MadGraph5_aMC@NLO

event weight not in agreement with xsection

Bug #1569506 reported by Eleni Vryonidou
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
MadGraph5_aMC@NLO
Fix Released
Undecided
Unassigned

Bug Description

Hi guys,

I was running some simple loop process 2>1 and noticed that for one of my parameter setups the event weight does not agree with the cross-section given on the screen at the end of the run or the one in the run_tag file.

This is the end of the run_tag file:

</slha>
<MGGenerationInfo>
# Number of Events : 5000
# Integrated weight (pb) : -2.3347
</MGGenerationInfo>
</header>
</LesHouchesEvents>

and the event file I have:

<MGGenerationInfo>
# Number of Events : 5000
# Integrated weight (pb) : -2.3347
</MGGenerationInfo>
</header>
<init>
2212 2212 6.500000e+03 6.500000e+03 0 0 25000 25000 -3 1
-2.334700e+00 9.036100e-03 4.669400e-04 0
<generator name='MadGraph5_aMC@NLO' version='5.2.4.0.beta'>please cite 1405.0301 </generator>
</init>
<event>
 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
       21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +5.7883565162e+01 5.7883565162e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
       21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -6.7484613103e+01 6.7484613103e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
       25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 -9.6010479412e+00 1.2536817827e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
</event>
<event>
 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
       21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +1.7900225751e+02 1.7900225751e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
       21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -2.1822350479e+01 2.1822350479e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
       25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 +1.5717990703e+02 2.0082460799e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
</event>

When the event weight should be -2.3347/5000=4.67E-4...

This seems to randomly happen. Other runs in the same output are fine. Do you have any idea what is causing this? I am using 2.3.4. I tried with more events and the event weight numbers seem to change randomly.

Cheers,

Eleni

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) wrote : Re: [Bug 1569506] [NEW] event weight not in agreement with xsection
Download full text (5.0 KiB)

Hi,

Are you sure that all your event have negative weight?
If not then this is normal.

Cheers,

Olivier

> On Apr 12, 2016, at 19:22, Eleni Vryonidou <email address hidden> wrote:
>
> Public bug reported:
>
> Hi guys,
>
> I was running some simple loop process 2>1 and noticed that for one of
> my parameter setups the event weight does not agree with the cross-
> section given on the screen at the end of the run or the one in the
> run_tag file.
>
> This is the end of the run_tag file:
>
> </slha>
> <MGGenerationInfo>
> # Number of Events : 5000
> # Integrated weight (pb) : -2.3347
> </MGGenerationInfo>
> </header>
> </LesHouchesEvents>
>
> and the event file I have:
>
> <MGGenerationInfo>
> # Number of Events : 5000
> # Integrated weight (pb) : -2.3347
> </MGGenerationInfo>
> </header>
> <init>
> 2212 2212 6.500000e+03 6.500000e+03 0 0 25000 25000 -3 1
> -2.334700e+00 9.036100e-03 4.669400e-04 0
> <generator name='MadGraph5_aMC@NLO' version='5.2.4.0.beta'>please cite 1405.0301 </generator>
> </init>
> <event>
> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +5.7883565162e+01 5.7883565162e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -6.7484613103e+01 6.7484613103e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 -9.6010479412e+00 1.2536817827e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
> </event>
> <event>
> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +1.7900225751e+02 1.7900225751e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -2.1822350479e+01 2.1822350479e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 +1.5717990703e+02 2.0082460799e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
> </event>
>
> When the event weight should be -2.3347/5000=4.67E-4...
>
> This seems to randomly happen. Other runs in the same output are fine.
> Do you have any idea what is causing this? I am using 2.3.4. I tried
> with more events and the event weight numbers seem to change randomly.
>
> Cheers,
>
> Eleni
>
> ** Affects: mg5amcnlo
> Importance: Undecided
> Status: New
>
> --
> You received this bug notification because you are subscribed to
> MadGraph5_aMC@NLO.
> https://bugs.launchpad.net/bugs/1569506
>
> Title:
> event weight not in agreement with xsection
>
> Status in MadGraph5_aMC@NLO:
> New
>
> Bug description:
> Hi guys,
>
> I was running some simple loop process 2>1 and noticed that for one of
> my parameter setups the event weight does not agree with the cross-
> section given on the screen at the end of the run or the one in the
> run_tag file.
>
> This is the end of the run_tag file:
>
> </slha>
> <MGGenerationInfo>
> # Number of Events : 5000
> # Integrated weight (pb) : -2.3347
> </MGGenerationInfo>
...

