Regarding Checking Gauge and Lorentz Invariance Test

Hi,

> 1.As this lines are coming, can I still believe the results ?

Yes, We support various method for the computation of the loop, (cut tools, irigi, pjfry and golem). The two first are installed by default with madgraph.
If you want to use the two other, you need to install them manually and then indicate to madgraph (in input/mg5_configuration.txt) where they are on your system.
Then in the mad loop_card, you can decide in which order each of those four method will be use for the computation.

The idea is that for each computation we check the stability of the result. If the result is unstable, then we pass to the next available method.
If all are unstable then we use cut tools in quadruple precision (which fix the problem in almost all cases).

So not having PJFRY/GOLEM is not problematic (I do not have those either) but typically allow your code to run faster.

Cheers,

Olivier

PS: I put Valentin in charge, if he wants to add more details on how to install those packages.

On 14 Oct 2015, at 10:52, shibasipu <email address hidden> wrote:

> New question #272412 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Dear Olivier,
> I am trying to check gauge invariance and lorentz invariance of a process in a New Physics Model. It is showing the following lines and giving the results.
>
> PJFRY not available on your system; it will be skipped.
> GOLEM not available on your system; it will be skipped.
>
>
> 1.As this lines are coming, can I still believe the results ?
> 2. How to install PJFRY and GOLEM ?
>
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                           I have another query regarding Lorentz Invariance Test. When I am checking the lorentz invariance test of p p > w- w+ in SM, it is giving the following results.

Lorentz invariance results:
Process Min element Max element Relative diff. Result
u u~ > w- w+ 3.2478225552e-02 3.2478225552e-02 3.6320086569e-15 Passed
d d~ > w- w+ 4.3951549579e-02 4.3951549579e-02 2.8417676161e-15 Passed
Summary: 2/2 passed, 0/2 failed
Not checked processes: c c~ > w- w+, s s~ > w- w+

My worry is why it is not checking the lorentz invariance of all the processes.

Regards,
Shiba

Those matrix elements should be identical to those tested. Therefore this is pointless to redo the same test.

Cheers,

Olivier
On 14 Oct 2015, at 18:47, shibasipu <email address hidden> wrote:

> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> I have another query regarding Lorentz Invariance Test. When I am checking the lorentz invariance test of p p > w- w+ in SM, it is giving the following results.
>
>
> Lorentz invariance results:
> Process Min element Max element Relative diff. Result
> u u~ > w- w+ 3.2478225552e-02 3.2478225552e-02 3.6320086569e-15 Passed
> d d~ > w- w+ 4.3951549579e-02 4.3951549579e-02 2.8417676161e-15 Passed
> Summary: 2/2 passed, 0/2 failed
> Not checked processes: c c~ > w- w+, s s~ > w- w+
>
>
> My worry is why it is not checking the lorentz invariance of all the processes.
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                              thanks for your reply. I have another query regarding the gauge invariance test. Suppose I have generated the UFO model file such that it is compatible with both gauges i.e. it can do the gauge invariance test. One particular process fails the gauge invariance test. Now I have generated the UFO model file in unitary gauge by changing " gauge = [0]" in the __init__.py file. Now I can not do the gauge invariance test of that particular process. My question is can I study the that particular process in the UFO model in unitary gauge ?

Regards,
Shiba

Hi,

> Now I have generated the UFO model file in unitary gauge by changing " gauge = [0]" in the __init__.py file.

> My question is can I study the that particular process in the UFO model in unitary gauge ?

If you have just changed the __init__.py file, then I would not trust the model.
If I were you, I would first try to understand why the gauge test fails and after that you can see if the unitary gauge is meaningful or not.

Cheers,

Olivier

On 15 Oct 2015, at 11:33, shibasipu <email address hidden> wrote:

> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> thanks for your reply. I have another query regarding the gauge invariance test. Suppose I have generated the UFO model file such that it is compatible with both gauges i.e. it can do the gauge invariance test. One particular process fails the gauge invariance test. Now I have generated the UFO model file in unitary gauge by changing " gauge = [0]" in the __init__.py file. Now I can not do the gauge invariance test of that particular process. My question is can I study the that particular process in the UFO model in unitary gauge ?
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                           Now I understood that to generate UFO model file in the unitary gauge I have to change at the .fr level i.e. I have to change "FeynmanGauge = True" to "FeynmanGauge = False" . Still I am not clear why the gauge test fails. when I generate the process
it looks all the diagrams are possible. For your reference, i am quoting the gauge invariance test test results.

