I’m trying to construct an N-dimensional histogram for use with a RooSimultaneous fit. It seems I have to add a weight of 1/N during the import step to get the correct bin entries with respect to the original TH1F. Is this correct? I have attached a minimal example script that reproduces the issue.
I’m using the example given in the v2.91 doc, pg 99,
This seems to cause problems with the errors reported in the simultaneous fit parameter dump/log file, but I’ll ask that question separately.
Thanks in advance,
PS: My root version: ROOT 5.31/01 (trunk@41357, Oct 12 2011, 16:00:36 on macosx64)
test.C (1003 Bytes)