gnuplot development

Sehr geehrte Damen und Herren,

wie kann ich in gnuplot
auf dem x( oder y, egal ) Achse scrt() schreiben,
ich meine wie ein Text, nicht scrt() zu berechnen?

Herzlcihe Gr=FCsse,

stefka

Dear madams and sirs,
how can I write sqrt() on the x axis?
Not to calculate the value of sqrt() but to
write  it as a text,  the sign  of sqrt()  ?

All the best,
stefka

Sch=F6nes Wochenende,

have a nice weekend,

stefka

Stefcheto wrote:

> auf dem x( oder y, egal ) Achse scrt() schreiben,
> ich meine wie ein Text, nicht scrt() zu berechnen?
[...]
> how can I write sqrt() on the x axis?
> Not to calculate the value of sqrt() but to
> write  it as a text,  the sign  of sqrt()  ?

Your question is a bit hard to parse, but you seem to be asking how to 
get the thing that looks roughly like this (use a fixed-width font do 
see it)
    _______
   /
\/

into an axis label?  Well, that depends on the terminal driver you're 
using.  Some have this letter in their Symbol fonts (PostScript, e.g., 
does), some even allow you write a real math equation using properly 
formatted sqrt signs (like the TeX family of drivers, including pslatex 
and epslatex:

	set xlabel '$\sqrt{x^2+5}$'

> When typing 'help boxes' we get the following explanation (I only keep the
> beginning) : 
> 
> "boxes
> 
> The boxes style is only relevant to 2-d plotting.  It draws a box centered
> about the given x coordinate from the x axis (not the graph border) to the
> given y coordinate.  The width of the box is obtained in one of three
> ways. If it is a data plot and the data file has a third column, this will
> be used to set the width of the box......"
> 
> I am very interested in the boxwidth question, but don't understand
> exactly the end (above in bold). We can put a coefficient in a third
> column in order to reduce the width od boxes ? 
> 
> thank you again
>

Luczak Fr=E9d=E9ric wrote:

>>I am very interested in the boxwidth question, but don't understand
>>exactly the end (above in bold). We can put a coefficient in a third
>>column in order to reduce the width od boxes ?=20

Not particularly to reduce it, but to determine it.  Boxes can either be =

made to touch each other automatically by gnuplot, or have a width which =

you specify.  This width specified by you can be a constant number, or=20
it can be taken from the data file, such that it can be different for box=
=2E

In a minimalistic example:

	set yrange [0:*]
	plot '-' with boxes
	  10 1 2
	  20 2 5
	  30 3 1
	 e

> I hope that it is the last piece of advices I ask you to answer.
> 
> I intend to plot a series with boxes (filled say in black color) with a
> grey-light background colour (I mean only inside the x and y axes). I use
> postscript output.
> Since it is not possible directly I use the following way to do it : 
> 
> 
> 
> plot 'myfile1'  using 1:2 with boxes fs solid 0.1 noborder, \
> # that is plotting a fictitious series from min y1 to max y1 in light grey
> color
>       ' myfile2 '  using 1:2 with boxes title border etc...
> # that is the real series I want to plot
> 
> the pb is that I would like the first boxes to be adjacent (for an
> homogeneous grey) and the second ones with a space between boxes
> (something like boxwidth 0.8 relative)
> 
> How can I do to manage that ?

Luczak Fr=E9d=E9ric wrote:

>>the pb is that I would like the first boxes to be adjacent (for an
>>homogeneous grey) and the second ones with a space between boxes
>>(something like boxwidth 0.8 relative)
>>
>>How can I do to manage that ? =20

You can't.  Box width is a global setting --- its value can't be=20
selected differently for individual data sets.

hi
i am learning java programing
i dont know how to plot gnuplot
i have to do the following program
i have completed up to second point
now i have difficulty in plotting gnuplot from the data
so please send me information about this "GNUPLOT"
this is my question

In the theory of telecommunication networks, the probability that a call
offered to a group of *r* circuits finds that all *r* circuits are busy is
given by Erlang's Loss Formula:

            r     /   r    i
           a     /  ---   a
E(r,a) =  ---   /   \    ---
           r!  /    /     i!
              /     ---
                    i=0

where *a* is the traffic offered. However, the recursive formula:

  1            r     1
-----  =  1 + --- ---------             r = 1, 2, 3, ...
E(r,a)         a   E(r-1,a)

with *E(0,a) = 1* offers a more efficient method of computing loss
probabilities.

   - Write a Java program which takes the parameters *R* and *a* and
   calculates *E(r,a), r = 1, 2, ... R*. Use the recursive version of
   Erlang's loss formula. (45% of marks)
   - Extend your program (25% of marks) to take the parameters *R, A* and
   *L* be able to generate data for the functions *E(r,a), r = 1, 2, ...
   R* in the range 0 <= *a* <= *A* and 10-*L* <= *E* <= 1. The horizontal
   *a*-axis should have a linear scale. The vertical *E*-axis can be
   linear but extra credit (5% of marks) will be given to those who wish to
   make the vertical axis have a logarithmic scale.

   - Use your program to generate data for

     R = 20, A = 12, L = 3


   - Write out the data into a file, and plot the file data using a
   spreadsheet program or graphing package such as gnuplot.

thankin you
Gajjar hiren

http://www.nabble.com/file/p15975854/bond_length.jpg 

i want to let the curves cover the curves behind, but as the plot showed, it
didn't work.

i tried  to type      
 
set hidden3d
set hidden

but it failed.

if anyone know how to let it work, please reply me. :-)
-- 
View this message in context: http://www.nabble.com/gnuplot-tp15975854p15975854.html
Sent from the Gnuplot - User mailing list archive at Nabble.com.

Hello.
My name is Dmitry Jaworski, I am a student of physics at Warsaw
University. I am currently engaged in processing of experimental data
for my master's work.
The data-processing program I use is gnuplot. I would like to ask you
if it is possible, when fitting, to establish certain relationships
between parameters. For example: for the fitting, I use a function f
(x) = a + b * x and parameter 'a' must be > -3b (larger than -3b).
I'll be very grateful for the help.
Sincerely.
D J

On 19.03.2011 19:44, Dmitrij Jaworski wrote:

> The data-processing program I use is gnuplot.

That may not be the right tool for the job.  gnuplot is a plotting 
program, not a data processing engine.  It does offer some data 
processing as an aid for the plotting work, but it's not supposed to 
replace an actual data management tool.

> I would like to ask you if it is possible, when fitting, to establish
> certain relationships between parameters.

No.  But you can re-define your parameters such that such relations are 
automatically taken care of.

> For example: for the
> fitting, I use a function f (x) = a + b * x and parameter 'a' must
> be>  -3b (larger than -3b).

So fit

	f(x)= (a_internal**2-3b) + b * x

then extract the actual a from sqrt(a_internal)
I'll be very grateful for the help.
> Sincerely.