gnuplot-info

[Gnuplot-info] Using histep in log mode

[John A. <Joh...@ma...>]

Dear gnuplot,

I am writing to suggest that for the histep verticals, gnuplot should =20=

use the halfway point calculated *before* taking logs.

I attach a plot of equally spaced data in 100 bins of 10 MeV.  The =20
first bin with content is the bin 10-20 MeV.  The lower vertical =20
should be halfway between 5 and 15 MeV, namely at 10 MeV, but gnuplot =20=

places it at the point halfway between ln(5) and ln(15), namely exp(ln=20=

(5) + (ln(15)-ln(5))/2) =3D 8.66.  Misleading.

As well as the plot I attach data and a file of gnuplot commands.

Regards,

John Allison


John Allison                   Tel: +44-1422-845624
Dept. of Physics & Astronomy   Mobile/Cell: +44-7810-242262
The Univ. of Manchester        Email: Joh...@ma...
MANCHESTER M13 9PL             Email(2): Joh...@3-...
UK                             WWW: http://www.hep.man.ac.uk/u/johna


=EF=BF=BC=EF=BF=BC=EF=BF=BC=EF=BF=BC=EF=BF=BC=EF=BF=BC

Re: [Gnuplot-info] Using histep in log mode

[Thomas S. <t.s...@fz...>]

you don't need 'histeps':

plot "integral.hist" us ($1-5.):2 title "integral input" with steps, \
     "differential.hist" us ($1-5.):2 title "differential input" with steps,
\
     "integralTrigger.hist" using ($1-5.):($2*10) title "integral trigger x
100" with steps, \
     "differentialTrigger.hist" using ($1-5.):($2/4.41) title "differential
trigger x 100" with steps

-- 
View this message in context: http://www.nabble.com/Using-histep-in-log-mode-tf4858579.html#a13914899
Sent from the Gnuplot - User mailing list archive at Nabble.com.

Re: [Gnuplot-info] Using histep in log mode

[<HBB...@t-...>]

John Allison wrote:

> I am writing to suggest that for the histep verticals, gnuplot should 
> use the halfway point calculated *before* taking logs.

I have to disagree.  histeps is simply the wrong plot style for this 
input.  'steps' or 'fsteps' would be a good deal better.

The horizontal edge for the bin [20:30] should go neither from 15 to 25, 
nor from [sqrt(200):sqrt(600)], but from 20 to 30.  Expressing this by 
an x position of 25 and then letting gnuplot reconstruct the original 
boundaries is quite circuitous.