RE: USRBIN-1D projections

From: Vasilis Vlachoudis <>
Date: Wed, 7 Mar 2012 14:05:25 +0000

"" <>

Hi Giuseppe,

On the "Detector Info" you can select the appropriate value to be plotted in X for each detector:
The following options are available:
- GeoMean = sqrt(low*high)
- Mean = (low+high)/2
- Low
- High
- Col.# (specific columns in the file)
Now if you plot the values as error bars or point then you should choose either the Geometric Mean or the Linear Mean depending if
the X value is scored in logarithmic bins or linear bins
If you want to plot it as a "histogram", since gnuplot doesn't really have histograms the ones that are resembling are the
steps, fsteps and histeps
The difference is in the x-starting value of each bin, briefly if you use the mean or geomean you should use the fsteps, steps
should only be used with the low value, and histeps only with the high value.

Vasilis Vlachoudis
CH-1211 GENEVA 23

Phone: +41-22 767 9851
GSM: +41-76 487 4378
Fax: +41-22 766 9644

From: [] on behalf of Giuseppe Battistoni []
Sent: 07 March 2012 11:35
To: Vasilis Vlachoudis;
Cc: Giuseppe Battistoni; Roberto Nicolini;
Subject: USRBIN-1D projections

Dear Vasilis
a question concerning FLAIR:
when we ask for 1-D USRBIN projection plots (let's take the example of a a logitudinal dose profile distribution),
we have noticed that the resulting ascii file containing the tabulations (in this example z, dose, error)
gives for z the lower value of each bin interval, when the dose value is instead the average on the
bin size.
Do you confirm that?
That is important when we interface those files to other analysis/graphic software
At present we are interpreting the file so that each dose value is assigned to the center of each spatial bin.

Thank you
Received on Wed Mar 07 2012 - 15:47:25 CET

This archive was generated by hypermail 2.2.0 : Wed Mar 07 2012 - 15:47:25 CET