.bell-labs.com>

Hans
-------------------------------------------------------------------------
                                   Hans Hagen | PRAGMA ADE | pragma@wxs.nl
                       Ridderstraat 27 | 8061 GH Hasselt | The Netherlands
  tel: +31 (0)38 477 53 69 | fax: +31 (0)38 477 53 74 | www.pragma-ade.com
-------------------------------------------------------------------------
                        information: http://www.pragma-ade.com/roadmap.pdf
                     documentation: http://www.pragma-ade.com/showcase.pdf
-------------------------------------------------------------------------