EPICS and scalers for E89044

Epics and Scaler Information in the E89044 Data Files

In /work/halla/e89044/disk2/kuss/nonCoda you will find two directories. ./epics contains files <datafile>.epics which contain the EPICS event information of the corresponding <datafile>, ./scaler is similar except for the scaler events.
These files were created by ~kuss/bin/SunOS/epics, which was running permanently on adaql2 since beginning of March. Files taken before were analyzed from either the work disks (if found there) or requested from the silo and analyzed. I checked that all files were completely analyzed. However, for the files copied to the work disks by other users I don't know if they were copied completely to the disk, e.g. the disk may have been full or the copy process was terminated for any reason.
The information contained in these files is conveniently accessible by two scripts.

~kuss/bin/grepScaler
usage: grepScaler spec,bank:channel[,bank:channel[,...]] run#
It is a simple script to grep scaler data from the *.scaler files. spec is either e or h. It is comma-separated from a list of also comma-separated pairs of bank #s and channel #s, which are separated by a colon.
example: grepScaler e,5:8,5:9 1006
greps the clock and Unser information of the electron arm from the scaler/e89044_1006.dat.*.scaler files and prints them to stdout. You can find the raw data structure, which is frequently updated by Bob, on the Hall A

example 2: grepScaler h,5:8,2:3,1:2,7:1 2007

greps the hardware clock, S1 right 3rd paddle, S1 left 2nd paddle, and gas Cerenkov 1st mirror of the hadron arm from the scaler/e89044_2007.dat.*.scaler files (actually, here is only a *.0.scaler).

~kuss/bin/Linux/grepEPICS
usage: grepEpics parameter[,parameter[,...]] firstrun [lastrun]
It is a simple script to grep epics data from the *.epics files.
example: grepEpics hac_bcm_average 1843
greps the hac_bcm_average information from the epics/e89044_1843.dat.*.epics files and prints it to stdout. Anyway, the column contains the date extracted from the epics event. The units for the date are the days since Nov 1 00:00:00 EST 1999.
example 2: grepEpics haITC502_1_Sensor_1_R,haITC502_1_Sensor_2_R 2500 2511
greps two target temperature readings of the first ITC502 from all files between run # 2500 and 2511 (inclusive) and prints them to stdout.
Flaws:
- It can only run on Linux (e.g. ifarml3) because it uses the functionality of Linux's date command.
- In order to convert the date string it has to create two temporary files which makes grepEpics very slow. I don't get the quote expansion off, and the system command in gawk does not return the command's output.

Please note that both scripts don't perform any data manipulation, they just extract the numbers they find in the files. After you have piped the script's output into a file you may want to consider further treatment. Currently, I think of two:

If I remember correctly, the first scaler event contains any numbers (probably from the previous run?). Sometime after that, the scalers are resetted, and the next event contains reasonable numbers. I don't have to much experience with this, so you may want to check for yourself.
Looking at e.g. the long term behaviour of the target temperatures one should exclude conditions where the target was not stable, e.g. by checking also the beam current or by requiring e.g. the last five reading to be within a small range. I will apply checks like these when studying the target behaviour.

Michael Kuss, May 24, 2000