Read more...

Revision history for this message
Eleni Vryonidou (evryonidou) wrote :
Download full text (8.0 KiB)

Hi Olivier,

I did this:

mac-1n0-345:run_07 evryonidou$ grep -rc '3 0 -4.1417436e-04' unweighted_events.lhe
unweighted_events.lhe:5000

It seems all events have the same weight.

Cheers,
Eleni

> On 12 Apr 2016, at 21:28, Olivier Mattelaer <email address hidden> wrote:
>
> Hi,
>
> Are you sure that all your event have negative weight?
> If not then this is normal.
>
> Cheers,
>
> Olivier
>
>> On Apr 12, 2016, at 19:22, Eleni Vryonidou <email address hidden> wrote:
>>
>> Public bug reported:
>>
>> Hi guys,
>>
>> I was running some simple loop process 2>1 and noticed that for one of
>> my parameter setups the event weight does not agree with the cross-
>> section given on the screen at the end of the run or the one in the
>> run_tag file.
>>
>> This is the end of the run_tag file:
>>
>> </slha>
>> <MGGenerationInfo>
>> # Number of Events : 5000
>> # Integrated weight (pb) : -2.3347
>> </MGGenerationInfo>
>> </header>
>> </LesHouchesEvents>
>>
>> and the event file I have:
>>
>> <MGGenerationInfo>
>> # Number of Events : 5000
>> # Integrated weight (pb) : -2.3347
>> </MGGenerationInfo>
>> </header>
>> <init>
>> 2212 2212 6.500000e+03 6.500000e+03 0 0 25000 25000 -3 1
>> -2.334700e+00 9.036100e-03 4.669400e-04 0
>> <generator name='MadGraph5_aMC@NLO' version='5.2.4.0.beta'>please cite 1405.0301 </generator>
>> </init>
>> <event>
>> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
>> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +5.7883565162e+01 5.7883565162e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
>> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -6.7484613103e+01 6.7484613103e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
>> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 -9.6010479412e+00 1.2536817827e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
>> </event>
>> <event>
>> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
>> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +1.7900225751e+02 1.7900225751e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
>> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -2.1822350479e+01 2.1822350479e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
>> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 +1.5717990703e+02 2.0082460799e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
>> </event>
>>
>> When the event weight should be -2.3347/5000=4.67E-4...
>>
>> This seems to randomly happen. Other runs in the same output are fine.
>> Do you have any idea what is causing this? I am using 2.3.4. I tried
>> with more events and the event weight numbers seem to change randomly.
>>
>> Cheers,
>>
>> Eleni
>>
>> ** Affects: mg5amcnlo
>> Importance: Undecided
>> Status: New
>>
>> --
>> You received this bug notification because you are subscribed to
>> MadGraph5_aMC@NLO.
>> https://bugs.launchpad.net/bugs/1569506
>>
>> Title:
>> event weight not in agreement with xsection
>>
>> Status in MadGraph5_aMC@NLO:
>> New
>>
>> Bug description:
>> Hi guys,
>>
>> I was running some simple loop process 2>1 and n...

Read more...

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) wrote :
Download full text (10.9 KiB)

Hi,

Could you give me the model/benchmark/process definition that you use?
I will try to reproduce it locally.

