4.2 Record keeping and data management

4.2.1 Calibrating data with multiple FQ entries

An observing run with the VLA may result in a uv data file containing multiple FQ entries. This may be convenient, but it has a number of costs. If a file contains multiple, independent frequencies, then it occupies more disk space and costs time in every program to skip the currently unwanted data (either a small cost when the index file is used or a rather larger cost when the file must be read sequentially). Since multiple frequencies are still not handled correctly in all programs (i.e., polarization calibration) and since it is not possible to calibrate all of the different FQ data in one pass, you might consider separating the multiple frequencies into separate files (use UVCOP). In either case, you must calibrate each frequency with a separate pass of the scheme outlined below. There are three adverbs to enable you to differentiate between the different FQ entries: FREQID enables the user to specify the FQ number directly (with -1 or 0 meaning to take the first found); SELFREQ and SELBAND enable the user to specify the observing frequency and bandwidth to be calibrated (the tasks then determine to which FQ number these adverbs correspond). If SELFREQ and SELBAND are specified they override the value of FREQID.

There are certain bookkeeping tasks that must be performed between calibrating each FQ set. First, you must ensure that you have reset the fluxes of your secondary calibrators by running SETJY with OPTYPE = ’REJY’ — if not, this will cause the amplitudes of your data to be incorrect. Second, it is wise to remove the SN tables associated with any previous calibration using the verb EXTDEST. Although this is not strictly necessary, it will simplify your bookkeeping.

4.2.2 Recommended record keeping

It is useful to print a summary of the time stamps and source names of the scans in your data set. This reminds you of the structure of your observing program when you decide on interpolation and editing strategies, and may help to clarify relationships between later, more detailed listings of parts of the data set. It is also useful to have a printed scan summary and a map of the antenna layout if you need to return to processing the data months or years later. Finally, it is also making sure that all  input parameters have their null (default) values before invoking the parts of the calibration package, such as CALIB, that have many inputs. The null settings of most parameters are arranged to be sensible ones so that basic VLA calibration can be done with a minimum of specific inputs; but some inputs may lose their default values if you interleave other  tasks with the calibration pattern recommended below. Therefore, you should always review the input parameters with INP taskname  C R before running task taskname.

We suggest that you begin a calibration session with the following inputs:

Note that the DEFAULT LISTR sets the adverbs to select all sources and all times and to send printed output to the terminal rather than the printer. Task LISUN is a useful alternative to LISTR. It is also very useful to have a printed summary of your antenna locations, especially a list of which ones you actually ended up using. To do this, enter

In looking over the output from LISTR, you may notice that some of the sources you wish to use as calibrators have a blank “Calcode”. To mark them as calibrators, use:

This operation will let you select the calibrators by their Calcodes rather than having to spell out their names over and over again. You may wish to consider separate calibrator codes for primary and secondary gain calibrators to make them easier to separate. You may reset a calibrator code to blank by specifying CALCODE = ’----’.