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

Groups > comp.graphics.apps.gnuplot > #936

Re: plot in hexagonal lattice

Show all headers | View raw

On Feb 10, 11:42 am, Péter Juhász <peter.juhas...@gmail.com> wrote:
> On Feb 9, 10:52 am, zsolti elter <elter.zs...@gmail.com> wrote:
>
>
>
>
>
>
>
>
>
> > Hello!
>
> > I would like to plot data-s in a hexagonal lattice. I have a .dat file
> > in which the 1st and 2nd column are the X and Y coordinates of the
> > hexagon's centers, and the other columns hold the data, what I want to
> > plot in the lattice.
>
> > I managed only to plot the datas with pm3d, but I don't know, how to
> > plot the lattice grids, and how to plot hexagons. I plotted the datas
> > by circles, and I increased the size of it, but it's not the best.
> > So I would like to get something like the lattice map on the left part
> > of the picture:http://www.kfki.hu/~aekihp/arlhome/images/verona1.gif
>
> > For the lattice I guess I should use a loop and set object polygon,
> > because from the center of a hexagon ! can evaluate the points of the
> > hexagon (and I know the lattice pitch), but I don't know how to read
> > in every step only one X and Y value from my data file, and how to
> > assign the value of the data (power in my case) to this hexagon.
>
> > Thanks for your help!
>
> > Best regards,
>
> > Zsolti Elter
>
> I see two ways to do this.
>
> The first is similar to what you propose: plot the file twice,
> building the command that declares the hexagons in the first pass, and
> plotting them in the second.
>
> ########################################
> reset
> fn = 'hexagons.dat'
> S = 0.5
> C = sqrt(3)/2
> A = 1/C
> a = 0.95*A
>
> hex_id = 1
>
> cmd = ''
>
> hex_cmd(x, y, c) = sprintf("set obj hex_id poly from %g,%g-a to %g+a*C,
> %g-a*S to %g+a*C,%g+a*S to %g,%g+a to %g-a*C,%g+a*S to %g-a*C,%g-a*S
> to %g,%g-a fs solid 0.8 border lc pal cb %g lw 2 fc pal cb %g; hex_id
> = hex_id + 1;", x,y,x,y,x,y,x,y,x,y,x,y,x,y,c,c)
>
> read_f(x, y, c) = (cmd=cmd.hex_cmd(x, y, c), y)
>
> set term unknown
>
> plot fn u 1:(read_f($1,$2,$3))
>
> set term pop
>
> eval cmd
>
> set size ratio -1
> set xr [GPVAL_X_MIN-A:GPVAL_X_MAX+A]
> set yr [GPVAL_Y_MIN-A:GPVAL_Y_MAX+A]
>
> plot fn u 1:2:3 w d lc pal notit
> ###################################
>
> Some test data:
>
> ###################################
> 3       1.73205080756888        0.627868065218539
> 2       3.46410161513775        0.0381587751648738
> 3       5.19615242270663        0.372697677268786
> 2       6.92820323027551        0.421502736487078
> 3       8.66025403784439        0.274237128801435
> 5       1.73205080756888        0.202129157245494
> 4       3.46410161513775        0.769320237819866
> 5       5.19615242270663        0.031500208931071
> 4       6.92820323027551        0.562379837027411
> 5       8.66025403784439        0.504284749332101
> 7       1.73205080756888        0.217345187922358
Thank you very much!

best regards,
Zsolti


> ###################################
>
> This method should work with most terminals with a relatively recent
> gnuplot.
>
> The second method is specific to the postscript terminal. You can
> modify the locally installed copy of the postscript prologue gnuplot
> prepends to every ps file it produces, changing one of the point types
> to a hexagon. I don't have time to do this, but it should be
> straightforward if you have some knowledge of the postscript language.
> Once this is done you can just plot the data 'with points pointtype X'
> where X is the point type you modified.
>
> Péter Juhász

Back to comp.graphics.apps.gnuplot | Previous | Next — Previous in thread | Next in thread | Find similar

Thread

plot in hexagonal lattice zsolti elter <elter.zsolt@gmail.com> - 2012-02-09 01:52 -0800
  Re: plot in hexagonal lattice Péter Juhász <peter.juhasz83@gmail.com> - 2012-02-10 02:42 -0800
    Re: plot in hexagonal lattice zsolti elter <elter.zsolt@gmail.com> - 2012-02-10 23:16 -0800
  Re: plot in hexagonal lattice Péter Juhász <peter.juhasz83@gmail.com> - 2012-02-10 02:53 -0800

csiph-web