A detailed description of the way in which the CCFs corresponding to a certain observation data set can be accessed, along with the parameters that need to be specified, can be found in § 3.11. The SAS command cifbuild retrieves the observation date from the observation to be analyzed and selects the calibration files for that given time period. Alternatively, the user can specify a given date for the CCFs to be used through one of the input parameters of the cifbuild task. All the different CCFs are kept in a repository accessible to the SAS user for building the calibration database.
Calling cifbuild with no parameters assumes the following defaults:
cifbuild calindexset = 'ccf.cif' withccfpath = no usecanonicalname = no \
ccfpath = '.' recurse = no fileglob = '*.ccf|*.CCF' fullpath = no \
withobservationdate = no observationdate = 'now' analysisdate = 'now'\
category = 'XMMCCF' ignorecategory = no masterindex = no \
withmasterindexset = no masterindexset = 'ccf.mif' append = no
The output of cifbuild is a file referred to as the Calibration Index File (CIF) (ccf.cif in the example above), created in the directory where cifbuild was run. We will refer to this directory from now onwards as /my_work. This directory will contain all the files corresponding to the results of the analysis of a particular observation. It is thus recommended to run all subsequent SAS tasks from this directory. This directory could be any directory.
Once cifbuild has been executed, a new environment variable, SAS_CCF, has to be set to point to this file, ccf.cif, in order for SAS to know where to locate it:
export SAS_CCF=/my_work/ccf.cif [sh, bash, ksh] setenv SAS_CCF /my_work/ccf.cif [csh, tcsh]