banner.txt lacks "Matched Cross-section"

Asked by Sho Iwamoto

Hi MadTeam,

For a process with `ICKKW=1` matching, the banner file loses several information when we do `launch` the process again with a same runname.
This is reproduced in v2.4.0 (May 12 2016), with pythia-pgs + delphes3 (latest).

For example, with ./bin/mg5_aMC, consider the workflow
```
generate p p > t t
add process p p > t t j
output tt
launch (with laststep=delphes, run_card automatically has ickkw=1)
launch --name=run_01
```

In this case, after the first `launch`, `tt/Events/run_01/run_01_tag_1_banner.txt` is generated.
This file at this point has all information:
```
$ tail run_01_tag_1_banner.txt

  add Branch ScalarHT/energy ScalarHT ScalarHT
}
</MGDelphesCard>
<MGGenerationInfo>
# Number of Events : 10000
# Integrated weight (pb) : 1058.914901
# Matched Integrated weight (pb) : 541.1
</MGGenerationInfo>
</header>
</LesHouchesEvents>
```

However, this file is modified during the second launch, i.e., `launch --name=run_01`. After the launch the tail is
```
DECAY 16 0.000000 # vt : 0.0
DECAY 21 0.000000 # g : 0.0
DECAY 22 0.000000 # a : 0.0
</slha>
<MGGenerationInfo>
# Number of Events : 10000
# Integrated weight (pb) : 1060.552755
</MGGenerationInfo>
</header>
</LesHouchesEvents>
```
which has lost Matched Cross section as well as delphes card etc.

(Note that the banner for the second launch is generated as `run_01_tag_2_banner.txt`.)

Also, if we run more, like
```
generate p p > t t
add process p p > t t j
output tt
launch
launch --name=run_01
launch --name=run_01
...
```
only the first banner `run_01_tag_1_banner.txt` loses the Matched cross section etc., and tag_2, tag_3, ... have complete information.

This behavior is probably due to `add_generation_info` in `bin/internal/banner.py`.

So, I guess this can be a bug, but rather I have the following question:

 * Is this workflow (multiple launch with same runname) correct/expected workflow?
 * What are the concepts `run` and `tag` ?

(I am a bit sorry that I may have read this question here before but I could not find it)

Question information

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

Dear Sho,

The run_name correspond to the directory name under the Events directory.
In such directory you can have only one events file (unweighted_events.lhe)
but we allow to run multiple times the shower/… and those file have therefore a “tag” associated.

Each couple (run_name,tag) are associated to their own banner.

When you do
launch —name=run_01
you overwrite the previous event file existing in run_01 by the new one.
This is clearly not a standard workflow.

A standard work flow would be

launch —name=name1
launch —name=name2

Cheers,

Olivier

> On May 21, 2016, at 14:07, Sho Iwamoto <email address hidden> wrote:
>
> New question #294015 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/294015
>
> Hi MadTeam,
>
> For a process with `ICKKW=1` matching, the banner file loses several information when we do `launch` the process again with a same runname.
> This is reproduced in v2.4.0 (May 12 2016), with pythia-pgs + delphes3 (latest).
>
>
> For example, with ./bin/mg5_aMC, consider the workflow
> ```
> generate p p > t t
> add process p p > t t j
> output tt
> launch (with laststep=delphes, run_card automatically has ickkw=1)
> launch --name=run_01
> ```
>
> In this case, after the first `launch`, `tt/Events/run_01/run_01_tag_1_banner.txt` is generated.
> This file at this point has all information:
> ```
> $ tail run_01_tag_1_banner.txt
>
> add Branch ScalarHT/energy ScalarHT ScalarHT
> }
> </MGDelphesCard>
> <MGGenerationInfo>
> # Number of Events : 10000
> # Integrated weight (pb) : 1058.914901
> # Matched Integrated weight (pb) : 541.1
> </MGGenerationInfo>
> </header>
> </LesHouchesEvents>
> ```
>
> However, this file is modified during the second launch, i.e., `launch --name=run_01`. After the launch the tail is
> ```
> DECAY 16 0.000000 # vt : 0.0
> DECAY 21 0.000000 # g : 0.0
> DECAY 22 0.000000 # a : 0.0
> </slha>
> <MGGenerationInfo>
> # Number of Events : 10000
> # Integrated weight (pb) : 1060.552755
> </MGGenerationInfo>
> </header>
> </LesHouchesEvents>
> ```
> which has lost Matched Cross section as well as delphes card etc.
>
> (Note that the banner for the second launch is generated as `run_01_tag_2_banner.txt`.)
>
>
> Also, if we run more, like
> ```
> generate p p > t t
> add process p p > t t j
> output tt
> launch
> launch --name=run_01
> launch --name=run_01
> ...
> ```
> only the first banner `run_01_tag_1_banner.txt` loses the Matched cross section etc., and tag_2, tag_3, ... have complete information.
>
>
> This behavior is probably due to `add_generation_info` in `bin/internal/banner.py`.
>
> So, I guess this can be a bug, but rather I have the following question:
>
> * Is this workflow (multiple launch with same runname) correct/expected workflow?
> * What are the concepts `run` and `tag` ?
>
> (I am a bit sorry that I may have read this question here before but I could not find it)
>
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Can you help with this problem?

Provide an answer of your own, or ask Sho Iwamoto for more information if necessary.

To post a message you must log in.