Cheers,

Olivier

> On Apr 12, 2016, at 20:39, Eleni Vryonidou <email address hidden> wrote:
>
> Hi Olivier,
>
> I did this:
>
>
> mac-1n0-345:run_07 evryonidou$ grep -rc '3 0 -4.1417436e-04' unweighted_events.lhe
> unweighted_events.lhe:5000
>
>
> It seems all events have the same weight.
>
> Cheers,
> Eleni
>
>
>> On 12 Apr 2016, at 21:28, Olivier Mattelaer <email address hidden> wrote:
>>
>> Hi,
>>
>> Are you sure that all your event have negative weight?
>> If not then this is normal.
>>
>> Cheers,
>>
>> Olivier
>>
>>> On Apr 12, 2016, at 19:22, Eleni Vryonidou <email address hidden> wrote:
>>>
>>> Public bug reported:
>>>
>>> Hi guys,
>>>
>>> I was running some simple loop process 2>1 and noticed that for one of
>>> my parameter setups the event weight does not agree with the cross-
>>> section given on the screen at the end of the run or the one in the
>>> run_tag file.
>>>
>>> This is the end of the run_tag file:
>>>
>>> </slha>
>>> <MGGenerationInfo>
>>> # Number of Events : 5000
>>> # Integrated weight (pb) : -2.3347
>>> </MGGenerationInfo>
>>> </header>
>>> </LesHouchesEvents>
>>>
>>> and the event file I have:
>>>
>>> <MGGenerationInfo>
>>> # Number of Events : 5000
>>> # Integrated weight (pb) : -2.3347
>>> </MGGenerationInfo>
>>> </header>
>>> <init>
>>> 2212 2212 6.500000e+03 6.500000e+03 0 0 25000 25000 -3 1
>>> -2.334700e+00 9.036100e-03 4.669400e-04 0
>>> <generator name='MadGraph5_aMC@NLO' version='5.2.4.0.beta'>please cite 1405.0301 </generator>
>>> </init>
>>> <event>
>>> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
>>> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +5.7883565162e+01 5.7883565162e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
>>> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -6.7484613103e+01 6.7484613103e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
>>> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 -9.6010479412e+00 1.2536817827e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
>>> </event>
>>> <event>
>>> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
>>> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +1.7900225751e+02 1.7900225751e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
>>> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -2.1822350479e+01 2.1822350479e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
>>> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 +1.5717990703e+02 2.0082460799e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
>>> </event>
>>>
>>> When the event weight should be -2.3347/5000=4.67E-4...
>>>
>>> This seems to randomly happen. Other runs in the same output are fine.
>>> Do you have any idea what is causing this? I am using 2.3.4. I tried
>>> with more events and the event weight numbers seem to change randomly.
>>>
>>> Cheers,
>>>
>>> Eleni
>>>
>>> ** Affects: mg5amcnlo
>>> Importance: Undecided
>>> Status: New
>>>
>>> --
>>> You received this bug...

Revision history for this message
Eleni Vryonidou (evryonidou) wrote :
Download full text (13.9 KiB)

Hi Olivier,

I sent them to you by email.

