MadGraph5_aMC@NLO

Re-calculation of the systematic uncertainties after event selection

Asked by Yannis Maznas

Dear MG experts,

As the title implies what I'd like to do is to recalculate the systematic uncertainties for scale and PDF variations after the event selection.

So after running the embedded python tool for systematics, the original lhe file is enriched with a bunch of reweighted weights based the following indexing table as quoted in the enriched lhe file.

<weightgroup name="Central scale variation" combine="envelope">
<weight id="1" MUR="0.5" MUF="0.5" PDF="262000" > MUR=0.5 MUF=0.5 </weight>
<weight id="2" MUR="0.5" MUF="0.5" DYN_SCALE="1" PDF="262000" > MUR=0.5 MUF=0.5 dyn_scale_choice=sum pt </weight>
<weight id="3" MUR="0.5" MUF="0.5" DYN_SCALE="2" PDF="262000" > MUR=0.5 MUF=0.5 dyn_scale_choice=HT </weight>
<weight id="4" MUR="0.5" MUF="0.5" DYN_SCALE="3" PDF="262000" > MUR=0.5 MUF=0.5 dyn_scale_choice=HT/2 </weight>
<weight id="5" MUR="0.5" MUF="0.5" DYN_SCALE="4" PDF="262000" > MUR=0.5 MUF=0.5 dyn_scale_choice=sqrts </weight>
<weight id="6" MUR="0.5" MUF="1.0" PDF="262000" > MUR=0.5 </weight>
<weight id="7" MUR="0.5" MUF="1.0" DYN_SCALE="1" PDF="262000" > MUR=0.5 dyn_scale_choice=sum pt </weight>
<weight id="8" MUR="0.5" MUF="1.0" DYN_SCALE="2" PDF="262000" > MUR=0.5 dyn_scale_choice=HT </weight>
<weight id="9" MUR="0.5" MUF="1.0" DYN_SCALE="3" PDF="262000" > MUR=0.5 dyn_scale_choice=HT/2 </weight>
<weight id="10" MUR="0.5" MUF="1.0" DYN_SCALE="4" PDF="262000" > MUR=0.5 dyn_scale_choice=sqrts </weight>
<weight id="11" MUR="0.5" MUF="2.0" PDF="262000" > MUR=0.5 MUF=2.0 </weight>
<weight id="12" MUR="0.5" MUF="2.0" DYN_SCALE="1" PDF="262000" > MUR=0.5 MUF=2.0 dyn_scale_choice=sum pt </weight>
<weight id="13" MUR="0.5" MUF="2.0" DYN_SCALE="2" PDF="262000" > MUR=0.5 MUF=2.0 dyn_scale_choice=HT </weight>
<weight id="14" MUR="0.5" MUF="2.0" DYN_SCALE="3" PDF="262000" > MUR=0.5 MUF=2.0 dyn_scale_choice=HT/2 </weight>
<weight id="15" MUR="0.5" MUF="2.0" DYN_SCALE="4" PDF="262000" > MUR=0.5 MUF=2.0 dyn_scale_choice=sqrts </weight>
<weight id="16" MUR="1.0" MUF="0.5" PDF="262000" > MUF=0.5 </weight>
<weight id="17" MUR="1.0" MUF="0.5" DYN_SCALE="1" PDF="262000" > MUF=0.5 dyn_scale_choice=sum pt </weight>
<weight id="18" MUR="1.0" MUF="0.5" DYN_SCALE="2" PDF="262000" > MUF=0.5 dyn_scale_choice=HT </weight>
<weight id="19" MUR="1.0" MUF="0.5" DYN_SCALE="3" PDF="262000" > MUF=0.5 dyn_scale_choice=HT/2 </weight>
<weight id="20" MUR="1.0" MUF="0.5" DYN_SCALE="4" PDF="262000" > MUF=0.5 dyn_scale_choice=sqrts </weight>
<weight id="21" MUR="1.0" MUF="1.0" DYN_SCALE="1" PDF="262000" > dyn_scale_choice=sum pt </weight>
<weight id="22" MUR="1.0" MUF="1.0" DYN_SCALE="2" PDF="262000" > dyn_scale_choice=HT </weight>
