Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Update README.md #944

Merged
merged 1 commit into from
Apr 17, 2024
Merged

Update README.md #944

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 6 additions & 6 deletions data/tutorials/groups/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,17 +12,17 @@ The files in this directory are:

For simplicity, this tutorial does not discuss scans using `-M MultiDimFit --algo=grid`,
but rather uses `-M MaxLikelihoodFit --robustFit=1`
Copy link
Contributor

@nucleosynthesis nucleosynthesis Apr 17, 2024 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I was too slow to add a comment but since we are updating this, we probably want to modify MaxLikelihoodFit -> FitDiagnostics since the former is no longer the method name

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@adavidzh @anigamova (fth), can we update again with this fix?

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

there is a Makefile in this folder that needs an update too, e.g. the fit options seem to be outdated https://github.com/cms-analysis/HiggsAnalysis-CombinedLimit/blob/main/data/tutorials/groups/Makefile#L113
I think this tutorial needs some refactoring along the lines https://cms-analysis.github.io/HiggsAnalysis-CombinedLimit/latest/part5/longexercise/#e-signal-strength-measurement-and-uncertainty-breakdown

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Or maybe we could just add some grouping instructions there? or do you prefer to keep the Makefile @adavidzh ?

to determine the crossings of the likelihood with the 1 sigma levels.
to determine the crossings of the likelihood with the 1σ levels.

## Groups of nuisances

Groups of nuisances are defined in the datacard, e.g. [here](./myanalysis.dc.txt#L30), [here](./myanalysis.dc.txt#L38-L40), [here](./myanalysis.dc.txt#L46-L47), and [here](./myanalysis.dc.txt#L56-L58).
They are parsed when building the binary datacard and result in two actions:
- creation in the workspace of a `RooArgSet` for each group, containing the concerned nuisance parameters.
- creation in the workspace of a `RooArgSet` for each group, containing the concerned nuisance parameters, and
- setting of a `group_%s` attribute in each nuisance parameter.

A nuisance parameter can belong to multiple groups of nuisances.
It is up to you to list or not list nuisances as part of groups in an exclusive, or overlapping way.
It is up to you to list or not list nuisances as part of groups in an exclusive or overlapping way.
In particular, there is no check that every nuisance is in at least one group.

At present, the syntax does not allow to declare groups as members of other groups;
Expand All @@ -44,12 +44,12 @@ by fitting the data would not be taken into account and even the central value w

One way to deconstruct the total uncertainty into different components is by freezing nuisances.
For instance, freezing all nuisances directly gives the statistical component as mentioned above.
By subtracting in quadrature the statistical uncertainty from the total uncertainty, one can derive the
systematic uncertainty.
By subtracting in quadrature the statistical uncertainty from the total uncertainty, one can estimate the
total systematic uncertainty.

When there are multiple components being disentangled,
the decomposition can proceed in many ways,
and an example of extracting the exp. syst. and theo. syst. is given in this tutorial, [here](./Makefile#L42-L70)
and an example of extracting the `exp. syst.` and `theo. syst.` is given in this tutorial, [here](./Makefile#L42-L70).

## Show me the money

Expand Down
Loading