Online analysis

From PREX Wiki
Revision as of 23:11, 21 June 2019 by Ajzec (Talk | contribs)

Jump to: navigation, search

In the process of editing and updating this page. Please contact Juliette Mammei crowder@jlab.org with questions or suggestions.

PREX Main<< Information for Shift Takers << Shift Leader

For Shift Workers

  • The prompt analysis should run automatically as a background process
  • To look at the output, use Panguin
  • For anything else, ask an expert!

For Experts

General Instructions

The expert shifter must ensure that the online japan analysis is running and monitor panguin plots. In addition, they must execute offline analysis as instructed below.

Online JAPAN

To run an online instance of the JAPAN decoder execute

gojapan
./build/qwparity --config prex.conf --add-config online_apar.conf

This will automatically grab the current CODA run (it attaches to the 64 bit ET system and analyzes data on the fly) Looking at live plots can be done with PANGUIN in live mode as well, as long as your PANGUIN macro has the line

watchfile

included in it. A live PANGUIN session will update the plots on a few second cycle, but it can take some time to loop over all of the data. It is possible to trim down the ROOT file output size, which is especially useful in these circumstances. See here for more information.

Offline JAPAN

First login to apar@aonl1.jlab.org. After the end of each run execute:

gojapan
./build/qwparity -r <run number> -c prex.conf

If auto-prompt is not running, then execute (with apar credentials)

cd ~/PREX/prompt
./auto_prompt.sh <run number> <0 or 1>

Use 0 option if you don't want to re-analyze existing runs, and 1 if you want to re-analyze existing runs.

Macro for Panguin

Experts can open a custom panguin session on completed runs. Instructions are listed in the "Expert" section on the Panguin page.