The building of a regular grid in CLASS is done in two
steps, in a similar fashion as the production of a PdBI
regular grid from UV tables. The first step involves the
building of a table: the spectra are written as rows in the
Gildas internal data format for efficiency (command
table). When creating the TABLE, the spectra are
not resampled on a grid. The second step is the
resampling of the non-regularly spaced spectra on a regular
grid (command xy_map).
A simple example is given below where the cube is produced
from a single .30m file. Default parameters are used by the
xy_map command to define both the convolution kernel
(1/3 of the 
) and the output grid (1/2 of the
).
file in map ! Open the input file find ! Build the index consistency ! Check first that the index is consistent let name thecube ! Use global SIC variables to define output cube name let type lmv ! The output cube is named ``thecube.lmv'' table 'name' new ! Build the non-gridded table xy_map 'name' ! Grid the data from the table go view ! Check the quality of the data reduction with the VIEW tool
The following example illustrates the way to build a cube from several .30m files. This is typically the case when the map has been repeated several times and the data have been calibrated and reduced in different files.
file in map1 ! Open the input file find ! Build the index consistency ! Check first that the index is consistent let name thecube ! Use global SIC variables to define output cube name let type lmv ! table 'name' new ! Build the non-gridded table as a new table file in map2 ! Open the input file find ! Build the index table 'name' ! Append the index to the current table file in map3 ! Open the input file find ! Build the index table 'name' ! Append the index to the current table ... xy_map 'name' ! Grid the data from the global table go view ! Check the quality of the data reduction with the VIEW tool