Thanks,

Eleni

> On 12 Apr 2016, at 21:53, Olivier Mattelaer <email address hidden> wrote:
>
> Hi,
>
> Could you give me the model/benchmark/process definition that you use?
> I will try to reproduce it locally.
>
> Cheers,
>
> Olivier
>
>> On Apr 12, 2016, at 20:39, Eleni Vryonidou <email address hidden> wrote:
>>
>> Hi Olivier,
>>
>> I did this:
>>
>>
>> mac-1n0-345:run_07 evryonidou$ grep -rc '3 0 -4.1417436e-04' unweighted_events.lhe
>> unweighted_events.lhe:5000
>>
>>
>> It seems all events have the same weight.
>>
>> Cheers,
>> Eleni
>>
>>
>>> On 12 Apr 2016, at 21:28, Olivier Mattelaer <email address hidden> wrote:
>>>
>>> Hi,
>>>
>>> Are you sure that all your event have negative weight?
>>> If not then this is normal.
>>>
>>> Cheers,
>>>
>>> Olivier
>>>
>>>> On Apr 12, 2016, at 19:22, Eleni Vryonidou <email address hidden> wrote:
>>>>
>>>> Public bug reported:
>>>>
>>>> Hi guys,
>>>>
>>>> I was running some simple loop process 2>1 and noticed that for one of
>>>> my parameter setups the event weight does not agree with the cross-
>>>> section given on the screen at the end of the run or the one in the
>>>> run_tag file.
>>>>
>>>> This is the end of the run_tag file:
>>>>
>>>> </slha>
>>>> <MGGenerationInfo>
>>>> # Number of Events : 5000
>>>> # Integrated weight (pb) : -2.3347
>>>> </MGGenerationInfo>
>>>> </header>
>>>> </LesHouchesEvents>
>>>>
>>>> and the event file I have:
>>>>
>>>> <MGGenerationInfo>
>>>> # Number of Events : 5000
>>>> # Integrated weight (pb) : -2.3347
>>>> </MGGenerationInfo>
>>>> </header>
>>>> <init>
>>>> 2212 2212 6.500000e+03 6.500000e+03 0 0 25000 25000 -3 1
>>>> -2.334700e+00 9.036100e-03 4.669400e-04 0
>>>> <generator name='MadGraph5_aMC@NLO' version='5.2.4.0.beta'>please cite 1405.0301 </generator>
>>>> </init>
>>>> <event>
>>>> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
>>>> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +5.7883565162e+01 5.7883565162e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
>>>> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -6.7484613103e+01 6.7484613103e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
>>>> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 -9.6010479412e+00 1.2536817827e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
>>>> </event>
>>>> <event>
>>>> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
>>>> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +1.7900225751e+02 1.7900225751e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
>>>> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -2.1822350479e+01 2.1822350479e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
>>>> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 +1.5717990703e+02 2.0082460799e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
>>>> </event>
>>>>
>>>> When the event weight should be -2.3347/5000=4.67E-4...
>>>>
>>>> This seems to randomly happen. Other runs in the same output are fine.
>>>> Do you have any idea what is causing this? I am using 2....

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) wrote :

Thanks,

I have fixed the problem in 2.4.1

Cheers,

Olivier

Changed in mg5amcnlo:
status: New → Fix Committed
Revision history for this message
Eleni Vryonidou (evryonidou) wrote : Re: [Bug 1569506] event weight not in agreement with xsection

It works fine now. Thanks a lot.

Cheers,
Eleni

> On 13 Apr 2016, at 02:44, Olivier Mattelaer <email address hidden> wrote:
>
> Thanks,
>
> I have fixed the problem in 2.4.1
>
> Cheers,
>
> Olivier
>
> ** Changed in: mg5amcnlo
> Status: New => Fix Committed
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1569506
>
> Title:
> event weight not in agreement with xsection
>
> Status in MadGraph5_aMC@NLO:
> Fix Committed
>
> Bug description:
> Hi guys,
>
> I was running some simple loop process 2>1 and noticed that for one of
> my parameter setups the event weight does not agree with the cross-
> section given on the screen at the end of the run or the one in the
> run_tag file.
>
> This is the end of the run_tag file:
>
> </slha>
> <MGGenerationInfo>
> # Number of Events : 5000
> # Integrated weight (pb) : -2.3347
> </MGGenerationInfo>
> </header>
> </LesHouchesEvents>
>
> and the event file I have:
>
> <MGGenerationInfo>
> # Number of Events : 5000
> # Integrated weight (pb) : -2.3347
> </MGGenerationInfo>
> </header>
> <init>
> 2212 2212 6.500000e+03 6.500000e+03 0 0 25000 25000 -3 1
> -2.334700e+00 9.036100e-03 4.669400e-04 0
> <generator name='MadGraph5_aMC@NLO' version='5.2.4.0.beta'>please cite 1405.0301 </generator>
> </init>
> <event>
> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +5.7883565162e+01 5.7883565162e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -6.7484613103e+01 6.7484613103e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 -9.6010479412e+00 1.2536817827e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
> </event>
> <event>
> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +1.7900225751e+02 1.7900225751e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -2.1822350479e+01 2.1822350479e+01 0.0000000000e+00 0.0000e+00 -1.0000e+00
> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 +1.5717990703e+02 2.0082460799e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
> </event>
>
> When the event weight should be -2.3347/5000=4.67E-4...
>
> This seems to randomly happen. Other runs in the same output are fine.
> Do you have any idea what is causing this? I am using 2.3.4. I tried
> with more events and the event weight numbers seem to change randomly.
>
> Cheers,
>
> Eleni
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/mg5amcnlo/+bug/1569506/+subscriptions

