little bit about mass distributions phenomenology

Asked by Safinaz

Dear authors,

I'm working on studying pp > S0 S0 > bb~bb~ process where S0 is a neutral octet scalar of mS = 2 TeV ..
generated two plots pf m(b,b~) versus the cross section(pb/bin) and versus number of events ..

I noticed that the plot of cross section(pb/bin) is a distribtuion starting from around 750 GeV, while
the plot of number of events is adistribution has a small peak at 2 TeV .. which is fine .. so why this peak dosn't
appear at the cross plot as on-shell decay of S0 .. like for example pp> tt~H, H>bb~ process in SM .. is that because I have in my process 4 b .. if so why the events plot differes from the sigma plot ?

Bests,
Safinaz

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 Safinaz,

What is your input in order to make your plots?
If you use unweighted events, then I don't see how you can have two different plots.
(since the cross-section is directly proportional to the number of events)
If you use a weighted sample then of course this is possible.

Is this before/after shower/hadronization/detector simulation?
Did you use the same selection routine for the two plots?

Cheers,

Olivier

On Oct 18, 2013, at 4:21 PM, Safinaz <email address hidden> wrote:

> New question #237560 on MadGraph5:
> https://answers.launchpad.net/madgraph5/+question/237560
>
>
> Dear authors,
>
> I'm working on studying pp > S0 S0 > bb~bb~ process where S0 is a neutral octet scalar of mS = 2 TeV ..
> generated two plots pf m(b,b~) versus the cross section(pb/bin) and versus number of events ..
>
> I noticed that the plot of cross section(pb/bin) is a distribtuion starting from around 750 GeV, while
> the plot of number of events is adistribution has a small peak at 2 TeV .. which is fine .. so why this peak dosn't
> appear at the cross plot as on-shell decay of S0 .. like for example pp> tt~H, H>bb~ process in SM .. is that because I have in my process 4 b .. if so why the events plot differes from the sigma plot ?
>
> Bests,
> Safinaz
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.

Revision history for this message
Safinaz (sramadan) said :
#2

Hi Dr. Olivier,

I drew the number of events at parton level by MadAnalysis by making "import<path> events.lhe as bb~bb~ "

While the cross section by Madgraph by " generate p p > S0 S0 , S0 > b b~ " .

Bests,
Safinaz

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

Hi Safinaz,

Since you are using the events.lhe, you are using a weighted sample, therefore the number of events in that file doesn't have any physical meaning.
Since some events are more statistically significant than some other (due their weight)
If you import unweighted_events.lhe, then you should reproduce the plot of the cross-section since they all have the same weight.

Cheers,

Olivier

PS: I'm surprised that you don't observe a peak at 2 TeV, what is the value of the width for this particles at such spectrum?

On Oct 18, 2013, at 5:16 PM, Safinaz <email address hidden> wrote:

> Question #237560 on MadGraph5 changed:
> https://answers.launchpad.net/madgraph5/+question/237560
>
> Safinaz posted a new comment:
>
> Hi Dr. Olivier,
>
> I drew the number of events at parton level by MadAnalysis by making
> "import<path> events.lhe as bb~bb~ "
>
> While the cross section by Madgraph by " generate p p > S0 S0 , S0 > b
> b~ " .
>
> Bests,
> Safinaz
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.

Revision history for this message
Safinaz (sramadan) said :
#4

Hi Dr. Olivier,

It took me some time to check my model ..

* The total width of S0 scalar is 0.5 GeV (ms0 =2 Tev) and they are coloed octet

* Still I have no the resonance peak for " p p > S0 S0 , S0 > b b~" cross section .. however I noticed that even
 for " p p > S0 S0 " process Madgraph dosn't plot the invariant mass of M(S0,S0) vs. sigma ..alternative for a SM process
like " p p > t t~ " and I don't why ..

* Even when I used the generated unweighted_events.lhe file for MadAnalysis (parton level) and said:
plot M(b b~ ) 30 0 8000 [logY] .. I got the events plot as a distribution has a peak centrerd at 2 TeV .. ?

Regards,
Safinaz

Revision history for this message
Olivier Mattelaer (olivier-mattelaer) said :
#5

Dear Safinaz,

> * Even when I used the generated unweighted_events.lhe file for MadAnalysis (parton level) and said:
> plot M(b b~ ) 30 0 8000 [logY] .. I got the events plot as a distribution has a peak centered at 2 TeV .. ?

And? You expect to have that peak, right?
In your email, I thought that you were saying that you didn't have the peak… Sorry I'm confuse.

> * Still I have no the resonance peak for " p p > S0 S0 , S0 > b b~" cross section .. however I noticed that even
> for " p p > S0 S0 " process Madgraph dosn't plot the invariant mass of M(S0,S0) vs. sigma ..alternative for a SM process
> like " p p > t t~ " and I don't why ..

In madgraph, we use by default MadAnalysis4 for the plot which is a bit crappy and not "smart" to discover new particles in the generation.

Cheers,

Olivier
On Oct 21, 2013, at 8:01 PM, Safinaz <email address hidden> wrote:

> Question #237560 on MadGraph5 changed:
> https://answers.launchpad.net/madgraph5/+question/237560
>
> Safinaz posted a new comment:
>
> Hi Dr. Olivier,
>
> It took me some time to check my model ..
>
> * The total width of S0 scalar is 0.5 GeV (ms0 =2 Tev) and they are
> coloed octet
>
> * Still I have no the resonance peak for " p p > S0 S0 , S0 > b b~" cross section .. however I noticed that even
> for " p p > S0 S0 " process Madgraph dosn't plot the invariant mass of M(S0,S0) vs. sigma ..alternative for a SM process
> like " p p > t t~ " and I don't why ..
>
> * Even when I used the generated unweighted_events.lhe file for MadAnalysis (parton level) and said:
> plot M(b b~ ) 30 0 8000 [logY] .. I got the events plot as a distribution has a peak centrerd at 2 TeV .. ?
>
> Regards,
> Safinaz
>
> --
> You received this question notification because you are a member of
> MadTeam, which is an answer contact for MadGraph5.

Revision history for this message
Safinaz (sramadan) said :
#6

Hi Dr. Olivier,

* In madgraph, we use by default MadAnalysis4 for the plot which is a bit crappy and not "smart" to discover new particles in the generation.

I think here you mean the MadAnalysis version that installed in Madgraph by : mg5>install MadAnalysis

So that's fine I can admit now the events plots made from the unweighted_events.lhe or even
 unweighted_events.root by ExRootAnalysis package (since the last one has also resonsce peak for M parameter)
and which could identify new physics ..
Espaically as you mentioned in the email , I got the plot of the weight of the events
and is a bin at one ..

Regards,
Safinaz

Revision history for this message
Best Olivier Mattelaer (olivier-mattelaer) said :
#7

On Oct 22, 2013, at 3:46 PM, Safinaz <email address hidden> wrote:

> Question #237560 on MadGraph5 changed:
> https://answers.launchpad.net/madgraph5/+question/237560
>
> Safinaz posted a new comment:
>
> Hi Dr. Olivier,
>
> * In madgraph, we use by default MadAnalysis4 for the plot which is a
> bit crappy and not "smart" to discover new particles in the generation.
>
> I think here you mean the MadAnalysis version that installed in Madgraph
> by : mg5>install MadAnalysis

yes

Olivier

Revision history for this message
Safinaz (sramadan) said :
#8

Thanks Olivier Mattelaer, that solved my question.