Description of how to produce k.dat files

The basic idea is described in Brandenburg (2001, ApJ 550, 824),
where however a sqrt was missing around the 2 in the denominator.

By default, we employ a domain of size (2pi)^3, so the spacing
of k in each direction is 1. The routine can then be adapted
"by hand" to suit the problem of interest.

In particular, you can change the numbers in the line:

  dkz=1. & ex=1. & ey=1. & ez=1. & kmax=10. & k1=4.5 & k2=5.5    ;(gives 350 vectors)

to

  dkx=1. & dky=1. & dkz=.5 & ex=1. & ey=1. & ez=1. & kmax=10. & k1=4.5 & k2=5.5    ;(gives 630 vectors)

if the domain has a z extent of 4*pi, so therefore dkz=.5.
You get then more vectors.

At the end of the routine, it prints the line:
  good=where(abs(kky) le .4) & plot,kkx(good),kkz(good),ps=1,/iso,xst=0,yst=0

You can cut+paste it into the command line to plot whether
the k-vectors look sufficiently well populated and isotropic.

If there are further questions, let me know.
Axe Brandenburg