20.18.3 /MODE

[ADVANCED\]UV_MERGE OutFile /FILES In1 In2 ... Inn /MODE [LINE|STACK|CONCATENATE|INTERSECT|STITCH|CONTINUUM [Index [Freq]]] [/IGNORE] [/SCALES F1 ... Fn] [/WEIGHTS W1 ... Wn ]

Specify the merging mode. The default mode is LINE.

In LINE mode, for spectral line UV tables (more than 1 channel), the default is that the spectral lines must have the same Rest Frequency. Resampling (in velocity, which is then identical in frequency) is done on the grid of the first UV table In1.

In mode CONCATENATE, a common spectral axis encompassing those of the initial UV tables is defined, and all UV tables are resampled on this one. Spectral resolution is set by the first UV table.

In mode INTERSECT, the common spectral axis is the intersection of the coverage of all UV tables. Spectral resolution is set by the first UV table.

In mode STACK, the Rest Frequencies can differ. Resampling is then done in velocity, and the (u,v) coordinates scaled as the Rest Frequency ratios to conserve the angular resolution of the data. The /SCALES and /WEIGHT factors can be used to incorporate prior knowledge of the expected line ratios to optimize S/N.

For mode CONTINUUM, a spectral index can be specified, as well as a reference Frequency. (u,v) coordinates are scaled appropriately, as well as Flux and Weights in this case. The /SCALES and /WEIGHTS factors are applied on top of this automatic spectral index scaling. Input Line UV Tables are treated as multi-frequency Continuum ones (as in UV_CONTINUUM command).

In mode STITCH, line (even single channel) tables with different frequencies but the same uv coverage are glued together to create a UV table with wider frequency coverage. The input UV tables must be time-baseline ordered. The output UV table is given a random frequency axis, where each channel has its own frequency. This operation may be needed for data coming from the EVLA for example.

**** mode STITCH is still under development ***