Size: 1444
Comment:
|
Size: 3285
Comment:
|
Deletions are marked like this. | Additions are marked like this. |
Line 9: | Line 9: |
* To read KAGRA data by using NDS, We need to edit the code (see [[http://gwwiki.icrr.u-tokyo.ac.jp/JGWwiki/KAGRA/Subgroups/DET/OmegaScan|instruction]].) | |
Line 12: | Line 12: |
conda activate o4_dqr_proto -y conda install M2Crypto -y |
conda activate o4_dqr_proto conda install m2crypto ciecplib python-nds2-client cryptography==3.3.1 -y |
Line 17: | Line 17: |
conda install ciecplib -y | |
Line 20: | Line 19: |
* We need to prepare the authentification to access GraceDB. * For that additional permission, we need to ask LIGO+Virgo colleague. * Now we are asking the KAGRA DetChar common account on gw-astronomy. We will ask to add the permission to GraceDB on that common account. |
* install ligo-proxy-init, from [[https://computing.docs.ligo.org/guide/software/ldg-client/|the LDG Client package]] |
Line 24: | Line 21: |
* After the account has the permission to access GraceDB, we need to prepare X.509 certification in local environment. * [[https://computing.docs.ligo.org/guide/authentication/#kagra|There are two options for KAGRA account]] * ecp-get-cert, from the ciecplib Python package |
{{{ wget http://www.lsc-group.phys.uwm.edu/lscdatagrid/doc/ldg-client.sh -O /tmp/ldg-client.sh && sudo bash /tmp/ldg-client.sh }}} |
Line 28: | Line 25: |
{{{ ecp-get-cert --identity-provider KAGRA --username albert.einstein --verbose }}} * ligo-proxy-init, from the LDG Client package |
* X.509 certification * For the additional permission to access GraceDB, we need to ask LIGO+Virgo colleague. * There are [[https://computing.docs.ligo.org/guide/authentication/#kagra|two options]] for KAGRA account. Here Yuzurihara used ligo-proxy-init. |
Line 37: | Line 32: |
* To test the certification, {{{ gracedb ping }}} {{{ grid-proxy-info }}} == Generate DAG files == * By following [[https://detchar.docs.ligo.org/dqrtasks/example-dqr.html#writing-configuration-files|instruction]], prepare ini files for DQR task manager. * This procedure will NOT require the access GraceDB, because the public alert will be used. * For test code, see {{{/users/yuzu/work/DQR/Script/o4-dqr-configuration:k1det1}}} {{{ mkdir ~/public_html/event dqr-create-dag -g S190924h -v -c ex_configs/defaults.ini ex_configs/condor.ini ex_configs/stationarity.ini ex_configs/omega1.ini }}} * The generated DAG files are located in {{{/home/controls/public_html/events/}}} * Because we don't store the O3 data in KAGRA site, we need to run this job with gps option. {{{ mkdir ~/public_html/event dqr-create-dag -g 1318260000 -v -c ex_configs/defaults.ini ex_configs/condor.ini ex_configs/stationarity.ini ex_configs/omega1.ini }}} == Submit DAG jobs == * For example on k1det1 {{{ cd /home/controls/public_html/events/ condor_submit_dag S190924h/data_quality_report_S190924h.dag }}} * Result plots and html are {{{~/public_html/eventsS190924h/omegascan/plots/}}} and {{{~/public_html/eventsS190924h/omegascan/about/index.html}}} == Remained taks == * Investigate typical data duration for DQR * Measure the computational time for one channel * Investigate the availability of how many channels we can use for DQR by our computers * (Although this will finish just before O4,) fix the channel list for DQR * Try to use other tool * For example, [[https://detchar.docs.ligo.org/dqrtasks/stationarity.html|user's manual for {dqr-stationarity}]] * I need to investigate how to use them, although the other tools are listed up in design document. |
Task Manager
Installation
Yuzurihara prepared the virtual environment o4_dqr_proto on k1det1.
- DQR tasks are supported by python 3.8
This procedure will require the authentication of git.ligo.org.
To read KAGRA data by using NDS, We need to edit the code (see instruction.)
conda create --name o4_dqr_proto python=3.8 -y conda activate o4_dqr_proto conda install m2crypto ciecplib python-nds2-client cryptography==3.3.1 -y pip install git+https://git.ligo.org/o4-dqr/o4-dqr-configuration.git pip install git+https://github.com/gwdetchar/gwdetchar.git pip install git+https://git.ligo.org/detchar/dqrtasks.git
install ligo-proxy-init, from the LDG Client package
wget http://www.lsc-group.phys.uwm.edu/lscdatagrid/doc/ldg-client.sh -O /tmp/ldg-client.sh && sudo bash /tmp/ldg-client.sh
- X.509 certification
- For the additional permission to access GraceDB, we need to ask LIGO+Virgo colleague.
There are two options for KAGRA account. Here Yuzurihara used ligo-proxy-init.
ligo-proxy-init -i shibbi.pki.itc.u-tokyo.ac.jp albert.einstein
- To test the certification,
gracedb ping
grid-proxy-info
Generate DAG files
By following instruction, prepare ini files for DQR task manager.
- This procedure will NOT require the access GraceDB, because the public alert will be used.
For test code, see /users/yuzu/work/DQR/Script/o4-dqr-configuration:k1det1
mkdir ~/public_html/event dqr-create-dag -g S190924h -v -c ex_configs/defaults.ini ex_configs/condor.ini ex_configs/stationarity.ini ex_configs/omega1.ini
The generated DAG files are located in /home/controls/public_html/events/
- Because we don't store the O3 data in KAGRA site, we need to run this job with gps option.
mkdir ~/public_html/event dqr-create-dag -g 1318260000 -v -c ex_configs/defaults.ini ex_configs/condor.ini ex_configs/stationarity.ini ex_configs/omega1.ini
Submit DAG jobs
- For example on k1det1
cd /home/controls/public_html/events/ condor_submit_dag S190924h/data_quality_report_S190924h.dag
Result plots and html are ~/public_html/eventsS190924h/omegascan/plots/ and ~/public_html/eventsS190924h/omegascan/about/index.html
Remained taks
- Investigate typical data duration for DQR
- Measure the computational time for one channel
- Investigate the availability of how many channels we can use for DQR by our computers
- (Although this will finish just before O4,) fix the channel list for DQR
- Try to use other tool
For example, user's manual for {dqr-stationarity}
- I need to investigate how to use them, although the other tools are listed up in design document.