"h->Fill();" problem

ROOT
1

Dear people,

I have one very strange problem, that looks totally crazy.

I do the following in interpreter:

h=new TH1F(“h”,“h”,100,0,0.1);
h->Fill(0.016);
h->Fill(0.017);
h->Draw();

And what I see? 0.016 comes to the bin 16 (0.015-0.016), and 0.017 comes to the bin 18 (0.017-0.018).

How can it be and what is wrong?

I use 528-00 version.
I tried it on 527 and 523 as well. Result is the same.
On 529 there is no such problem.

Thank you in advance!

2

[quote=“Dasha”]Dear people,

I have one very strange problem, that looks totally crazy.

I do the following in interpreter:

h=new TH1F(“h”,“h”,100,0,0.1);
h->Fill(0.016);
h->Fill(0.017);
h->Draw();

And what I see? 0.016 comes to the bin 16 (0.015-0.016), and 0.017 comes to the bin 18 (0.017-0.018).

How can it be and what is wrong?
[/quote]

My guess is that you are are just being hit by the floating point precision.

root [5] 0.016
(const double)1.60000000000000003e-02
root [6] 0.017
(const double)1.70000000000000012e-02

If you really need to make sure that 0.016 and 0.017 fall in sequential bins, then I suggest you rebin your histogram so that these values fall in the middle of the bins instead of the edges.

h=new TH1F("h","h",100,0.005,0.105);

Good luck,
Charles