full LEFT beta function added

[jasonaebischerGIT]

Here the beta function in the LEFT, including all terms.

[jasonaebischerGIT]

Goes together with PR #93

[peterstangl]

I have made some small changes such that the beta functions can be used with the current WCxf basis definition. In particular, I have

• commented out the beta functions of all undefined operators
• set all undefined Wilson coefficients to zero
• defined the Kronecker delta inside the function beta

Since the new file added in this PR does not interfere with anything else in wilson, but is necessary both for PR #93 and for my plan to automatically extract the ADMs from the beta functions, I think this PR can be merged already.

[jackypheno]

Yes maybe its good idea to merge it!

[jackypheno]

@jasonaebischerGIT @peterstangl I think still the shape of the Kronecker delta needs to be adjusted for up-type (2 flavours) and down-type (3-flavours).

[jackypheno]

@jasonaebischerGIT @peterstangl I think still the shape of the Kronecker delta needs to be adjusted for up-type (2 flavours) and down-type (3-flavours).

OK, I probably this can be handled by using pad_C method.

[peterstangl]

@jackypheno yes exactly, you can use the pad_C method. The beta function can be used as follows:

import numpy as np
from wilson.test_wilson import get_random_wc
from wilson.util import wetutil
from wilson.run.wet.beta import beta

# generate some wcxf input
wc = get_random_wc('WET', 'JMS', 90)

# turn WCxf Wilson coefficients into symmetrized arrays (the redundant basis)
C = wetutil.wcxf2arrays_symmetrized(wc.dict)

# add the SM parameters (here I just set them all to 1)
C = {**C,**{k: 1 if v==1 else np.ones(v) for k,v in wetutil._dim4_keys_shape.items()}}

# pad with zeros to bring all arrays to full 3-generation shape (needed for u-type quarks)
C = wetutil.pad_C(C)

# evaluate the beta functions
result = beta(C)

# unpad the result to remove zeros and bring all arrays to their canonical shape
result = wetutil.unpad_C(result)

[jackypheno]

@jackypheno yes exactly, you can use the pad_C method. The beta function can be used as follows:

import numpy as np
from wilson.test_wilson import get_random_wc
from wilson.util import wetutil
from wilson.run.wet.beta import beta

# generate some wcxf input
wc = get_random_wc('WET', 'JMS', 90)

# turn WCxf Wilson coefficients into symmetrized arrays (the redundant basis)
C = wetutil.wcxf2arrays_symmetrized(wc.dict)

# add the SM parameters (here I just set them all to 1)
C = {**C,**{k: 1 if v==1 else np.ones(v) for k,v in wetutil._dim4_keys_shape.items()}}

# pad with zeros to bring all arrays to full 3-generation shape (needed for u-type quarks)
C = wetutil.pad_C(C)

# evaluate the beta functions
result = beta(C)

# unpad the result to remove zeros and bring all arrays to their canonical shape
result = wetutil.unpad_C(result)

Yes! It has been implemented already in #93 !

[dvandyk]

@peterstangl You approved this PR but never merged it. Is there a particular reason?

[peterstangl]

@peterstangl You approved this PR but never merged it. Is there a particular reason?

Sorry @dvandyk I somehow overlooked your message. This PR was originally made to assist fixing issue #54 and was also needed for PR #93. Issue #54 has been fixed without needing to merge this PR and #93 did not have any activity for more than two years. The code in this PR is not used anywhere in wilson, so I think there is not really a point in adding it. It might be useful to have it in the future, so I think we could keep the PR open, but I now marked it as a draft.

[jackypheno]

This PR is superseded by #114 and #115 since new EFTevolve class also works for LEFT. So I am closing it.

[peterstangl]

This PR is superseded by #114 and #115 since new EFTevolve class also works for LEFT. So I am closing it.

Hi @jackypheno, I don't really understand. This PR you closed contains only the LEFT beta functions. Please see my last comment in the conversation above. I don't see that #114 and #115 contain the LEFT beta functions, so how do they supersede this PR? Maybe you meant to close the related PR #93?