different options for smeft_accuracy result in very different results after running

[Yong-Du]

Hi,

I was trying to use the wilson package for running some Wilson coefficients from the mZ scale up to 1TeV, but I got very different results (differ by ~1e12) when choosing different options for smeft_accuracy. I tried other operators (like phiD) as well, but got similar discrepancy. Since I do not expect the RGEs for the Wilson coefficients or the SM parameters to be stiff, I would agree with the methods defined in your rge.py file. However, I have a difficulty in understanding this huge discrepancy. Do you have any comments/suggestions? (My code and the output are attached below.)

Thanks,
Yong

Code:

import wilson as wc
from wilson import Wilson

w = Wilson({'lequ1_1111':1.},
           scale = 91.1876,
           eft = 'SMEFT', basis = 'Warsaw')
w.set_option('smeft_matching_order',1)
w.set_option('smeft_accuracy','leadinglog') 
wc = w.match_run(1000.,'SMEFT','Warsaw') 
print(wc['lequ1_1111'])
w.set_option('smeft_accuracy','integrate') 
wc = w.match_run(1000.,'SMEFT','Warsaw') 
print(wc['lequ1_1111'])

The output

(0.8122261045241321+3.455055738367379e-09j)
(2.6547291756603105e-12-3.7133250374864686e-14j)

[DavidMStraub]

Hi,

I haven't looked at your specific example, but in general this is something that is to be expected if the mixing of Wilson coefficient B into Wilson coefficient A is zero in leading-log approximation. Note also that the numerical value for the Wilson coefficient you are choosing is physically meaningless as 1/GeV² invalidates the EFT expansion.

[Yong-Du]

Hi,

I get it now, thank you for the comment.

Best,
Yong