The db-class containing the data information from which a subpart must be extracted.

List of names of attributes that must be extracted. For more information see db.ident. If 'names' is not specified, all the attributes are extracted.

When TRUE, only the active samples are extracted. When FALSE, the selection is ignored.

When defined, this argument gives the array of indices (between 1 and db$nech) that must be kept in the output db-class.

This flag only makes sense when initial db-class structure is organized as a regular grid. When TRUE, the resulting file is a db-class (also organized as a grid) whose dimensions are calculated so as to contain the active samples of the initial Db.

When using the 'flag.keep.grid', the resulting db-class is dimensioned to the smallest area containing the active samples of the initial db-class. Nevertheless, you can wish to dilate this area. The argument 'nabove' provides the number of nodes for this dilation upwards along each axis of the working space.

If the length of 'nabove' is not equal to the space dimension, the argument is repeated up to the space dimension.

When using the 'flag.keep.grid', the resulting db-class is dimensioned to the smallest area containing the active samples of the initial db-class. Nevertheless, you can wish to dilate this area. The argument 'nbelow' provides the number of nodes for this dilation upwards along each axis of the working space.

If the length of 'nbelow' is not equal to the space dimension, the argument is repeated up to the space dimension.