4.9 Completing the initial calibration

When you are satisfied with the initial calibration (pre self-calibration) of your data set, you should back up your full multi-source data set on magnetic tape. Then you can apply the calibration to the data for each program source, creating a separate single-source uv data set for each. These data sets are used with the imaging and self-calibration tasks to be described in the following chapters. For the impatient, the recommended imaging task reads the multi-source data set directly, applying any calibration,

4.9.1 Using FITTP and FITAB to write multi-source data to tape

The recommended way out of AIPS for multi-source uv data is to use FITTP to write a FITS-format tape. This will preserve the data and all associated calibration and editing tables in a machine-independent form. FITAB also writes a FITS-format tape or disk file using tables rather than random groups. This has the advantages of allowing a compressed format and of allowing uv files to be broken into “pieces” for increased reliability and control of space. FITAB output from earlier than 15-Oct-2007 can be read by versions of AIPS between 15APR99 and 15-Oct-2007, after 15-Oct-2007 it can only be read by versions of UVLOD or FITLD later than 15-Oct-2007. Also note that FITAB output cannot be read by other uv-data software packages, except obit. FITTP output can be read by some other packages. Consult 3.9 about magnetic tapes in AIPS. That section tells you to mount your tape on the hardware device and then to do a software mount in AIPS. For example,

> INTAP n  C R

to specify which tape drive to use.

> DENSITY 6250  C R

to set the density to 6250-bpi, if needed.

> MOUNT  C R

to mount the tape in software.

This step used to be optional for some operating systems. However, in recent versions of AIPS, it is required on all operating systems.

To write the data to tape:

> TASK FITTP  C R

> IND m ; GETN n  C R

to specify the multi-source data set.

> DOEOT TRUE  C R

to write at the end of tape — if there are other data files on the tape you wish to preserve.

> OUTTA INTAP  C R

to write to tape just mounted.

> DOTABLE TRUE  C R

to write associated tables.

> FORMAT 3  C R

to use IEEE floating format for data.

> BLOCKING 10  C R

to use blocked FITS for tape efficiency.

> INP  C R

to review the inputs.

> GO  C R

to run the program when inputs set correctly.

Most people use 8mm Exabyte or 4mm DAT tapes today. These have very large capacities. However, if you must still use half-inch reel tapes, you will find that many data sets (particularly spectral line) may be too large to fit on one 6250 bpi tape even with BLOCKING = 10. Since it is not possible to write multi-volume FITS tapes, it is recommended that you back up the single-source data sets formed after applying the calibration tables in SPLIT (see 4.9.2). FITAB allows you to break up the data set into pieces which can fit on your tape. Multiple executions will be needed for multiple tapes. Alternatively, since all of the calibration information is contained in the extension tables, you may copy these to a dummy uv file with task TASAV and write this new file to tape with FITTP.

Be sure to run task PRTTP to make sure that the data were written successfully on your tape before you delete your multi-source uv data set!

4.9.2 Creating single-source data files with SPLIT

When you are happy with the calibration and editing represented by the current set of calibration and flag tables, you can convert the multi-source file into single-source files, applying your calibration and editing tables. Remember that only one FREQID can be SPLIT at a time.

> TASK SPLIT  C R

> SOURCE sou1’ , ’sou2’ ,  C R

to select sources, ’ ’ means all.

> TIMERANG 0  C R

to keep all times.

> BIF 1 ; EIF 2  C R

to keep both IFs

> FREQID 1  C R

to set the one FQ value to use.

> DOCALIB 1  C R

to apply calibration to the data and the weights.

> GAINUSE 0  C R

to use the highest numbered CL table.

> DOPOL TRUE  C R

to correct for feed polarization.

> DOBAND 3  C R

to correct bandpass with time smoothing.

> BPVER 1  C R

to select BP table to apply.

> STOKES ’ ’  C R

to write the input Stokes type.

> DOUVCOMP FALSE  C R

to write visibilities in uncompressed format.

> APARM 0  C R

to avoid channel averaging and autocorrelation data.

> INP  C R

to review the inputs.

> GO  C R

to run the program when inputs set correctly.

The files produced by this process should be completely calibrated and edited and ready to be imaged or further processed as described in later chapters. Note that one may wish to defer the DOPOL 1 part of this until after self-calibration of the parallel-hand visibilities.

It is not necessary to run SPLIT to make images with IMAGR and it is probably a good idea to make a couple of quick images to make sure that the calibration is okay. However, for serious imaging, it is probably best to run SPLIT and then use the single-source output files. See 5.2 for details of the imaging process.

4.9.3 Making images from multi-source data with IMAGR

IMAGR can be used to make images from multi-source data files. It is probably a good idea to make a couple of quick images to make sure that the calibration is okay. An example set of inputs to IMAGR is:

> TASK IMAGR’ ; DEFAULT  C R

to select task and initialize all its parameters. This selects the usual convolution and weighting functions among other things.

> IND m ; GETN n  C R

to specify the multi-source data set.

> SOURCE sou1’ , ’ ’  C R

to choose one source to image.

> STOKES ’I’ ; TIMERANG 0  C R

to image total intensity from all times.

> FREQID 1  C R

to select FQ value to image.

> BIF 1 ; EIF 0  C R

to image all IFs — multi-channel mode images only one IF.

> BCHAN n ; ECHAN m  C R

to combine a range of channels.

> NCHAV N  C R

to include N spectral channels in each image where N (m - n + 1); for each spectral channel, IFs bif through eif are also included. Note that each channel and IF included in the “average” image is handled individually at its correct frequency.

> DOCALIB 1  C R

to apply calibration. Use DOCAL 100  C R if the weights should not be calibrated.

> GAINUSE 0  C R

to use highest numbered CL table.

> FLAGVER 1  C R

to edit data.

> DOPOL TRUE  C R

to correct for feed polarization.

> DOBAND 3  C R

to correct bandpass with time smoothing.

> BPVER 1  C R

to select BP table to apply.

> OUTNAME sou1  C R

to set the output file name to the source name.

> OUTDISK 0  C R

to use any output disk with enough space.

> IMSIZE 512 512  C R

to set the size in cells of image.

> CELLSIZE 0.25 , 0.25  C R

to set the size of each image cell in arc-seconds.

> RASHIFT 0 ; DECSHIFT 0  C R

to (not) shift image center.

> NFIELD 1 ; NGAUSS 0 C R

to make only one image at high resolution.

> UVWTFN ’ ’  C R

to use uniform weighting.

> ZEROSP 0  C R

to introduce no zero-spacing flux.

> NITER 0  C R

to do no Cleaning.

> INP  C R

to review the inputs.

> GO  C R

to run IMAGR when the inputs are set correctly.