Gauge results (switching between Unitary/Feynman):
Process Unitary Feynman Relative diff. Result
u u~ > h+ w- h2 1.4798311271e-11 1.0052063142e-07 9.9985278334e-01 Failed
   JAMP 0 1.7757973525e-10 1.2062475770e-06 9.9985278334e-01 Failed
u c~ > h+ w- h2 2.9730535555e-01 2.1009855216e-06 9.9999293324e-01 Failed
   JAMP 0 3.5676642666e+00 2.5211826260e-05 9.9999293324e-01 Failed
c u~ > h+ w- h2 3.8828022181e-01 2.1009641411e-06 9.9999458905e-01 Failed
   JAMP 0 4.6593626617e+00 2.5211569693e-05 9.9999458905e-01 Failed
c c~ > h+ w- h2 3.6297368270e-05 5.1737380265e-05 2.9843049487e-01 Failed
   JAMP 0 4.3556841924e-04 6.2084856318e-04 2.9843049487e-01 Failed
d d~ > h+ w- h2 2.5075813830e-10 7.3446025853e-10 6.5858174709e-01 Failed
   JAMP 0 3.0090976596e-09 8.8135231023e-09 6.5858174709e-01 Failed
d s~ > h+ w- h2 2.6775931425e-01 1.3845276706e-08 9.9999994829e-01 Failed
   JAMP 0 3.2131117710e+00 1.6614332047e-07 9.9999994829e-01 Failed
s d~ > h+ w- h2 9.5678330203e-02 1.3851051486e-08 9.9999985523e-01 Failed
   JAMP 0 1.1481399624e+00 1.6621261783e-07 9.9999985523e-01 Failed
s s~ > h+ w- h2 2.5236069686e-07 2.6075144961e-07 3.2179122151e-02 Failed
   JAMP 0 3.0283283623e-06 3.1290173953e-06 3.2179122151e-02 Failed
Summary: 0/8 passed, 8/8 failed
Failed processes: u u~ > h+ w- h2, u c~ > h+ w- h2, c u~ > h+ w- h2, c c~ > h+ w- h2, d d~ > h+ w- h2, d s~ > h+ w- h2, s d~ > h+ w- h2, s s~ > h+ w- h2

Actually, I am not getting which thing should I look for so that I can understand why this gauge invariance test fails.

Regards,
Shiba

HI,

Do you have the yukawa coupling equal to the pole mass?

Cheers,

Olivier

On 15 Oct 2015, at 12:57, shibasipu <email address hidden> wrote:

> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> Now I understood that to generate UFO model file in the unitary gauge I have to change at the .fr level i.e. I have to change "FeynmanGauge = True" to "FeynmanGauge = False" . Still I am not clear why the gauge test fails. when I generate the process
> it looks all the diagrams are possible. For your reference, i am quoting the gauge invariance test test results.
>
>
> Gauge results (switching between Unitary/Feynman):
> Process Unitary Feynman Relative diff. Result
> u u~ > h+ w- h2 1.4798311271e-11 1.0052063142e-07 9.9985278334e-01 Failed
> JAMP 0 1.7757973525e-10 1.2062475770e-06 9.9985278334e-01 Failed
> u c~ > h+ w- h2 2.9730535555e-01 2.1009855216e-06 9.9999293324e-01 Failed
> JAMP 0 3.5676642666e+00 2.5211826260e-05 9.9999293324e-01 Failed
> c u~ > h+ w- h2 3.8828022181e-01 2.1009641411e-06 9.9999458905e-01 Failed
> JAMP 0 4.6593626617e+00 2.5211569693e-05 9.9999458905e-01 Failed
> c c~ > h+ w- h2 3.6297368270e-05 5.1737380265e-05 2.9843049487e-01 Failed
> JAMP 0 4.3556841924e-04 6.2084856318e-04 2.9843049487e-01 Failed
> d d~ > h+ w- h2 2.5075813830e-10 7.3446025853e-10 6.5858174709e-01 Failed
> JAMP 0 3.0090976596e-09 8.8135231023e-09 6.5858174709e-01 Failed
> d s~ > h+ w- h2 2.6775931425e-01 1.3845276706e-08 9.9999994829e-01 Failed
> JAMP 0 3.2131117710e+00 1.6614332047e-07 9.9999994829e-01 Failed
> s d~ > h+ w- h2 9.5678330203e-02 1.3851051486e-08 9.9999985523e-01 Failed
> JAMP 0 1.1481399624e+00 1.6621261783e-07 9.9999985523e-01 Failed
> s s~ > h+ w- h2 2.5236069686e-07 2.6075144961e-07 3.2179122151e-02 Failed
> JAMP 0 3.0283283623e-06 3.1290173953e-06 3.2179122151e-02 Failed
> Summary: 0/8 passed, 8/8 failed
> Failed processes: u u~ > h+ w- h2, u c~ > h+ w- h2, c u~ > h+ w- h2, c c~ > h+ w- h2, d d~ > h+ w- h2, d s~ > h+ w- h2, s d~ > h+ w- h2, s s~ > h+ w- h2
>
> Actually, I am not getting which thing should I look for so that I can
> understand why this gauge invariance test fails.
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                          Yes. Please have a look to the below.

