forums.silverfrost.com

[sospel]

Hello !

I would need an advice

Here is my problem :

I calculate some comets trajectories and I would want to make them draw in 1 or 2 window which I open on the screen, with a length and a width given, a given thorough color, a given title, and every trajectory drawn point by point, of a given color.

My problem is this one: I have well the graphic software SIMFIT, SIMDEM , CLEARWIN+ and OPENGL in SILVERFROST package but " there is too much "
and I don't know which to use ....

Could one of you recommend me on the best way to begin ? Especially also because after the plan of comets, I would like to draw the "chaotic" trajectories which aim towards the strange attractors ...

Thank you in advance !!

Cordially,
Sospel

[DanRRight]

Post a link on example of what you'd ideally wanted to see?

[LitusSaxonicum]

Sospel,

If your points on the comet trajectory are close enough to each other, you can just draw straight lines between them and it will look like a curve onscreen. When they are too far apart, you need to interpolate more points. As a comet trajectory is a conic section, you should be able to interpolate positions using a parabola over a short enough segment (or a hyperbola if you are fussy), and calculate enough extra points to give a smooth curve appearance when you join the points with straight lines! You can fit a parabola through any 3 points. Do it all with Clearwin+ graphics. Comets don't change direction suddenly (at least while they don't hit planets full of dinosaurs), so the trajectory curves must be gradual. I've drawn lots of curves this way, usually circles or circular arcs, and they look fine.

Eddie

[sospel]

Hello @ DanRRight and LitusSaxonicum

Thank you for your answers.

In reality, I wished to have a "global" advice on the graphics softwares to choose the one who is the easiest to use for many graphs : trajectories of comets, but also trajectories of "mathematical objects " such as those converging on a strange attractor, or Julia's " curves " or Mandelbrot "fractals" ...

I think that LitusSaxonicum suggests with good reason to using CLEARWIN+ : I browsed the notice and I have effectively the feeling that I can open several windows simultaneously, to study for example the influence of certain parameters.
I have moreover just found in the directory of CLEARWIN an example of program which corresponds about to the fact that I wish to make : "PLOTFUNC.F95" . It is short and very easy to modify : I am going to be inspired by it to begin my graphs !!

Thus, thank you again, and see you soon for other advice

Cordially
Sospel

[LitusSaxonicum]

PLOTFUNC makes use of SimplePlot, and elsewhere in the forum you will learn that this is no longer supported. Even a few years ago I decided not to use it because of that, but (despite a few issues with modern versions of Windows) it appears to still work, and to have some really nice features.

I always draw my own stuff from the simple primitives in FTN95/Clearwin+. Here is a simple and rough program to show you how to draw 2 separate graphics areas in one Window (but you could have several free-floating windows as well):

[LitusSaxonicum]

[Kenneth_Smith]

I've played about with Clearwin+ and found it very useful, although a difficult to (initially) get my head round some of the programing concepts. The drawing space in the graphics region being in the 4th quadrant caused me a fair bit of grief in the beginning!

However, I'd recommend you give it a go, Clearwin+ has allowed me to do things where other plotting tools have failed.

The following is an example of what's possible with just a few lines of code:-

http://www.avao.co.uk/plot.bmp

Cheers

Ken

[LitusSaxonicum]

Having Y increase downwards also confused me to begin with. I use the following two statement functions to convert to pixels:

[JohnCampbell]

Ken,

An interesting plot.

You appear to have the same problem as I do of trying to get a rainbow palette where the displayed colours are visually differentiated, especially in the greens.
It would be good to get an algorithm that can give a rainbow of colours, with even perception of colour difference. Using a uniform numerical difference in the RGB numbers does not give a good solution.

John

[Kenneth_Smith]

John,

It is indeed a problem, especially with the green part of the spectrum - for a while i thought it was just my eyesight.

The basic algorithim i use is as follows:-

[Kenneth_Smith]

I very much agree with you - it would be good to find a better way to do this. One thought, perhaps having a second "vector" which runs around the colour wheel faster to generate a second set of RGB values, then multiply R(I) from the first vector by R(I) from the second etc. might work. Mathematically this would look something like an amplitude modulating function, with the possibiltiy of varying both the amplitude, phase and frequency of the modulation.

Off to the day job now - which is much less interesting!

Ken

[LitusSaxonicum]

Yes I agree, the rainbow spectrum in the diagram is excellent. The rainbow isn't the only workable 'spectrum', there are others used in geography for example, representing the highest mountains to the deepest oceans. You can also do a single colour (hue) and change its intensity, that gives you less obvious contours.
At a guess, I would say that someone somewhere is very knowledgeable about this!

Eddie

[Kenneth_Smith]

I did a little bit of googling at lunch time as came across this:-

http://paulbourke.net/texture_colour/colourspace/

About 3/4 down the page there is a dicussion "Simple method of storing colour ramps"

There are 22 different "colour ramps" displayed along with a data file for each, containing the 256 different RGB values in each ramp.

Ramp 1 looks very like the output of the ANGLE_TO_RGB subroutine above, which is problematic due to the wide range of the greens. Ramp 16 looks more promising. So I guess some form of interpolation on the corresponding table of data may be useful for some applications. Not exactly an algorithm that can give a rainbow of colours, with even perception of colour difference, that John is looking for, but it might be good enough for some applications?

Ken

[Kenneth_Smith]

I took a quick look at the data files for RAMP1 and RAMP16, to see how the individual RGB values vary with position around the colour wheel. RAMP1 is a trapazoidal fuction and RAMP16 is a triangular function - which consequently has the maximum G value for a shorter duration. What's needed is a sort of average between the two, and an offset cosine wave very approximately, fits the bill. Hence:-

[DanRRight]