<weight id="23" MUR="1.0" MUF="1.0" DYN_SCALE="3" PDF="262000" > dyn_scale_choice=HT/2 </weight>
<weight id="24" MUR="1.0" MUF="1.0" DYN_SCALE="4" PDF="262000" > dyn_scale_choice=sqrts </weight>
<weight id="25" MUR="1.0" MUF="2.0" PDF="262000" > MUF=2.0 </weight>
<weight id="26" MUR="1.0" MUF="2.0" DYN_SCALE="1" PDF="262000" > MUF=2.0 dyn_scale_choice=sum pt </weight>
<weight id="27" MUR="1.0" MUF="2.0" DYN_SCALE="2" PDF="262000" > MUF=2.0 dyn_scale_choice=HT </weight>
<weight id="28" MUR="1.0" MUF="2.0" DYN_SCALE="3" PDF="262000" > MUF=2.0 dyn_scale_choice=HT/2 </weight>
<weight id="29" MUR="1.0" MUF="2.0" DYN_SCALE="4" PDF="262000" > MUF=2.0 dyn_scale_choice=sqrts </weight>
<weight id="30" MUR="2.0" MUF="0.5" PDF="262000" > MUR=2.0 MUF=0.5 </weight>
<weight id="31" MUR="2.0" MUF="0.5" DYN_SCALE="1" PDF="262000" > MUR=2.0 MUF=0.5 dyn_scale_choice=sum pt </weight>
<weight id="32" MUR="2.0" MUF="0.5" DYN_SCALE="2" PDF="262000" > MUR=2.0 MUF=0.5 dyn_scale_choice=HT </weight>
<weight id="33" MUR="2.0" MUF="0.5" DYN_SCALE="3" PDF="262000" > MUR=2.0 MUF=0.5 dyn_scale_choice=HT/2 </weight>
<weight id="34" MUR="2.0" MUF="0.5" DYN_SCALE="4" PDF="262000" > MUR=2.0 MUF=0.5 dyn_scale_choice=sqrts </weight>
<weight id="35" MUR="2.0" MUF="1.0" PDF="262000" > MUR=2.0 </weight>
<weight id="36" MUR="2.0" MUF="1.0" DYN_SCALE="1" PDF="262000" > MUR=2.0 dyn_scale_choice=sum pt </weight>
<weight id="37" MUR="2.0" MUF="1.0" DYN_SCALE="2" PDF="262000" > MUR=2.0 dyn_scale_choice=HT </weight>
<weight id="38" MUR="2.0" MUF="1.0" DYN_SCALE="3" PDF="262000" > MUR=2.0 dyn_scale_choice=HT/2 </weight>
<weight id="39" MUR="2.0" MUF="1.0" DYN_SCALE="4" PDF="262000" > MUR=2.0 dyn_scale_choice=sqrts </weight>
<weight id="40" MUR="2.0" MUF="2.0" PDF="262000" > MUR=2.0 MUF=2.0 </weight>
<weight id="41" MUR="2.0" MUF="2.0" DYN_SCALE="1" PDF="262000" > MUR=2.0 MUF=2.0 dyn_scale_choice=sum pt </weight>
<weight id="42" MUR="2.0" MUF="2.0" DYN_SCALE="2" PDF="262000" > MUR=2.0 MUF=2.0 dyn_scale_choice=HT </weight>
<weight id="43" MUR="2.0" MUF="2.0" DYN_SCALE="3" PDF="262000" > MUR=2.0 MUF=2.0 dyn_scale_choice=HT/2 </weight>
<weight id="44" MUR="2.0" MUF="2.0" DYN_SCALE="4" PDF="262000" > MUR=2.0 MUF=2.0 dyn_scale_choice=sqrts </weight>
</weightgroup> # scale
<weightgroup name="NNPDF30_lo_as_0118" combine="replicas"> # 262000: NNPDF3.0 LO global fit, alphas(MZ)=0.118. mem=0 ; average on replicas; mem=1-100 ; PDF replicas
<weight id="45" MUR="1.0" MUF="1.0" PDF="262000" > </weight>
<weight id="46" MUR="1.0" MUF="1.0" PDF="262001" > PDF=262000 MemberID=1 </weight>
<weight id="47" MUR="1.0" MUF="1.0" PDF="262002" > PDF=262000 MemberID=2 </weight>
<weight id="48" MUR="1.0" MUF="1.0" PDF="262003" > PDF=262000 MemberID=3 </weight>
