Re: [ROOT] questions on histogram errors

From: Stephane Plaszczynski (plaszczy@lal.in2p3.fr)
Date: Thu Sep 18 2003 - 15:14:55 MEST


OK both work well (even if I don't understnd why the Update() should be 
moved...)

thanks for your patience,

Stephane

--On Thursday, September 18, 2003 03:07:11 PM +0200 Rene Brun 
<Rene.Brun@cern.ch> wrote:

>
>> --On Thursday, September 18, 2003 12:02:44 PM +0200 Rene Brun
>> <Rene.Brun@cern.ch> wrote:
>>
>> > I don't know what you mean by sigma_y?
>> > If you just want to draw a set of points with eerrors, use TGraphErrors
>> > TGraphErrors(n,x,y,ex,ey)
>>
>> OK. So I guess my question is whether there is an "automated" way of
>> creating this Graph directly from the ntuple entries (I am used to a tool
>> that can do it): something like Draw("y:x") but where each point
>> displayed would have an error bar given by two other ntuple entries
>> "sigmax,sigmay"?
>>
>
> You can do the following
>  nt->Draw("x:y:sigmax","sigmay","goff");
>  TGraphErrors g(nt->GetSelectedRows(),nt->GetV1(),nt->GetV2(),nt->V3(),
>                 nt->GetW());
>  g.Draw("ap");
>
> see TTree::Draw doc section "How to obtain more info from TTree::Draw"
>
>> >
>> II tried the following unamed script (following your suggestion). It is
>> in "myscript.C"
>> I log in interactively to root and type
>> .x myscript.C
>> but the ntuple points do not appear....
>>
>
> You should move the c1->Update() statement immediatly after having drawn
> the histogram. See change below



This archive was generated by hypermail 2b29 : Thu Jan 01 2004 - 17:50:15 MET