V_112 = Vertex(name = 'V_112',
               particles = [ P.T, P.t, P.h ],
               color = [ 'Identity(1,2)' ],
               lorentz = [ L.FFS2 ],
               couplings = {(0,0):C.GC_87})

GC_87 = Coupling(name = 'GC_87',
                 value = '-((complex(0,1)*MT)/v)',
                 order = {'QED':1})
 where MT = 175 GeV

Is this the reason to fail the gauge invariance test ?

Regards,
Shiba

This sounds correct.

I have heard that FR will release a new version of the 2HDM model with some change for the goldstone sector.
Please check if this has an impact on your process.

Cheers,

Olivier

Dear Olivier,
                         There is a small error in the 2HDM potential also. Now, Celine fixed it. Now the same process is gauge invariant. But I am checking the bug in the Inert doublet Model file where that process is not gauge invariant. I am looking into the model file. If I am not able to solve this issue, I may need your help.

Regards,
Shiba

Dear Olivier,
                              As far I understand MG5 does the gauge invariant check by Ward Identities. When I am checking i.e. "check p p > h+ w- h2" it is showing no ward identity for the subprocesses. I am pasting the results in the below.

INFO: No ward identity for process: u u > h+ w- h2
INFO: No ward identity for process: u c > h+ w- h2
INFO: No ward identity for process: u d > h+ w- h2
INFO: No ward identity for process: u s > h+ w- h2
INFO: No ward identity for process: u u~ > h+ w- h2
INFO: No ward identity for process: u c~ > h+ w- h2
INFO: No ward identity for process: u d~ > h+ w- h2
INFO: No ward identity for process: u s~ > h+ w- h2
INFO: No ward identity for process: c c > h+ w- h2
INFO: No ward identity for process: c d > h+ w- h2
INFO: No ward identity for process: c s > h+ w- h2
INFO: No ward identity for process: c u~ > h+ w- h2
INFO: No ward identity for process: c c~ > h+ w- h2
INFO: No ward identity for process: c d~ > h+ w- h2
INFO: No ward identity for process: c s~ > h+ w- h2
INFO: No ward identity for process: d d > h+ w- h2
INFO: No ward identity for process: d s > h+ w- h2
INFO: No ward identity for process: d u~ > h+ w- h2
INFO: No ward identity for process: d c~ > h+ w- h2
INFO: No ward identity for process: d d~ > h+ w- h2
INFO: No ward identity for process: d s~ > h+ w- h2
INFO: No ward identity for process: s s > h+ w- h2
INFO: No ward identity for process: s u~ > h+ w- h2
INFO: No ward identity for process: s c~ > h+ w- h2
INFO: No ward identity for process: s d~ > h+ w- h2
INFO: No ward identity for process: s s~ > h+ w- h2
INFO: No ward identity for process: u~ u~ > h+ w- h2
INFO: No ward identity for process: u~ c~ > h+ w- h2
INFO: No ward identity for process: u~ d~ > h+ w- h2
INFO: No ward identity for process: u~ s~ > h+ w- h2
INFO: No ward identity for process: c~ c~ > h+ w- h2
INFO: No ward identity for process: c~ d~ > h+ w- h2
INFO: No ward identity for process: c~ s~ > h+ w- h2
INFO: No ward identity for process: d~ d~ > h+ w- h2
INFO: No ward identity for process: d~ s~ > h+ w- h2
INFO: No ward identity for process: s~ s~ > h+ w- h2

Is there problem with this ?

