[ROOT] Graph Versus TBrowser's Histogram

From: Tommaso Chiarusi (Tommaso.Chiarusi@bo.infn.it)
Date: Wed Jan 16 2002 - 12:20:25 MET


Dear Rooters,
I've tried to plot two variables using the features of 
TBrowser's StartViewer.
I dragged the a and b variables (leaves in my data set) onto the x and y
places and then I pressed the
graph button to issue a plot on a canvas.
The result is in the picture 

/afs/cern.ch/user/c/chiarusi/public/aVSb_browser.ps


The variables a and b are respectively the major and minor semiaxis of
some ellipses, so  b MUST be <= a.
Well  I startled to the plot since
I saw many point getting below the y=x line (i.e. b>a!!!!!).

I tried to us<e the scan feature to see which of those were such events
with b>a.  
Root told me that NONE of them was "b>a compliant"!

So I wrote a small script verifying inipendently from Root, and I got the
same result: none of the events was with b>a!

Then I plotted the same events using the TGraph, and I got the
plot 

/afs/cern.ch/user/c/chiarusi/public/aVSb_graph.ps 

where is perfectly clear that the geometric condition b<=a is completely
respected!

So, my  question is:
Is it possible that the StartViewer way of plotting is made via
histogramming and not via "TGraph-ing" ?

What do you think about swapping this features ? So Scan's results
could be confirmed directly by  StartViewer's plots. 

Thanks!
Tommaso 


P.S.
I'm using old 3.0 version of root.
Do the GetHistogram() or GetXaxis() functions work with a TGraph 
(let's say in case if I would like to zoom throug the axis ranges )?

Thank you very much again.
Tommaso



This archive was generated by hypermail 2b29 : Sat Jan 04 2003 - 23:50:38 MET