Revision history for this message
Eleni Vryonidou (evryonidou) wrote :
Download full text (3.7 KiB)

Hi Olivier,

I seem to find the same problem again with another process in 2.4.1

I noticed it by running SysCalc which gives slightly different central scale results compared to the output on the screen. Again the sum of the weights is not equal to the cross-section.

<MGGenerationInfo>
# Number of Events : 200000
# Integrated weight (pb) : -1.272194
</MGGenerationInfo>

scale fact cross-section :0 -1.74537
scale fact cross-section :1 -1.43312
scale fact cross-section :2 -1.19754
scale fact cross-section :3 -1.57691
scale fact cross-section :4 -1.29508
scale fact cross-section :5 -1.08237
scale fact cross-section :6 -1.43046
scale fact cross-section :7 -1.17502
scale fact cross-section :8 -0.982175

The sum of weights in the event file is equal to the scale fact cross-section :4

In this case there are both negative and positive event weights. Maybe the fix was not general enough?

Thanks,

Eleni

> On 13 Apr 2016, at 02:44, Olivier Mattelaer <email address hidden> wrote:
>
> Thanks,
>
> I have fixed the problem in 2.4.1
>
> Cheers,
>
> Olivier
>
> ** Changed in: mg5amcnlo
> Status: New => Fix Committed
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1569506
>
> Title:
> event weight not in agreement with xsection
>
> Status in MadGraph5_aMC@NLO:
> Fix Committed
>
> Bug description:
> Hi guys,
>
> I was running some simple loop process 2>1 and noticed that for one of
> my parameter setups the event weight does not agree with the cross-
> section given on the screen at the end of the run or the one in the
> run_tag file.
>
> This is the end of the run_tag file:
>
> </slha>
> <MGGenerationInfo>
> # Number of Events : 5000
> # Integrated weight (pb) : -2.3347
> </MGGenerationInfo>
> </header>
> </LesHouchesEvents>
>
> and the event file I have:
>
> <MGGenerationInfo>
> # Number of Events : 5000
> # Integrated weight (pb) : -2.3347
> </MGGenerationInfo>
> </header>
> <init>
> 2212 2212 6.500000e+03 6.500000e+03 0 0 25000 25000 -3 1
> -2.334700e+00 9.036100e-03 4.669400e-04 0
> <generator name='MadGraph5_aMC@NLO' version='5.2.4.0.beta'>please cite 1405.0301 </generator>
> </init>
> <event>
> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +5.7883565162e+01 5.7883565162e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -6.7484613103e+01 6.7484613103e+01 0.0000000000e+00 0.0000e+00 1.0000e+00
> 25 1 1 2 0 0 +0.0000000000e+00 +0.0000000000e+00 -9.6010479412e+00 1.2536817827e+02 1.2500000000e+02 0.0000e+00 0.0000e+00
> </event>
> <event>
> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
> 21 -1 0 0 502 501 +0.0000000000e+00 +0.0000000000e+00 +1.7900225751e+02 1.7900225751e+02 0.0000000000e+00 0.0000e+00 -1.0000e+00
> 21 -1 0 0 501 502 +0.0000000000e+00 +0.0000000000e+00 -2.1822350479e+01 2.1822350479e+01 0.00...