Regards,
Shiba

That test works only if you have a photon/gluon in the initial/final state.

Cheers,

Olivier
On 16 Oct 2015, at 19:32, shibasipu <email address hidden> wrote:

> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> shibasipu gave more information on the question:
> Dear Olivier,
> As far I understand MG5 does the gauge invariant check by Ward Identities. When I am checking i.e. "check p p > h+ w- h2" it is showing no ward identity for the subprocesses. I am pasting the results in the below.
>
> INFO: No ward identity for process: u u > h+ w- h2
> INFO: No ward identity for process: u c > h+ w- h2
> INFO: No ward identity for process: u d > h+ w- h2
> INFO: No ward identity for process: u s > h+ w- h2
> INFO: No ward identity for process: u u~ > h+ w- h2
> INFO: No ward identity for process: u c~ > h+ w- h2
> INFO: No ward identity for process: u d~ > h+ w- h2
> INFO: No ward identity for process: u s~ > h+ w- h2
> INFO: No ward identity for process: c c > h+ w- h2
> INFO: No ward identity for process: c d > h+ w- h2
> INFO: No ward identity for process: c s > h+ w- h2
> INFO: No ward identity for process: c u~ > h+ w- h2
> INFO: No ward identity for process: c c~ > h+ w- h2
> INFO: No ward identity for process: c d~ > h+ w- h2
> INFO: No ward identity for process: c s~ > h+ w- h2
> INFO: No ward identity for process: d d > h+ w- h2
> INFO: No ward identity for process: d s > h+ w- h2
> INFO: No ward identity for process: d u~ > h+ w- h2
> INFO: No ward identity for process: d c~ > h+ w- h2
> INFO: No ward identity for process: d d~ > h+ w- h2
> INFO: No ward identity for process: d s~ > h+ w- h2
> INFO: No ward identity for process: s s > h+ w- h2
> INFO: No ward identity for process: s u~ > h+ w- h2
> INFO: No ward identity for process: s c~ > h+ w- h2
> INFO: No ward identity for process: s d~ > h+ w- h2
> INFO: No ward identity for process: s s~ > h+ w- h2
> INFO: No ward identity for process: u~ u~ > h+ w- h2
> INFO: No ward identity for process: u~ c~ > h+ w- h2
> INFO: No ward identity for process: u~ d~ > h+ w- h2
> INFO: No ward identity for process: u~ s~ > h+ w- h2
> INFO: No ward identity for process: c~ c~ > h+ w- h2
> INFO: No ward identity for process: c~ d~ > h+ w- h2
> INFO: No ward identity for process: c~ s~ > h+ w- h2
> INFO: No ward identity for process: d~ d~ > h+ w- h2
> INFO: No ward identity for process: d~ s~ > h+ w- h2
> INFO: No ward identity for process: s~ s~ > h+ w- h2
>
> Is there problem with this ?
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                          Sorry, I need one clarification regarding your comment. Are you saying the gauge invariant test will only work when I have a photon/gluon in the initial/final state ? i.e. the command " check gauge p p > h+ w- h2" will work only when I have a photon/gluon in the initial/final state ? Please suggest some references so that I will learn more about these tests. Actually I am looking the ref. 1209.0297 (Page 23 - 25) for these test. But the detailed thing about these tests are not given.

Regards,
Shiba

We can only test ward identity if we have a photon/gluon in the initial/final state.
If you do not have a photon/gluon, we print the message that you quote and pass to the next subprocesses.

> Actually I am looking the ref. 1209.0297 (Page 23 - 25) for these test. But the detailed thing about these tests are not given.

That’s the only reference we have.