<weight id="49" MUR="1.0" MUF="1.0" PDF="262004" > PDF=262000 MemberID=4 </weight>
<weight id="50" MUR="1.0" MUF="1.0" PDF="262005" > PDF=262000 MemberID=5 </weight>
<weight id="51" MUR="1.0" MUF="1.0" PDF="262006" > PDF=262000 MemberID=6 </weight>
<weight id="52" MUR="1.0" MUF="1.0" PDF="262007" > PDF=262000 MemberID=7 </weight>
<weight id="53" MUR="1.0" MUF="1.0" PDF="262008" > PDF=262000 MemberID=8 </weight>
<weight id="54" MUR="1.0" MUF="1.0" PDF="262009" > PDF=262000 MemberID=9 </weight>
<weight id="55" MUR="1.0" MUF="1.0" PDF="262010" > PDF=262000 MemberID=10 </weight>
<weight id="56" MUR="1.0" MUF="1.0" PDF="262011" > PDF=262000 MemberID=11 </weight>
<weight id="57" MUR="1.0" MUF="1.0" PDF="262012" > PDF=262000 MemberID=12 </weight>
<weight id="58" MUR="1.0" MUF="1.0" PDF="262013" > PDF=262000 MemberID=13 </weight>
<weight id="59" MUR="1.0" MUF="1.0" PDF="262014" > PDF=262000 MemberID=14 </weight>
<weight id="60" MUR="1.0" MUF="1.0" PDF="262015" > PDF=262000 MemberID=15 </weight>
<weight id="61" MUR="1.0" MUF="1.0" PDF="262016" > PDF=262000 MemberID=16 </weight>
<weight id="62" MUR="1.0" MUF="1.0" PDF="262017" > PDF=262000 MemberID=17 </weight>
<weight id="63" MUR="1.0" MUF="1.0" PDF="262018" > PDF=262000 MemberID=18 </weight>
<weight id="64" MUR="1.0" MUF="1.0" PDF="262019" > PDF=262000 MemberID=19 </weight>
<weight id="65" MUR="1.0" MUF="1.0" PDF="262020" > PDF=262000 MemberID=20 </weight>
<weight id="66" MUR="1.0" MUF="1.0" PDF="262021" > PDF=262000 MemberID=21 </weight>
<weight id="67" MUR="1.0" MUF="1.0" PDF="262022" > PDF=262000 MemberID=22 </weight>
<weight id="68" MUR="1.0" MUF="1.0" PDF="262023" > PDF=262000 MemberID=23 </weight>
<weight id="69" MUR="1.0" MUF="1.0" PDF="262024" > PDF=262000 MemberID=24 </weight>
<weight id="70" MUR="1.0" MUF="1.0" PDF="262025" > PDF=262000 MemberID=25 </weight>
<weight id="71" MUR="1.0" MUF="1.0" PDF="262026" > PDF=262000 MemberID=26 </weight>
<weight id="72" MUR="1.0" MUF="1.0" PDF="262027" > PDF=262000 MemberID=27 </weight>
<weight id="73" MUR="1.0" MUF="1.0" PDF="262028" > PDF=262000 MemberID=28 </weight>
<weight id="74" MUR="1.0" MUF="1.0" PDF="262029" > PDF=262000 MemberID=29 </weight>
<weight id="75" MUR="1.0" MUF="1.0" PDF="262030" > PDF=262000 MemberID=30 </weight>
<weight id="76" MUR="1.0" MUF="1.0" PDF="262031" > PDF=262000 MemberID=31 </weight>
<weight id="77" MUR="1.0" MUF="1.0" PDF="262032" > PDF=262000 MemberID=32 </weight>
<weight id="78" MUR="1.0" MUF="1.0" PDF="262033" > PDF=262000 MemberID=33 </weight>
<weight id="79" MUR="1.0" MUF="1.0" PDF="262034" > PDF=262000 MemberID=34 </weight>
<weight id="80" MUR="1.0" MUF="1.0" PDF="262035" > PDF=262000 MemberID=35 </weight>
<weight id="81" MUR="1.0" MUF="1.0" PDF="262036" > PDF=262000 MemberID=36 </weight>