Read more...

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) wrote :
Download full text (7.0 KiB)

Hi Eleni,

This is just expected. The quantity which is preserved by the unweighting is the \sum abs(wgt)
and therefore the \sum wgt will not be exactly equal to the cross-section after unweighting.
The same actually occur in all NLO sample where the sum of the weight is not equal to the cross-sections.
(The difference is by construction within the statistical error, but can be higher than the precision on the cross-section since it depends on the number of event requested)

Cheers,

Olivier

In presence of negative and positive wei

Cheers,

Olivier

> On Apr 13, 2016, at 17:06, Eleni Vryonidou <email address hidden> wrote:
>
> Hi Olivier,
>
> I seem to find the same problem again with another process in 2.4.1
>
> I noticed it by running SysCalc which gives slightly different central
> scale results compared to the output on the screen. Again the sum of the
> weights is not equal to the cross-section.
>
> <MGGenerationInfo>
> # Number of Events : 200000
> # Integrated weight (pb) : -1.272194
> </MGGenerationInfo>
>
> scale fact cross-section :0 -1.74537
> scale fact cross-section :1 -1.43312
> scale fact cross-section :2 -1.19754
> scale fact cross-section :3 -1.57691
> scale fact cross-section :4 -1.29508
> scale fact cross-section :5 -1.08237
> scale fact cross-section :6 -1.43046
> scale fact cross-section :7 -1.17502
> scale fact cross-section :8 -0.982175
>
> The sum of weights in the event file is equal to the scale fact cross-
> section :4
>
> In this case there are both negative and positive event weights. Maybe
> the fix was not general enough?
>
> Thanks,
>
> Eleni
>
>> On 13 Apr 2016, at 02:44, Olivier Mattelaer <email address hidden> wrote:
>>
>> Thanks,
>>
>> I have fixed the problem in 2.4.1
>>
>> Cheers,
>>
>> Olivier
>>
>> ** Changed in: mg5amcnlo
>> Status: New => Fix Committed
>>
>> --
>> You received this bug notification because you are subscribed to the bug
>> report.
>> https://bugs.launchpad.net/bugs/1569506
>>
>> Title:
>> event weight not in agreement with xsection
>>
>> Status in MadGraph5_aMC@NLO:
>> Fix Committed
>>
>> Bug description:
>> Hi guys,
>>
>> I was running some simple loop process 2>1 and noticed that for one of
>> my parameter setups the event weight does not agree with the cross-
>> section given on the screen at the end of the run or the one in the
>> run_tag file.
>>
>> This is the end of the run_tag file:
>>
>> </slha>
>> <MGGenerationInfo>
>> # Number of Events : 5000
>> # Integrated weight (pb) : -2.3347
>> </MGGenerationInfo>
>> </header>
>> </LesHouchesEvents>
>>
>> and the event file I have:
>>
>> <MGGenerationInfo>
>> # Number of Events : 5000
>> # Integrated weight (pb) : -2.3347
>> </MGGenerationInfo>
>> </header>
>> <init>
>> 2212 2212 6.500000e+03 6.500000e+03 0 0 25000 25000 -3 1
>> -2.334700e+00 9.036100e-03 4.669400e-04 0
>> <generator name='MadGraph5_aMC@NLO' version='5.2.4.0.beta'>please cite 1405.0301 </generator>
>> </init>
>> <event>
>> 3 0 -4.1417436e-04 1.25000000e+02 7.81860800e-03 1.28338700e-01
>> 21 -1 0 0 502 501 +0.0000000000e+00 +0.00000...

