post-ap 0.0.4

Project with code used to ease data manipulation tasks

Description

This project is used to carry out checks on Run3 data. Check this for installation instructions. Regardless, the code will need the following variable to be defined:

export LXNAME=$USER # This is the username when running in LXPLUS
# This is the value of VENVS used to create the virtual environment that will be used
export VENVS=/afs/ihep.ac.cn/users/c/campoverde/VENVS

such that the code that is ran, will be taken from a tarball in the grid, and it will be associated to a specific user.

and instead of running in a virtual environment one will have to run in an environment with DIRAC with:

. /cvmfs/lhcb.cern.ch/lib/LbEnv

# This will open a new shell
lb-dirac

# And you will work here
export PATH+=:$HOME/.local/bin

Specifying configuration for filtering and slimming

For this to work, configs need to be uploaded to the grid with the scripts below. The scripts need to know the place in the grid where the user LFNs live. For that, the following line needs to be issued:

export LXNAME=$USER # This is the username when running in LXPLUS

The configuration file is updated with:

update_config -u 1

The -u flag will update the config file if its LFN is already in the grid. The script runs with:

1. The LHCb environment set up.
2. With a valid grid token.
3. Within the working virtual environment. lb-dirac and the script need to be used. No conflict between the VENV and the LHCb environments seems to happen.

Save lists of PFNs

The PFNs to be processed will be stored once with the AP api and will be read as package data when processing ntuples. The list of PFNs is created with, e.g.:

save_pfns -c dt_2024_turbo_comp

where -c will correspond to the config file.

Submitting jobs

---

All the jobs below require code that lives in a virtual environment, there should be multiple versions of this environment and the latest one should be obtained by running:

lb-dirac dirac-dms-user-lfns -w dcheck.tar -b /lhcb/user/${LXNAME:0:1}/$LXNAME/run3/venv

---

The instructions below need to be done outside the virtual environment in an environment with access to dirac and in the post_ap_grid directory.

First run a test job with:

./job_filter -d dt_2024_turbo -c comp -j 1211 -e 003 -m local -n test_flt

where -j specifies the number of jobs. For tests, this is the number of files to process, thus, the test job does only one file. The -n flag is the name of the job, for tests it will do/send only one job if either:

1. Its name has the substring test.
2. It is a local job.

Thus one can do local or grid tests running over a single file.

For real jobs:

./job_filter -d dt_2024_turbo -c comp -j 200 -e 003 -m wms -n flt_001

Downloading ntuples

A test would look like:

run3_download_ntuples -j flt_004 -n 3 [-d $PWD/files]

where:

-j: Is the name of the job, which has to coincide with the directory name, where the ntuples are in EOS, e.g. /eos/lhcb/grid/user/lhcb/user/a/acampove/flt_004.
-n: Number of ntuples to download, if not pased, will download everything.
-d: Directory where output ntuples will go, if not passed, directory pointed by DOWNLOAD_NTUPPATH will be used.

A real download would look like:

run3_download_ntuples -j flt_001 -m 40

Where -m denotes the number of threads used to download, -j the name of the job.

Notes

• The downloads can be ran many times, if a file has been downloaded already, it will not be downloaded again.

Linking and merging

Once the ntuples are downloaded these need to be linked and merged with:

link_merge -j flt_002 -v v1

where -j is the name of the job and the files are linked to a directory named as -v v1. For tests run:

link_merge -j flt_002 -d 1 -m 10 -v v1

which will do the same with at most 10 files, can use debug messages with -l 10.

Making basic plots

For this run:

plot_vars -y 2024 -v v2 -c bukee_opt -d data_ana_cut_bp_ee:Data ctrl_BuToKpEE_ana_ee:Simulation

which will run the plotting of the variables in a config specified by bukee_opt where also axis, names, ranges, etc are specified. This config is in post_ap_data. The script above will overlay data and MC.