<weight id="82" MUR="1.0" MUF="1.0" PDF="262037" > PDF=262000 MemberID=37 </weight>
<weight id="83" MUR="1.0" MUF="1.0" PDF="262038" > PDF=262000 MemberID=38 </weight>
<weight id="84" MUR="1.0" MUF="1.0" PDF="262039" > PDF=262000 MemberID=39 </weight>
<weight id="85" MUR="1.0" MUF="1.0" PDF="262040" > PDF=262000 MemberID=40 </weight>
<weight id="86" MUR="1.0" MUF="1.0" PDF="262041" > PDF=262000 MemberID=41 </weight>
<weight id="87" MUR="1.0" MUF="1.0" PDF="262042" > PDF=262000 MemberID=42 </weight>
<weight id="88" MUR="1.0" MUF="1.0" PDF="262043" > PDF=262000 MemberID=43 </weight>
<weight id="89" MUR="1.0" MUF="1.0" PDF="262044" > PDF=262000 MemberID=44 </weight>
<weight id="90" MUR="1.0" MUF="1.0" PDF="262045" > PDF=262000 MemberID=45 </weight>
<weight id="91" MUR="1.0" MUF="1.0" PDF="262046" > PDF=262000 MemberID=46 </weight>
<weight id="92" MUR="1.0" MUF="1.0" PDF="262047" > PDF=262000 MemberID=47 </weight>
<weight id="93" MUR="1.0" MUF="1.0" PDF="262048" > PDF=262000 MemberID=48 </weight>
<weight id="94" MUR="1.0" MUF="1.0" PDF="262049" > PDF=262000 MemberID=49 </weight>
<weight id="95" MUR="1.0" MUF="1.0" PDF="262050" > PDF=262000 MemberID=50 </weight>
<weight id="96" MUR="1.0" MUF="1.0" PDF="262051" > PDF=262000 MemberID=51 </weight>
<weight id="97" MUR="1.0" MUF="1.0" PDF="262052" > PDF=262000 MemberID=52 </weight>
<weight id="98" MUR="1.0" MUF="1.0" PDF="262053" > PDF=262000 MemberID=53 </weight>
<weight id="99" MUR="1.0" MUF="1.0" PDF="262054" > PDF=262000 MemberID=54 </weight>
<weight id="100" MUR="1.0" MUF="1.0" PDF="262055" > PDF=262000 MemberID=55 </weight>
<weight id="101" MUR="1.0" MUF="1.0" PDF="262056" > PDF=262000 MemberID=56 </weight>
<weight id="102" MUR="1.0" MUF="1.0" PDF="262057" > PDF=262000 MemberID=57 </weight>
<weight id="103" MUR="1.0" MUF="1.0" PDF="262058" > PDF=262000 MemberID=58 </weight>
<weight id="104" MUR="1.0" MUF="1.0" PDF="262059" > PDF=262000 MemberID=59 </weight>
<weight id="105" MUR="1.0" MUF="1.0" PDF="262060" > PDF=262000 MemberID=60 </weight>
<weight id="106" MUR="1.0" MUF="1.0" PDF="262061" > PDF=262000 MemberID=61 </weight>
<weight id="107" MUR="1.0" MUF="1.0" PDF="262062" > PDF=262000 MemberID=62 </weight>
<weight id="108" MUR="1.0" MUF="1.0" PDF="262063" > PDF=262000 MemberID=63 </weight>
<weight id="109" MUR="1.0" MUF="1.0" PDF="262064" > PDF=262000 MemberID=64 </weight>
<weight id="110" MUR="1.0" MUF="1.0" PDF="262065" > PDF=262000 MemberID=65 </weight>
<weight id="111" MUR="1.0" MUF="1.0" PDF="262066" > PDF=262000 MemberID=66 </weight>
<weight id="112" MUR="1.0" MUF="1.0" PDF="262067" > PDF=262000 MemberID=67 </weight>
<weight id="113" MUR="1.0" MUF="1.0" PDF="262068" > PDF=262000 MemberID=68 </weight>
