Skip to content

Finding uncertainty on parameter of interest with non-parabolic likelihood #844

Closed Answered by HDembinski
JaLuka98 asked this question in Q&A
Discussion options

You must be logged in to vote

You have to call flatten in your model, because the data you want to fit is 1D. The model prediction must match the data. You can also fit multivariate data with a multivariate model with iminuit.

If you have binned data (a histogram), you cannot use UnbinnedNLL, that is for unbinned samples. That you need to set m.errordef = Minuit.LEAST_SQUARES is a consequence of the misuse of UnbinnedNLL, this would be wrong if you used UnbinnedNLL in its correct context. The correct likelihood for binned data is BinnedNLL, but you cannot use that either, because your histogram does not contain poisson-distributed counts but has bin-to-bin correlations which are captured by the cov matrix.

There is cu…

Replies: 3 comments 2 replies

Comment options

You must be logged in to vote
2 replies
@HDembinski
Comment options

@HDembinski
Comment options

Answer selected by JaLuka98
Comment options

You must be logged in to vote
0 replies
Comment options

You must be logged in to vote
0 replies
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Category
Q&A
Labels
None yet
2 participants