Read more...

Revision history for this message
Eleni Vryonidou (evryonidou) wrote :
Download full text (9.9 KiB)

Hi Olivier,

thanks for the clarification.

Eleni

> On 13 Apr 2016, at 18:24, Olivier Mattelaer <email address hidden> wrote:
>
> Hi Eleni,
>
> This is just expected. The quantity which is preserved by the unweighting is the \sum abs(wgt)
> and therefore the \sum wgt will not be exactly equal to the cross-section after unweighting.
> The same actually occur in all NLO sample where the sum of the weight is not equal to the cross-sections.
> (The difference is by construction within the statistical error, but can be higher than the precision on the cross-section since it depends on the number of event requested)
>
> Cheers,
>
> Olivier
>
>
> In presence of negative and positive wei
>
> Cheers,
>
> Olivier
>
>> On Apr 13, 2016, at 17:06, Eleni Vryonidou <email address hidden> wrote:
>>
>> Hi Olivier,
>>
>> I seem to find the same problem again with another process in 2.4.1
>>
>> I noticed it by running SysCalc which gives slightly different central
>> scale results compared to the output on the screen. Again the sum of the
>> weights is not equal to the cross-section.
>>
>> <MGGenerationInfo>
>> # Number of Events : 200000
>> # Integrated weight (pb) : -1.272194
>> </MGGenerationInfo>
>>
>> scale fact cross-section :0 -1.74537
>> scale fact cross-section :1 -1.43312
>> scale fact cross-section :2 -1.19754
>> scale fact cross-section :3 -1.57691
>> scale fact cross-section :4 -1.29508
>> scale fact cross-section :5 -1.08237
>> scale fact cross-section :6 -1.43046
>> scale fact cross-section :7 -1.17502
>> scale fact cross-section :8 -0.982175
>>
>> The sum of weights in the event file is equal to the scale fact cross-
>> section :4
>>
>> In this case there are both negative and positive event weights. Maybe
>> the fix was not general enough?
>>
>> Thanks,
>>
>> Eleni
>>
>>> On 13 Apr 2016, at 02:44, Olivier Mattelaer <email address hidden> wrote:
>>>
>>> Thanks,
>>>
>>> I have fixed the problem in 2.4.1
>>>
>>> Cheers,
>>>
>>> Olivier
>>>
>>> ** Changed in: mg5amcnlo
>>> Status: New => Fix Committed
>>>
>>> --
>>> You received this bug notification because you are subscribed to the bug
>>> report.
>>> https://bugs.launchpad.net/bugs/1569506
>>>
>>> Title:
>>> event weight not in agreement with xsection
>>>
>>> Status in MadGraph5_aMC@NLO:
>>> Fix Committed
>>>
>>> Bug description:
>>> Hi guys,
>>>
>>> I was running some simple loop process 2>1 and noticed that for one of
>>> my parameter setups the event weight does not agree with the cross-
>>> section given on the screen at the end of the run or the one in the
>>> run_tag file.
>>>
>>> This is the end of the run_tag file:
>>>
>>> </slha>
>>> <MGGenerationInfo>
>>> # Number of Events : 5000
>>> # Integrated weight (pb) : -2.3347
>>> </MGGenerationInfo>
>>> </header>
>>> </LesHouchesEvents>
>>>
>>> and the event file I have:
>>>
>>> <MGGenerationInfo>
>>> # Number of Events : 5000
>>> # Integrated weight (pb) : -2.3347
>>> </MGGenerationInfo>
>>> </header>
>>> <init>
>>> 2212 2212 6.500000e+03 6.500000e+03 0 0 25000 25000 -3 1
>>> -2.334700e+00 9.036100e-03 4.66940...

Olivier Mattelaer (olivier-mattelaer)
Changed in mg5amcnlo:
status: Fix Committed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.
Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.