| Re: Writing eps files from iplot [message #69069] |
Wed, 23 December 2009 09:10  |
John Mariska
Messages: 4
Registered: November 2003
|
Junior Member |
|
|
pp wrote:
> On Dec 23, 2:00 pm, John Mariska <mari...@nrl.navy.mil> wrote:
>> Producing decent eps plots for publication requires extensive control
>> over all aspects of the plot, such as the x- and y-sizes, the font
>> sizes, etc. I've always exercised this control using direct graphics.
>> The downside, of course, is what you see on the screen is not what ends
>> up in the eps file when you change the device--something we have all
>> asked to be fixed for so many years I've lost count. Once or twice a
>> year, I try to use itools to do publication quality plots. I just tried
>> again, and things have improved a bit, but not enough. Does anyone know
>> how to specify the x- and y-sizes of the bounding box in an eps file
>> written from an iplot window? All I can seem to get is 6"x4".
>>
>> John M.
>
> All publication plots I made in recent years were done with the
> itools. I choose the aspect ratio with the dimensions keyword. The
> bounding box is the same as the window area (assuming you are
> exporting the whole window), so the placement of things in it is
> determined by the margins used.
>
> With only vector elements in the plot, all that matters is the aspect
> ratio and the relative dimensions of the things in the plot (fonts,
> line widths, symbol sizes, margins), not an absolute size, since
> afterwards the eps can be rendered at any scale.
Relative dimensions do help quite a bit, but it is frustrating that the
units are device units. If one can work in cm or inches, then deciding
on things like font sizes and other layout parameters is much easier.
One does have to get all the plot elements with the right relative sizes
so that they all scale well together.
I'm also still trying to figure out how to do a filled circle for data
points in an iplot.
|
|
|
|
| Re: Writing eps files from iplot [message #69070 is a reply to message #69069] |
Wed, 23 December 2009 08:44  |
penteado
Messages: 866
Registered: February 2018
|
Senior Member
Administrator |
|
|
On Dec 23, 2:00 pm, John Mariska <mari...@nrl.navy.mil> wrote:
> Producing decent eps plots for publication requires extensive control
> over all aspects of the plot, such as the x- and y-sizes, the font
> sizes, etc. I've always exercised this control using direct graphics.
> The downside, of course, is what you see on the screen is not what ends
> up in the eps file when you change the device--something we have all
> asked to be fixed for so many years I've lost count. Once or twice a
> year, I try to use itools to do publication quality plots. I just tried
> again, and things have improved a bit, but not enough. Does anyone know
> how to specify the x- and y-sizes of the bounding box in an eps file
> written from an iplot window? All I can seem to get is 6"x4".
>
> John M.
All publication plots I made in recent years were done with the
itools. I choose the aspect ratio with the dimensions keyword. The
bounding box is the same as the window area (assuming you are
exporting the whole window), so the placement of things in it is
determined by the margins used.
With only vector elements in the plot, all that matters is the aspect
ratio and the relative dimensions of the things in the plot (fonts,
line widths, symbol sizes, margins), not an absolute size, since
afterwards the eps can be rendered at any scale.
|
|
|
|
| Re: Writing eps files from iplot [message #69253 is a reply to message #69069] |
Thu, 24 December 2009 09:20 |
JM[1]
Messages: 8
Registered: October 2009
|
Junior Member |
|
|
Years ago I wrote (for myself) a set of routines to work transparently
between X-windows and the PS device. They were written in total
ignorance of David Fanning's routines (and no doubt all the poorer for
that), but I've carried on using them because they cover my needs.
Just like with David's routines, there's a 'head' and a 'tail'
procedure that goes into every plotting procedure. The end-goal was
nearly always either hardcopy or PDF files.
Put simply, the user specifies all page, plot and image dimensions in
cm and, in the resulting PS/EPS file, the dimensions will be exactly
what the user specified. Any windows that are displayed are scaled to
very nearly match the specified dimensions - there's an associated
image display routine that resamples images for display to the window
device, but of course 'prints' the image at the full resolution to the
PS device.
The extra coding is minimal, and now it's second nature to work up my
plots in cm space - I find that especially useful for multiple plots
Your welcome to the code - let me know.
-John Mardaljevic
PS. Numerous examples of plots to exact cm co-ordinates in thesis
chapters 3 to 6 here:
http://www.iesd.dmu.ac.uk/~jm/doku.php?id=resources:thesis
|
|
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
