Variable size binning

ROOT
1

i am dealing with the following problem:
i have to create an histogram with variable bin size(let’s call this binning A). i am reading abscissa x and ordinate y from an ascii file, but they have a different binning (let’s call this other binning B) from what i want to get at the end
my idea was to read the ascii file and put it into a TGraph and then put the TGraph into a TH1F that I create with the binning I desire (A). Is it possible,is there any method to do that? i didn’t find it anyway.
how should i solve my problem?

alternatively, i can create from the ascii file an TH1F with binning B,but then how to get a TH1F with the binning desired (A)?

regards and thanks for help,

2

Why don’t you simply use TGraph ?
You have X and Y coordinates, TGraph seems more natural.

3

i cannot use TGraph because i need to get each point at the position of a certain well determined bin (fixed by what i called binning A in the first post).

4

so I guess you should book a TH1 with the number of bins you need.
Then you fill it for each X(i) position with the weight Y(i)…
I hope I understood what you really want…

5

thanks for helping.
the problem is that the ascii file has values for y corresponding to values of x that are different from what i have to fix in my histogram. maybe it is not clear and should i try to explain better?

6

Yes,

Please explain better your problem.
What do you mean put a TGraph in a TH1. A TGraph is a set of (x,y) observation, a TH1 is an estimator for the density of a set of (x) observations. So, they are very different objects!

Lorenzo

7

hi,

concerning ROOT the problem is:

how can I pass from a TH1F with a binning (variable) to another TH1F with a different binning.

TGraph was just an idea to overcome the problem, but it is not the solution for me.

8

Hi,
This is possible only if the bin overlap, i.e. if the new binning is larger than the old one and contains exactly the same bins. See the function TH1::Rebin.
Otherwise, by binning you loose information on the original observations, so you cannot move to a different binning

Lorenzo