<weight id="114" MUR="1.0" MUF="1.0" PDF="262069" > PDF=262000 MemberID=69 </weight>
<weight id="115" MUR="1.0" MUF="1.0" PDF="262070" > PDF=262000 MemberID=70 </weight>
<weight id="116" MUR="1.0" MUF="1.0" PDF="262071" > PDF=262000 MemberID=71 </weight>
<weight id="117" MUR="1.0" MUF="1.0" PDF="262072" > PDF=262000 MemberID=72 </weight>
<weight id="118" MUR="1.0" MUF="1.0" PDF="262073" > PDF=262000 MemberID=73 </weight>
<weight id="119" MUR="1.0" MUF="1.0" PDF="262074" > PDF=262000 MemberID=74 </weight>
<weight id="120" MUR="1.0" MUF="1.0" PDF="262075" > PDF=262000 MemberID=75 </weight>
<weight id="121" MUR="1.0" MUF="1.0" PDF="262076" > PDF=262000 MemberID=76 </weight>
<weight id="122" MUR="1.0" MUF="1.0" PDF="262077" > PDF=262000 MemberID=77 </weight>
<weight id="123" MUR="1.0" MUF="1.0" PDF="262078" > PDF=262000 MemberID=78 </weight>
<weight id="124" MUR="1.0" MUF="1.0" PDF="262079" > PDF=262000 MemberID=79 </weight>
<weight id="125" MUR="1.0" MUF="1.0" PDF="262080" > PDF=262000 MemberID=80 </weight>
<weight id="126" MUR="1.0" MUF="1.0" PDF="262081" > PDF=262000 MemberID=81 </weight>
<weight id="127" MUR="1.0" MUF="1.0" PDF="262082" > PDF=262000 MemberID=82 </weight>
<weight id="128" MUR="1.0" MUF="1.0" PDF="262083" > PDF=262000 MemberID=83 </weight>
<weight id="129" MUR="1.0" MUF="1.0" PDF="262084" > PDF=262000 MemberID=84 </weight>
<weight id="130" MUR="1.0" MUF="1.0" PDF="262085" > PDF=262000 MemberID=85 </weight>
<weight id="131" MUR="1.0" MUF="1.0" PDF="262086" > PDF=262000 MemberID=86 </weight>
<weight id="132" MUR="1.0" MUF="1.0" PDF="262087" > PDF=262000 MemberID=87 </weight>
<weight id="133" MUR="1.0" MUF="1.0" PDF="262088" > PDF=262000 MemberID=88 </weight>
<weight id="134" MUR="1.0" MUF="1.0" PDF="262089" > PDF=262000 MemberID=89 </weight>
<weight id="135" MUR="1.0" MUF="1.0" PDF="262090" > PDF=262000 MemberID=90 </weight>
<weight id="136" MUR="1.0" MUF="1.0" PDF="262091" > PDF=262000 MemberID=91 </weight>
<weight id="137" MUR="1.0" MUF="1.0" PDF="262092" > PDF=262000 MemberID=92 </weight>
<weight id="138" MUR="1.0" MUF="1.0" PDF="262093" > PDF=262000 MemberID=93 </weight>
<weight id="139" MUR="1.0" MUF="1.0" PDF="262094" > PDF=262000 MemberID=94 </weight>
<weight id="140" MUR="1.0" MUF="1.0" PDF="262095" > PDF=262000 MemberID=95 </weight>
<weight id="141" MUR="1.0" MUF="1.0" PDF="262096" > PDF=262000 MemberID=96 </weight>
<weight id="142" MUR="1.0" MUF="1.0" PDF="262097" > PDF=262000 MemberID=97 </weight>
<weight id="143" MUR="1.0" MUF="1.0" PDF="262098" > PDF=262000 MemberID=98 </weight>
<weight id="144" MUR="1.0" MUF="1.0" PDF="262099" > PDF=262000 MemberID=99 </weight>
<weight id="145" MUR="1.0" MUF="1.0" PDF="262100" > PDF=262000 MemberID=100 </weight>
</weightgroup>
</initrwgt>

