[SciPy-User] works in gnuplot, but not quite in Python/Gnuplot.py - splot (sin(x)/x) * (sin(y)/y)

[Esmail]

Hello all,

I am trying to generate a plot with Python/Gnuplot.py that I can
create interactively with gnuplot.

----------

   gnuplot> set isosample 40
   gnuplot> set hidden3d
   gnuplot> splot (sin(x)/x) * (sin(y)/y)

----------


gives me a *beautiful* continuous plot.

I am trying to do the same with Python/Gnuplot.py - but no joy.

I have the following short Python code to generate my data which I
then funnel into a file called 'somb.dat'

----------

#!/usr/bin/env python

# somb.py
# compute the sombrero function

from math import sin
from numpy import arange

def f(x, y):
     return (sin(x)/x) * (sin(y)/y)


for x in arange(-10, 11):
     for y in arange(-10, 11):
         print '%3d  %3d   %f' % (x, y, f(x, y))
     print

----------

e.g.,

    somb.py > somb.dat

and I use this to plot the data:

----------
#!/usr/bin/env python

import Gnuplot
g = Gnuplot.Gnuplot(debug=1)

g('set hidden3d')
g('set isosample 40')
g('set datafile missing "nan"')
g.splot(Gnuplot.File('somb.dat', with_='lines linewidth 1'))
raw_input()  # to pause

----------

Is there a way to achieve the same sort of continuous output that
gnuplot generates? I tried using 0.1 as 3rd parameter for arange, but
it didn't help. I suspect the problem is with the nan values (and/or
the numeric representation in Python)?. How does gnuplot handle
this? How can I achieve the same with Python/Gnuplot.py?

Any suggestions/hints?

(I am thrilled that I can use gnuplot via Python)

Thanks,
Esmail