TTree::Draw supports only one dimension. In case you have a two-dim
array, you should compute the 1d index number corresponding to your
2d array. For example, assuming an array a[10][20]
specify a[0] for a[0][0]
specify a[10] for a[1][0]
In case of 2-d arrays or more complex constructs, I encourage you
to automatically generate the analysis code corresponding to
your ntuple definition via TTree::MakeCode. if you have
an ntuple/tree h1, simply doing h1->MakeCode() will generate
a file h1.C with the ntuple skeleton analysis function.
In this code, you can use the normal two-dim array indexing
mechanism.
Rene Brun
> My problem is the following:
>
> I have a tree generated with h2root from a PAW CWN tuple.
> Here's the output of the PAW n/print command:
>
> PAW> n/print 3011
>
> ******************************************************************
> * Ntuple ID = 3011 Entries = 10 MDC11
> ******************************************************************
> * Var numb * Type * Packing * Range * Block * Name *
> ******************************************************************
> * 1 * U*4 * * * IEVT * IDEVT
> * 1 * I*4 * * [0,100] * S1D1 * NHIT
> * 2 * U*4 * 13 * [0,8000] * S1D1 * MDCTRK(NHIT)
> * 3 * R*4 * * * S1D1 * MDCX(6,NHIT)
> * 4 * R*4 * * * S1D1 * MDCY(6,NHIT)
> * 5 * R*4 * * * S1D1 * MDCTHETA(6,NHIT)
> * 6 * R*4 * * * S1D1 * MDCPHI(6,NHIT)
> * 7 * R*4 * * * S1D1 * MDCTOF(6,NHIT)
> ******************************************************************
> * Block * Entries * Unpacked * Packed * Packing Factor *
> ******************************************************************
> * IEVT * 10 * 4 * 4 * 1.000 *
> * S1D1 * 10 * 12404 * Var. * Variable *
> * Total * --- * 12408 * Var. * Variable *
> ******************************************************************
> * Blocks = 2 Variables = 8 Max. Columns = 3102 *
> ******************************************************************
>
> and the corresponding h3011->Print() output is:
>
> root [10] h3011->Print()
> ******************************************************************************
> *Tree :h3011 : MDC11 *
> *Entries : 10 : Total Size = 25608 bytes File Size = 25608 *
> * : : Tree compression factor = 1.00 *
> ******************************************************************************
> *Branch :Idevt : Idevt/i *
> *Entries : 10 : Total Size = 0 bytes File Size = 0 *
> *Baskets : 0 : Basket Size = 8000 bytes Compression= 1.00 *
> *............................................................................*
> *Branch :Nhit : Nhit/I *
> *Entries : 10 : Total Size = 0 bytes File Size = 0 *
> *Baskets : 0 : Basket Size = 8000 bytes Compression= 1.00 *
> *............................................................................*
> *Branch :Mdctrk : Mdctrk[Nhit]/i *
> *Entries : 10 : Total Size = 0 bytes File Size = 0 *
> *Baskets : 0 : Basket Size = 8000 bytes Compression= 1.00 *
> *............................................................................*
> *Branch :Mdcx : Mdcx[Nhit][6]/F *
> *Entries : 10 : Total Size = 0 bytes File Size = 0 *
> *Baskets : 0 : Basket Size = 8000 bytes Compression= 1.00 *
> *............................................................................*
> *Branch :Mdcy : Mdcy[Nhit][6]/F *
> *Entries : 10 : Total Size = 0 bytes File Size = 0 *
> *Baskets : 0 : Basket Size = 8000 bytes Compression= 1.00 *
> *............................................................................*
> *Branch :Mdctheta : Mdctheta[Nhit][6]/F *
> *Entries : 10 : Total Size = 0 bytes File Size = 0 *
> *Baskets : 0 : Basket Size = 8000 bytes Compression= 1.00 *
> *............................................................................*
> *Branch :Mdcphi : Mdcphi[Nhit][6]/F *
> *Entries : 10 : Total Size = 0 bytes File Size = 0 *
> *Baskets : 0 : Basket Size = 8000 bytes Compression= 1.00 *
> *............................................................................*
> *Branch :Mdctof : Mdctof[Nhit][6]/F *
> *Entries : 10 : Total Size = 0 bytes File Size = 0 *
> *Baskets : 0 : Basket Size = 8000 bytes Compression= 1.00 *
> *............................................................................*
>
> Now, in PAW I could do nice things on the 2dim arrays, like
>
> PAW> n/plo 3011.mdcx(1,2)
>
> and
>
> PAW> n/plo 3011.mdcx(1,1:nhit)
>
> or even select subranges
>
> PAW> n/plo 3011.mdcx(1:3,nhit-10:nhit)
>
> i.e. I had full control over BOTH indices!
>
> How much of this can I do with the TTree::Draw() member function?
>
> Thanks a lot for any help.
>
> Romain Holzmann
>
> p.s. Tom Moog from ANL posted a similar question quite some time ago, but no
> answer has apparently been threaded.