Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.graphics.apps.gnuplot > #3695 > unrolled thread

A couple of questions

Back to article view | Back to comp.graphics.apps.gnuplot

Contents

  A couple of questions layzarc@aol.com - 2017-07-24 12:50 -0700
    Re: A couple of questions Ethan A Merritt <EAMerritt@gmail.com> - 2017-07-24 21:16 -0700
      Re: A couple of questions layzarc@aol.com - 2017-07-25 07:02 -0700
        Re: A couple of questions Ethan A Merritt <EAMerritt@gmail.com> - 2017-07-25 08:33 -0700
          Re: A couple of questions layzarc@aol.com - 2017-07-25 14:34 -0700
            Re: A couple of questions Ethan A Merritt <EAMerritt@gmail.com> - 2017-07-25 18:59 -0700

#3695 — A couple of questions

1. When was "set autoscale noextend" implemented? I know 4.2 doesn't like it.

2. I'm getting some extraneous messages now that don't crop up in 4.6.
   "splot 'file' binary" works ok and no messages
   "splot 'file' binary with image" works ok but I get the messages
     No dimension information for ... pixels total. Try ...  

Al

[toc] | [next] | [standalone]

#3696

layzarc@aol.com wrote:

> 1. When was "set autoscale noextend" implemented? I know 4.2 doesn't
> like it.

  The feature itself has been there all along, but the required syntax
  was non-obvious

	set auto xfix

  "noextend" was added as a synonym so that you could use a more easily
  remembered command

	set xrange [*:*] noextend
	
  You can still use the older form if you prefer.


> 2. I'm getting some extraneous messages now that don't crop up in 4.6.
>    "splot 'file' binary" works ok and no messages
>    "splot 'file' binary with image" works ok but I get the messages
>      No dimension information for ... pixels total. Try ...


  Not enough information to answer that one.  
  Is your binary file in some standard image format (png jpeg etc)?
  Did you tell gnuplot that?
  If not, where are you expecting gnuplot to find the image dimensions?


> Al

[toc] | [prev] | [next] | [standalone]

#3697

Ethan,

1. Thanks for "xfix" synonym. I wanted to make sure my files worked with older versions.
2. The format of my binary (matrix) file as described in the manual:
The second interpretation assumes a non-uniform grid with explicit x and y coordinates. The 1st row of input data contains the y coordinates; the 1st column of input data contains the x coordinates. For binary input data, the 1st element of the 1st row must contain the number of columns. This is the default for binary matrix input.

Al

[toc] | [prev] | [next] | [standalone]

#3698

layzarc@aol.com wrote:

> Ethan,
>2. The format of my binary (matrix) file as
> described in the manual: The second interpretation assumes a
> non-uniform grid with explicit x and y coordinates. The 1st row of
> input data contains the y coordinates; the 1st column of input data
> contains the x coordinates. For binary input data, the 1st element of
> the 1st row must contain the number of columns. This is the default
> for binary matrix input.

That's the default _matrix_ format.  It says nothing about images.
The pixels in a 2D image constitute a uniform rectangular grid,
but as you just quoted the binary matrix format in general describes a
non-uniform grid.   So "with image" is not an option for this data.

If the points are spaced closely enough you may get the effect you
want by using "with points pointtype 5".  That draws each of your
points as an individual square "pixel".  You can adjust the point size
to fill in gaps or reduce overlap.

	Ethan

[toc] | [prev] | [next] | [standalone]

#3699

In most cases the data is on uniform grid and as I said, I get the expected image so "binary matrix with image" does work.  I just get the extra message which were not present in earlier versions.  Anyway its not a big deal.  I was just wondering what changed in later versions to generate the messages.

[toc] | [prev] | [next] | [standalone]

#3700

layzarc@aol.com wrote:

> In most cases the data is on uniform grid and as I said, I get the
> expected image so "binary matrix with image" does work.  I just get
> the extra message which were not present in earlier versions.  Anyway
> its not a big deal.  I was just wondering what changed in later
> versions to generate the messages.

Earlier versions deferred interpretation of image data until the
first few pixels were about to be rendered.  The on-screen coordinates
of those pixels were inspected and some empirical rules were applied to
[figure out/guess] what the image dimensions were supposed to be.
These empirical rules were not always right and, worse, they were
reapplied each time the image was rendered.  That meant that if you
rotated a 3D plot  containing image data it could flip around and
distort as you dragged the mouse.

Starting with version 5, image data is instead inspected at the time
it is read in to determine the dimensions from the input data itself
rather than from the rendered coordinates on the screen.  The dimensions
are then stored with the image and not re-determined when the image
is later rendered.

If the image dimensions are ambiguous when read in, as in your case,
gnuplot will make a guess and issue a warning.  That's probably what
you are seeing.  In some cases you can fix this by describing the
data as "binary array=(dim1,dim2)" rather than as "binary matrix",
but  if I understand your earlier description that is not the case for
your data format.  

	Ethan

[toc] | [prev] | [standalone]

Back to top | Article view | comp.graphics.apps.gnuplot

csiph-web