Cheers,

Olivier

On 17 Oct 2015, at 06:27, shibasipu <email address hidden> wrote:

> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> Sorry, I need one clarification regarding your comment. Are you saying the gauge invariant test will only work when I have a photon/gluon in the initial/final state ? i.e. the command " check gauge p p > h+ w- h2" will work only when I have a photon/gluon in the initial/final state ? Please suggest some references so that I will learn more about these tests. Actually I am looking the ref. 1209.0297 (Page 23 - 25) for these test. But the detailed thing about these tests are not given.
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                          I need one clarification regarding gauge invariance test. When I am checking "check u u~ > h+ w- h2", it is giving following results.

Gauge results (switching between Unitary/Feynman):
Process Unitary Feynman Relative diff. Result
u u~ > h+ w- h2 2.3518224324e-11 1.0054281360e-07 9.9976608747e-01 Failed
   JAMP 0 2.8221869189e-10 1.2065137633e-06 9.9976608747e-01 Failed
Summary: 0/1 passed, 1/1 failed
Failed processes: u u~ > h+ w- h2

Basically, I want to know the meaning of these columns. I think it calculates the matrix elements in both the gauges for a particular process and then takes the diffrences. Is it correct ?

Why the second line is coming? what is the significance ?
JAMP 0 2.8221869189e-10 1.2065137633e-06 9.9976608747e-01 Failed

Regards,
Shiba

Hi,

The JAMP information corresponds to the various color flow of the events.
In this case, you have only one color-flow and this information is therefore not very interesting.

Cheers,

Olivier

> On Oct 19, 2015, at 05:21, shibasipu <email address hidden> wrote:
>
> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> I need one clarification regarding gauge invariance test. When I am checking "check u u~ > h+ w- h2", it is giving following results.
>
> Gauge results (switching between Unitary/Feynman):
> Process Unitary Feynman Relative diff. Result
> u u~ > h+ w- h2 2.3518224324e-11 1.0054281360e-07 9.9976608747e-01 Failed
> JAMP 0 2.8221869189e-10 1.2065137633e-06 9.9976608747e-01 Failed
> Summary: 0/1 passed, 1/1 failed
> Failed processes: u u~ > h+ w- h2
>
> Basically, I want to know the meaning of these columns. I think it
> calculates the matrix elements in both the gauges for a particular
> process and then takes the diffrences. Is it correct ?
>
> Why the second line is coming? what is the significance ?
> JAMP 0 2.8221869189e-10 1.2065137633e-06 9.9976608747e-01 Failed
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                           thanks for your reply. Please comment something to some portion of my previous query ( I think you forgot to comment )

"Gauge results (switching between Unitary/Feynman):
Process Unitary Feynman Relative diff. Result
u u~ > h+ w- h2 2.3518224324e-11 1.0054281360e-07 9.9976608747e-01 Failed
   JAMP 0 2.8221869189e-10 1.2065137633e-06 9.9976608747e-01 Failed
Summary: 0/1 passed, 1/1 failed
Failed processes: u u~ > h+ w- h2

Basically, I want to know the meaning of these columns. I think it calculates the matrix elements in both the gauges for a particular process and then takes the diffrences. Is it correct ?

"

Hi,

> Basically, I want to know the meaning of these columns. I think it
> calculates the matrix elements in both the gauges for a particular
> process and then takes the diffrences. Is it correct ?

yes that’s the value of the matrix-element for a given PS point.
The third column corresponds to
| ME_F - ME_U | / max(ME_F, ME_U)

Cheers,

Olivier

> On Oct 19, 2015, at 11:22, shibasipu <email address hidden> wrote:
>
> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> thanks for your reply. Please comment something to some portion of my previous query ( I think you forgot to comment )
>
> "Gauge results (switching between Unitary/Feynman):
> Process Unitary Feynman Relative diff. Result
> u u~ > h+ w- h2 2.3518224324e-11 1.0054281360e-07 9.9976608747e-01 Failed
> JAMP 0 2.8221869189e-10 1.2065137633e-06 9.9976608747e-01 Failed
> Summary: 0/1 passed, 1/1 failed
> Failed processes: u u~ > h+ w- h2
>
> Basically, I want to know the meaning of these columns. I think it
> calculates the matrix elements in both the gauges for a particular
> process and then takes the diffrences. Is it correct ?
>
> "
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
(1) are these tests are at amplitude level or amplitude squared level ? I think, since it is calculating matrix element, it must be at amplitude level. Am I correct ?