My first guess was that the systematics tool, based on these weights, calculates the systematic uncertainties as given on the terminal output that ends like this:

INFO: #***************************************************************************
#
# original cross-section: 3.32022e-07
# scale variation: +18.9% -15.1%
# central scheme variation: +37% -5.41e-10%
# PDF variation: +7.96% -7.96%
#
# dynamical scheme # 1 : 3.7871e-07 +20.1% -15.8% # \sum ET
# dynamical scheme # 2 : 3.7871e-07 +20.1% -15.8% # \sum\sqrt{m^2+pt^2}
# dynamical scheme # 3 : 4.54874e-07 +21.7% -16.7% # 0.5 \sum\sqrt{m^2+pt^2}
# dynamical scheme # 4 : 3.32022e-07 +18.9% -15.1% # \sqrt{\hat s}

First of all I'd like to ask if these variation percentages correspond to the standard error or to a confidence interval. If the second, at which CL?

Secondly, the fact that the only seemingly symmetrical variation shown above is the PDF variation, drives me to think that, if not all of them, most of them are based directly on error propagation calculations (as derived from https://arxiv.org/abs/1110.4738) rather than the event (re-weighted) weights themselves. Is that the case?

For the PDF variation, I tried calculating the standard error of the sum of re-weighted weights but the percentage, although close enough to the one printed out, is not identical. e.g. I got +-7.89%.
Also, out of curiosity, I plotted quickly the sum of event weights (for 82500 events) for each PDF variation and I got the following histogram showing a distribution of the 101 PDF variations that declines a bit from a Gaussian. https://prnt.sc/p7m3v3
Can you confirm if that is expected?

Finally, my main question is, how can I re-calculate these uncertainties based on the new weights after my event selection?
I would like to avoid the hardcore way of removing the events from the lhe file and re-feeding it to the systematics tool.

Thanks for bearing with me,
Yannis

Question information

Language:
English Edit question
Status:
Solved
For:
MadGraph5_aMC@NLO Edit question
Assignee:
No assignee Edit question
Solved by:
Olivier Mattelaer
Solved:
Last query:
Last reply:
Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said : 		#1

Hi,

>First of all I'd like to ask if these variation percentages correspond to the standard error or to a confidence interval. If the second, at which CL?

