gnuplot-beta

Tics, minor tics, labels...

[Ethan A M. <merritt@u.washington.edu>]

Two questions:

1) The documentation says you can assign labels to minor tics
      set xtics ("Major" x1 0, "Minor" x2 1, ...)
   but the code explicitly ignores minor tic labels.  Was there any
   reason why?  You can always give an empty string if you don't
   want a label there. I propose that if the user specifies a label,
   we should print it.

2) I am inclined to apply the patchset currently listed as #1104018,
   having worked with the original proposer, Eddie Kohler, to pare it
   down to a very simple modification to the existing code.
   In a nutshell, it now adds a keyword "include" to the 'set <xyz>tics'
   command. In the presense of this keyword, the command does
   not clear the existing list of user-specified tics (if any).
   So
	set xtics auto
	set xtics include ("Pi" pi 0)
	set xtics include ("2Pi" pi*2 0)
   will auto-generate tics and include specific labels at pi and 2pi.
   Omitting the keyword retains the current behaviour; each
   command supersedes the previous ones.

   Please speak up if you have any objections or counter-proposals.
   Also please suggest a better name for the keyword if you can think
   of one.  "add" is the another possibility on the table at the
   moment.


-- 
Ethan A Merritt
Biomolecular Structure Center
University of Washington, Seattle 98195-7742

Re: Tics, minor tics, labels...

[Robert H. <en...@no...>]

On Sat, 13 Aug 2005, Ethan A Merritt wrote:

> 	set xtics auto
> 	set xtics include ("Pi" pi 0)
> 	set xtics include ("2Pi" pi*2 0)
>    will auto-generate tics and include specific labels at pi and 2pi.
>    Omitting the keyword retains the current behaviour; each
>    command supersedes the previous ones.
> 
>    Please speak up if you have any objections or counter-proposals.
>    Also please suggest a better name for the keyword if you can think
>    of one.  "add" is the another possibility on the table at the
>    moment.

I suppose the other way of looking at is:

a) Is this a syntax that could consistenly be applied to other commands
should this kind of "addition" behaviour be applicable to them. 

Similar commands are things like "set label", which IIRC behaves in a much
different way - no extra keyword is needed. Of course, using the same sort
of syntax as set label would break any existing scripts that contain
multiple explicit tics statements, which I imagine are relatively
uncommon.

b) Is it actually the "verb" that needs to be changed, as in:
unset xtics
set xtics auto
add xtics ("pi" pi 0)
add xtics ("2pi" 2*pi 0)
show xtics

This is, I guess, more akin to the plot and replot case in the current
syntax. I'm guessing (although I haven't really looked at the code) that
this would be more trouble to implement.

Rob


This message has been checked for viruses but the contents of an attachment
may still contain software viruses, which could damage your computer system:
you are advised to perform your own checks. Email communications with the
University of Nottingham may be monitored as permitted by UK legislation.

Re: Tics, minor tics, labels...

[Ethan A M. <merritt@u.washington.edu>]

On Sunday 14 August 2005 03:38 pm, you wrote:
> On Sat, 13 Aug 2005, Ethan A Merritt wrote:
> > 	set xtics auto
> > 	set xtics include ("Pi" pi 0)
> > 	set xtics include ("2Pi" pi*2 0)
> >    will auto-generate tics and include specific labels at pi and 2pi.
>
> Similar commands are things like "set label", which IIRC behaves in a
> much different way - no extra keyword is needed. Of course, using the
> same sort of syntax as set label would break any existing scripts that
> contain multiple explicit tics statements, which I imagine are
> relatively uncommon.

Yeah. This is where the goal of backwards compatibility runs headlong
into the goal of fixing design bugs.  I think it is a bug that successive
user tic specifications are not cumulative.  But that is not nearly so 
clear for the case of switching between automatic and manual tic placement.
If we were to make the default behaviour cumulative, then we would have to
add a mechanism to explicitly turn *off* the previous tic scheme before
selecting a new one:
	set xtics auto
	plot "foo"
	set xtics noauto
	set xtics ("Manual placement" ...)
	replot

That would definitely break backwards compatibility.


> b) Is it actually the "verb" that needs to be changed, as in:
> unset xtics
> set xtics auto
> add xtics ("pi" pi 0)
> add xtics ("2pi" 2*pi 0)
> show xtics

I like that idea, but it's too radical a change to the syntax.
If we were to introduce "add" as a major command, people
would expect to be able to use it for other purposes as well.

-- 
Ethan A Merritt
Biomolecular Structure Center
University of Washington, Seattle 98195-7742

Re: Tics, minor tics, labels...

[Daniel J S. <dan...@ie...>]

Ethan A Merritt wrote:
> Two questions:
> 
> 1) The documentation says you can assign labels to minor tics
>       set xtics ("Major" x1 0, "Minor" x2 1, ...)
>    but the code explicitly ignores minor tic labels.  Was there any
>    reason why?  You can always give an empty string if you don't
>    want a label there. I propose that if the user specifies a label,
>    we should print it.

Might as well.  Why limit flexibility?  And it may be easier to have it than to disallow it.

> 2) I am inclined to apply the patchset currently listed as #1104018,
>    having worked with the original proposer, Eddie Kohler, to pare it
>    down to a very simple modification to the existing code.
>    In a nutshell, it now adds a keyword "include" to the 'set <xyz>tics'
>    command. In the presense of this keyword, the command does
>    not clear the existing list of user-specified tics (if any).
>    So
> 	set xtics auto
> 	set xtics include ("Pi" pi 0)
> 	set xtics include ("2Pi" pi*2 0)
>
>    Please speak up if you have any objections or counter-proposals.
>    Also please suggest a better name for the keyword if you can think
>    of one.  "add" is the another possibility on the table at the
>    moment.

"with" ?  e.g.,

set xtics with ("Pi" pi 0)

Dan