[RF] New tutorial about bias and numeric issues in binned fits #12056
Conversation
|
Starting build on |
|
Build failed on ROOT-performance-centos8-multicore/cxx17. Errors:
|
lmoneta
left a comment
lmoneta
left a comment
There was a problem hiding this comment.
Thank you Jonas for the nice tutorial.
Maybe the tutorial could be split in two, one sowing the bias problem and the other showing the numerical stability.
| /// \date January 2023 | ||
| /// \author Jonas Rembser | ||
|
|
||
| // Generate binned Asimuv dataset for a continuouts pdf. |
There was a problem hiding this comment.
Here also you are creating a data set with expected events. In generated binned the values for each bin will be generated according to a Poisson process. In principle you could add a gRandom->Poisson(nEvents * integ->getVal()) to have the correct result of generatedBinned
There was a problem hiding this comment.
It is the intention that the dataset has the expected number of events. The generateBinned() function would do the same if you don't specify the Extended() command argument.
A new `rf614_binned_fit_problems` tutorial is added in both C++ and Python.
guitargeek
force-pushed
the
roofit_tuts_1
branch
from
cfb2b84 to
c801843
Compare
|
Starting build on |
|
Build failed on ROOT-ubuntu2004/python3. Failing tests:
And 2 more |
|
Build failed on ROOT-performance-centos8-multicore/cxx17. Errors:
|
|
Build failed on mac12/noimt. |
|
Build failed on windows10/cxx14. Failing tests: |
|
Build failed on mac11/cxx14. Errors:
|
|
Build failed on ROOT-debian10-i386/soversion. Errors:
|
3 participants
A new tutorial is added that explains how to deal with the bias and numerical instabilities in binned fits.