plot hbar*g*1000*(1+4*x**2)/(8*abs(x))/kb, "data" u ($9/80):1:5 w yerr \
pt 3 , "data" u ($9/80):2:6 w yerr pt 3
I have defined the constants there beforehand.
This produces errorbars around the data points with dashed lines
instead of a solid line. This can be fixed by not defining the pt
style but then it is difficult to tell the end of the error bars from
the data points since the two data are pretty close to each other.
So my question is, how do I make gnuplot plot the errorbars with the
data points as stars and the lines in the errorbars as solids and with
different colors.
Thanks.
Oli
A little bit more information.
These dashed lines do not show when I print the file normally (in a
wxt terminal with 0 options) but when I use
set term postscript eps enhanced color
and print out the graph to an eps file then I see these dashed lines
in the error bars.
Thanks.
Oli
I do not recognize this description, but...
>So my question is, how do I make gnuplot plot the errorbars with the
>data points as stars and the lines in the errorbars as solids and with
>different colors.
Maybe you could use the "candlesticks" plot type.
See
http://gnuplot.sourceforge.net/demo/candlesticks.html
--
Ethan A Merritt
That would be because your terminal setting has patterned linetypes (no
"solid" option to the postscript driver), and the errorbars have
linetype 2 and 3, which happen to be dashed.
> So my question is, how do I make gnuplot plot the errorbars with the
> data points as stars and the lines in the errorbars as solids and with
> different colors.
By actually specifying the colour, i.e. the linetype.
Ok, I tried all different linetypes and most of them give me a dashed
errorbar line (only lt -1 and lt 1 dont give me a solid line).
That is not good enough for me, I would like to be able to select from
all the colors available and get the errobarlines as solid.
Any more ideas?
Cheers
Oli
Ahhh, solved it by adding the solid option to the term setting:
set term postscript eps enhanced color solid
by default it uses the dash setting, silly gnuplot :-)
Cheers.
Oli