atlas logo.gif (621 bytes) Running NORM on GEM data 

NORM is used to add the spectra, convert to a Q-scale, and perform a dead-time correction.

This page emphasises those aspects of running the NORM program which are specific to GEM.

NORM should be run from the a_b batch menu.

When you run NORM for GEM data, it produces a .SUM file containing output spectra which are not divided by monitor spectrum, in the same way as running NORM on SANDALS data. The full contents of a .SUM file may easily be read into GENIE by typing @g_f:allgem .

For a typical experiment you will have data for sample, container, vanadium and background. In this case NORM will need to be run 4 times, to add the data for each of these 4 experimental conditions.

When running NORM you will need to specify the Q-scale for your data, by giving values for the Q-spacing and maximum Q. You are advised to initially choose a large maximum value, Qmax, because this can be reduced at a later stage of the analysis, but it is not possible to increase Qmax after running NORM. Similarly you should not attempt to decrease the Q-step subsequent to running NORM.

It is most advisable to use the same Qmax and the same Q-step for all runs of NORM for an experiment. i.e. for the typical experiment you should use the same Qmax and Q-step when running NORM on the data for the sample, container, vanadium and background.

Calibration

Values for the calibration for the detectors (scattering angle 2-theta, flight path L2 and azimuthal angle phi) are stored in the RAW file together with the counting data. However, it is unavoidable that the values stored in the RAW file are not always the best and most recent values. In order to access the most up-to-date values you should run NORM as follows:

  1. Before running NORM copy the appropriate calibration file into the directory you will be using for your data analysis. The name of the appropriate calibration file may be found by clicking here. (If you have trouble finding the calibration file then you probably have not set up ATLAS correctly for GEM.)
  2. Rename the calibration file so that it is called detector.calib.
  3. Initiate NORM by typing a_b to start the ATLAS batch menu.
  4. When asked "Give 1 to change the calibration or 0 otherwise" type 1.

(If you run NORM without changing calibration on a run with bad calibration values stored in the RAW file, then you may find that the peaks observed in different detector banks do not line up.)

Once you have run NORM for the sample and vanadium data (and also optionally for the empty can and background) you can use the GEMSQRAW program in OpenGenie to obtain uncorrected plots of the differential cross-section for each detector bank.

Grouping of Detectors

A groups file is used by NORM to indicate which detectors are to be added together into each detector group. A number of different groups files is available in GEM's norm_par area, to be used when running NORM on GEM data. These groups files are described here. You are strongly advised not to choose the default groups file offered when setting up a NORM run. The reason is that GEM is still changing frequently, as detector banks are installed, and the default may well not be suitable for your data.

Further information about the GEM detector banks is available on the detector specification page and the GEM news page.

Dead-time Correction

At present the necessary files have not been created for doing a dead-time correction of GEM data and you should not attempt to perform a dead-time correction when running NORM. The correction will be implemented in the future, but its effect is likely to be very small because the count rate in individual GEM detector elements is actually very low.

Batch Queue

It is recommended to use the batch queue gem$batch for running NORM on GEM data and the batch menu should offer this as the default.

GEM User Rights

You will need to have GEM USER rights set up for your username. You can check whether you have such rights by logging on to ISISA (e.g. type SET HOST ISISA) and typing RIGHTS. If you do not have GEM user rights, please send an e-mail to support@isise.rl.ac.uk and ask for these rights.

Paging File Quota

If you have problems running ATLAS on GEM data, then one possible cause may be that you have an inadequate paging file quota (due to the large quantities of data generated by GEM). You can determine your current paging file quota by use of this command; sh proc/quo. Send an e-mail to support@isise.rl.ac.uk to ask for your paging file quota to be increased. I recommend a value of 400,000 blocks, although the support staff may be reluctant to give you such a large quota!

Large Log Files

NORM produces log files named like scratch$disk:[gem]gem00216.normb_log. For GEM data these log files can be very big, due to the large number of detectors, and there is a danger of exceeding your scratch disk quota after a few runs of NORM. Hence you are advised to check the log files for errors and then delete them. You can search the log files using an editor, or by using the following type of command:

search scratch$disk:[gem]gem00216.normb_log warning

search scratch$disk:[gem]gem00216.normb_log error

Once you have completed the NORM stage of the analysis, the log files produced by the subsequent programs of the ATLAS suite should be much smaller.

Last updated on 08 Apr 2002 by Alex Hannon (a.c.hannon@rl.ac.uk)