(2) In the Inert Doublet model , u d~ > w+ z , c s~ > w+ z passes the gauge invariant test and u s~ > w+ z and c d~ > w+ z
do not pass the test. Is there some problem with .fr file ? Since all the above processes are of same type, I am not sure that
there will be some problem in the .fr file.

Regards,
Shiba

Regards,
Shiba

Hi,

> (1) are these tests are at amplitude level or amplitude squared level ? I think, since it is calculating matrix element, it must be at amplitude level. Am I correct ?

No it is amplitude square.

> (2) In the Inert Doublet model , u d~ > w+ z , c s~ > w+ z passes the gauge invariant test and u s~ > w+ z and c d~ > w+ z
> do not pass the test. Is there some problem with .fr file ? Since all the above processes are of same type, I am not sure that
> there will be some problem in the .fr file.

I do not know.
If the lorentz test is passing, then the problem is likely to be in the goldstone sector.
If the permutation tests is failing then it typically means that the error is on the madgraph/aloha side.

Cheers,

Olivier

> On Oct 19, 2015, at 13:46, shibasipu <email address hidden> wrote:
>
> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> (1) are these tests are at amplitude level or amplitude squared level ? I think, since it is calculating matrix element, it must be at amplitude level. Am I correct ?
>
>
> (2) In the Inert Doublet model , u d~ > w+ z , c s~ > w+ z passes the gauge invariant test and u s~ > w+ z and c d~ > w+ z
> do not pass the test. Is there some problem with .fr file ? Since all the above processes are of same type, I am not sure that
> there will be some problem in the .fr file.
>
>
> Regards,
> Shiba
>
>
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                                I need one clarification regarding previous query no. 1. I think GI test only valid at amplitude level . How MG doing it at amplitude square level ? Please clarify.

 (2) the processes u s~ > w+ z and c d~ > w+ z fails the GI test but fails the LI test. So as you have already pointed about it, I will check .fr file. Also, these processes pass the permutation test, so there is no problem from Madgraph side .

If some processes pass the LI test, then fail GI test, then error from which side i.e. from .fr file or from Madgraph ?

Regards,
Shiba

Dear Olivier,
                         sorry. there is a typo in my query 2 i.e. Please read the processes u s~ > w+ z and c d~ > w+ z fail the GI test but pass the LI test.

Regards,
Shiba

Hi,

> I think GI test only valid at amplitude level .

If you have a=b then you have a^2 = b^2.
So if you think that a test is valid at the amplitude level, the same test should be valid at the amplitude square level.

For the gauge test, we just compute the amplitude square in both gauge and check if they are equal.

> If some processes pass the LI test, then fail GI test, then error from which side i.e. from .fr file or from Madgraph ?

This suggests a wrong implementation of the Feynman gauge at the model level.
It can be at the FR level or just the fact that you have a difference between the pole mass and the the yukawa coupling.

Cheers,

Olivier

> On Oct 19, 2015, at 16:07, shibasipu <email address hidden> wrote:
>
> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> I need one clarification regarding previous query no. 1. I think GI test only valid at amplitude level . How MG doing it at amplitude square level ? Please clarify.
>
> (2) the processes u s~ > w+ z and c d~ > w+ z fails the GI
> test but fails the LI test. So as you have already pointed
> about it, I will check .fr file. Also, these processes pass
> the permutation test, so there is no problem from Madgraph side
> .
>
>
> If some processes pass the LI test, then fail GI test, then error from which side i.e. from .fr file or from Madgraph ?
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                              when I am checking the gauge invariance test, the following line is coming?

" Note That all width have been set to zero for those checks . "

While checking the gauge invariance , MG5 sets all the width automatically to zero ? or I have to set all the widths to zero, then do the gauge invariance check ?

Regards,
Shiba

It is automatically done.

