MadGraph5_aMC@NLO

How are errors determined in HwU files?

Asked by matteo maltoni

Dear MadGraph experts,

I'm trying to compare different ways to compute errors for some K-factor values, which I obtained from fixed-order generation.

In order to be more confident with this, I would like to know how are the errors assigned to the bin contents in the HwU file, since it seems to me that the "dy" values aren't related to the maximum and minimum results in the 9-point envelope, for each bin.

Can you please give me a short explanation, or point me to some reference?

Best,

Matteo

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
Rikkert Frederix Edit question
Solved by:
Rikkert Frederix
Solved:
Last query:
Last reply:
Revision history for this message
Rikkert Frederix (frederix) said : 		#1

Dear Matteo,

The 'dy' values are the statistical uncertainties.
"delta_mu_cen -1 @aux ", "delta_mu_min -1 @aux" and "delta_mu_max -1 @aux" are the central, minimum and maximum values of the scale variation related to scale choice "-1" (which is the default in the run_card). The latter two can be computed from the 9 next values in the HwU files by considering the full envelope of these 9 values.

Best,
Rikkert

Revision history for this message
matteo maltoni (matteo-maltoni) said : 		#2

Dear Rikkert,

Thank you for your answer, now it's clearer to me.

Statistical uncertainties and scale variation ones aren't related, right?
Is there any reason for which one should be smaller than the other?

Best,

Matteo

Revision history for this message
Best Rikkert Frederix (frederix) said : 		#3

Dear Matteo,

They are unrelated and hence one can be smaller than the other. In fact, statistical uncertainties can simply be reduced by generating more events, while scale variation can only be reduced by going from leading-order to next-to-leading order (or even higher-order) accuracy in the expansion in the strong coupling constant.

Best,
Rikkert

Revision history for this message
matteo maltoni (matteo-maltoni) said : 		#4

Thanks Rikkert Frederix, that solved my question.

Revision history for this message
Doris Mulcahy (gmontson13) said : 		#5

How do we measure error? Two common measures of error include the standard error and the relative standard error. Standard Error (SE) is a measure of the variation between any estimated population value that is based on a sample rather than true value for the population. https://inloggengids.com/

There are three types of errors that are classified based on the source they arise from; They are: Gross Errors. Random Errors. Systematic Errors.
Systematic Error (determinate error) The error is reproducible and can be discovered and corrected. Random Error (indeterminate error) Caused by uncontrollable variables, which can not be defined/eliminated.