Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

gnuplot rendering engine bug?

64 views
Skip to first unread message

usen...@gmail.com

unread,
Apr 28, 2013, 10:19:43 AM4/28/13
to
Hi

I'm having a play around with the 3d cylindrical coordinate system, and have come up against a posible bug. The front half of my graph isn't rendering.

my script is:
--------------
reset
set title "3D version using cylindrical coordinate system"
set ticslevel 0.0
set view 30,56,0.98
set angles degrees
set mapping cylindrical
set parametric
set zrange[-60:0]
unset key
splot 'freqResp.dat' u 1:3:2 with pm3d
--------------

And I've put the "freqResp.dat" file here: http://www.d-thompson.co.uk/uploads/3/4/5/9/3459097/freqresp.dat

When I run the script, I only get the back half of the graph being rendered: http://i.stack.imgur.com/js4hz.png

If I run a different script in normal cartesian coordinates, then all the data is present: http://i.stack.imgur.com/XgS5O.png

Normal script:
-------------
set xlabel "Arrival Angle (degrees)" font "arial,8"
set ylabel "Frequency (Hz)" font "arial,8"
set zlabel "Gain (dB)" font "arial,8"

set grid lc rgbcolor "#BBBBBB"

set xrange[-180:180]
set yrange[0:20000]
set zrange[-60:0]

unset key
set view 30,56,0.98

splot 'freq.dat' u 1:2:3 with pm3d
-------------

So can anybody confirm if this is a bug? (or if it's a problem with my code?)

Thanks
Dave

sfeam

unread,
Apr 28, 2013, 5:46:33 PM4/28/13
to
usen...@gmail.com wrote:

> Hi
>
> I'm having a play around with the 3d cylindrical coordinate system,
> and have come up against a posible bug. The front half of my graph
> isn't rendering.

You need to add
set pm3d depthorder

In short, unless you specify that the plot elements are to be sorted
on Z, they are drawn in the order they appear in the input file.
Your "front half" plot elements are near the top of the file so they
get overwritten by elements that appear later in the file.

Ethan

usen...@gmail.com

unread,
Apr 29, 2013, 6:34:04 AM4/29/13
to
That's excellent - thanks very much
Dave
0 new messages