Cheers,

Olivier
> On Oct 22, 2015, at 13:37, shibasipu <email address hidden> wrote:
>
> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> when I am checking the gauge invariance test, the following line is coming?
>
> " Note That all width have been set to zero for those checks . "
>
> While checking the gauge invariance , MG5 sets all the width
> automatically to zero ? or I have to set all the widths to zero,
> then do the gauge invariance check ?
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                       thanks. I need another clarification regarding gauge invariance test. I am checking GI i.e. " check gauge c u~ > h+ w- h2" in Inert doublet model(setting yukawas to zero except for t and b). Why it is failing GI test as for mu = mc =0, there is no gauge dependence ?

Regards,
Shiba

The code do not know anything about the gauge,it just runs the computation done in two different gauge.
i.e. the spin1 particle have different propagator/ the goldstone are allowed/forbidden.
It fails if the two computation do not return the same number.

If this fails and not the other tests, it means that your model has a problem with the goldstone coupling or something like that.

Cheers,

Olivier

> On Oct 22, 2015, at 14:27, shibasipu <email address hidden> wrote:
>
> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> thanks. I need another clarification regarding gauge invariance test. I am checking GI i.e. " check gauge c u~ > h+ w- h2" in Inert doublet model(setting yukawas to zero except for t and b). Why it is failing GI test as for mu = mc =0, there is no gauge dependence ?
>
>
>
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                           Now the processes p p > W+ Z , p p > W+ h , p p > h+ h2 , p p > h+ h3 passes the gauge invariance test (previously they are not) in the Inert Doublet Model file after imposing massless restrictions (setting all yukawas/masses to zero except top , bottom ,tau lepton). But our desired process p p > h+ w- h2 still not passing Gauge Invariance test. I have also written to the Authors of the model file. They are also going to this faq. But still they are also not getting the error any where. Please suggest where to look the bug in the .fr file ?

Regards,
Shiba

I have no idea of what a .fr looks like. So I can not help.

Cheers,

Olivier
> On Oct 23, 2015, at 08:13, shibasipu <email address hidden> wrote:
>
> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> Now the processes p p > W+ Z , p p > W+ h , p p > h+ h2 , p p > h+ h3 passes the gauge invariance test (previously they are not) in the Inert Doublet Model file after imposing massless restrictions (setting all yukawas/masses to zero except top , bottom ,tau lepton). But our desired process p p > h+ w- h2 still not passing Gauge Invariance test. I have also written to the Authors of the model file. They are also going to this faq. But still they are also not getting the error any where. Please suggest where to look the bug in the .fr file ?
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Dear Olivier,
                        apologizing for lengthing this thread ! I want to understand the option " check lorentz process name".
From help option : lorentz_invariance:
   Check that the amplitude is lorentz invariant by
   comparing the amplitiude in different frames

I understood the above thing. I want to know what MG5 is doing ? for example :

Lorentz invariance results:
Process Min element Max element Relative diff. Result
u u~ > w- w+ h 9.7536026347e-09 9.7536026347e-09 1.6673193949e-13 Passed
Summary: 1/1 passed, 0/1 failed

what is the significance of these columns i.e. Min element column ? Max element column ?

Regards,
Shiba

Hi,

We compute the matrix element in 4 frame.
1) the lab.
2) boosted in the x direction (the beta=0.5 if I remember correctly)
3) boosted in the y direction (same beta)
4) boosted in the z direction

This provides 4 different value for the matrix element square.
In this output, we display the minimal and maximal value of those four.
The “relative difference” is the (max - min)/max

Cheers,

Olivier

> On Oct 23, 2015, at 13:13, shibasipu <email address hidden> wrote:
>
> Question #272412 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/272412
>
> Status: Answered => Open
>
> shibasipu is still having a problem:
> Dear Olivier,
> apologizing for lengthing this thread ! I want to understand the option " check lorentz process name".
>> From help option : lorentz_invariance:
> Check that the amplitude is lorentz invariant by
> comparing the amplitiude in different frames
>
> I understood the above thing. I want to know what MG5 is doing ?
> for example :
>
> Lorentz invariance results:
> Process Min element Max element Relative diff. Result
> u u~ > w- w+ h 9.7536026347e-09 9.7536026347e-09 1.6673193949e-13 Passed
> Summary: 1/1 passed, 0/1 failed
>
> what is the significance of these columns i.e. Min element column ?
> Max element column ?
>
>
> Regards,
> Shiba
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.