Unexpected behavior of TProfile::Multiply(TF1*, double)

ROOT
1

Hi,

I’ve got a profile histogram and I’m multiplying by a function 1/x using TProfile::Multiply(TF1*, double). I was expecting the vertical position of each bin to change by a factor of 1/x. Instead, it seems as if I may be multiplying the weights (errors). Is this the expected behavior? If so, is there an option to modify the vertical bin position?

I’ve included a test macro and associated datafile. I use root 4.02.00

mike kordosky
root_profile.root (18.4 KB)
root_profile.C (372 Bytes)

2

Mike,

I have fixed the scaling in this function (quite tricky).
Thanks for reporting this case.
Now in CVS.

Rene