For the PDF uncertainty, you should actually refer to the associate PDF paper/reference (or sometimes simply name of the pdf).
It is in general a Confindence interval but the CL depend of the set (we even have two set with the same central value but with different CL associated to the error)

For the muR/muF factorization this is a pure rule of thumb to estimate uncertainty. In term of baysian statistic, the CL that you associated to that number is a pure prior of your statistical model.

>Secondly, the fact that the only seemingly symmetrical variation shown above is the PDF variation, drives me to think that, if not all of them, most of them are based directly on error propagation calculations (as derived from https://arxiv.org/abs/1110.4738) rather than the event (re-weighted) weights themselves. Is that the case?

I'm confused, that paper tells how to compute the weights and this is what the python follows as prescription. The method to combine those weights follows the PDF description (which is not universal and should follow the prescription of the pdf set).
So I do not understand your question here.

>Can you confirm if that is expected?
I do not see anything wrong

>Finally, my main question is, how can I re-calculate these uncertainties based on the new weights after my event selection?
I would like to avoid the hardcore way of removing the events from the lhe file and re-feeding it to the systematics tool.

What we do is to use lhapdf library to compute such numbers via the routine (Via python here):
pdferr = pdfset.uncertainty(values)

Cheers,

Olivier

Revision history for this message
Yannis Maznas (imaznas) said : 		#2

Hi Olivier,

Thanks for your quick reply!

So let me get this straight.
Scale uncertainties depend on the cross section calculated by MG (and by extension on the sum of weights) and as the code implies in the systematics.py on the maximum and minimum values of it (max_scale, min_scale variables), right?

On the other hand the pdf variation calculations solely depend on the pdf selected in the run card and has nothing to do with the events themselves. i.e. I will get the same variation even if I discard events from a dataset based on kinematics criteria.

I'm not familiar with your code but my understanding is that the "values" parameter corresponds to a lhapdf set, correct?

Best regards,
Yannis

Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said : 		#3

HI,

> Scale uncertainties depend on the cross section calculated by MG (and by extension on the sum of weights) and as the code implies in the systematics.py on the maximum and minimum values of it (max_scale, min_scale variables), right?

correct

> On the other hand the pdf variation calculations solely depend on the
> pdf selected in the run card and has nothing to do with the events
> themselves.

This is not True. For each PDF replica you can calculate the cross-section as the sum of the weights (or the average depending of the normalisation choice).

Then in order to get the uncertainty band associated to the cross-section given by that set of replica, you need to follow the method of computation of the error of that given PDF.

> i.e. I will get the same variation even if I discard events
> from a dataset based on kinematics criteria.

No you will get a different number (how different might depend of your sample and of your cut obviously)

> I'm not familiar with your code but my understanding is that the
> "values" parameter corresponds to a lhapdf set, correct?

This is the list of cross-section calculated as the sum (or average) of the weights.
You should look at the lhapdf manual if you want to use that method.

Cheers,

Olivier

> On 19 Sep 2019, at 13:47, Yannis Maznas <email address hidden> wrote:
>
> Question #684007 on MadGraph5_aMC@NLO changed:
> https://answers.launchpad.net/mg5amcnlo/+question/684007
>
> Status: Answered => Open
>
> Yannis Maznas is still having a problem:
> Hi Olivier,
>
> Thanks for your quick reply!
>
> So let me get this straight.
> Scale uncertainties depend on the cross section calculated by MG (and by extension on the sum of weights) and as the code implies in the systematics.py on the maximum and minimum values of it (max_scale, min_scale variables), right?
>
> On the other hand the pdf variation calculations solely depend on the
> pdf selected in the run card and has nothing to do with the events
> themselves. i.e. I will get the same variation even if I discard events
> from a dataset based on kinematics criteria.
>
> I'm not familiar with your code but my understanding is that the
> "values" parameter corresponds to a lhapdf set, correct?
>
> Best regards,
> Yannis
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Revision history for this message
Yannis Maznas (imaznas) said : 		#4

Thanks Olivier Mattelaer, that solved my question.