Description

CIAO extends the concept of type 1 and type 2 PHA files to general tables. A type 1 file of size N is a simple table consisting of scalar columns each with N rows. A type 2 file of size N x M is a table with M rows in which each column is either scalar or an array of size N. Each row of the type 2 file corresponds to a whole type 1 file, and `dmtype2split' performs this mapping.

The mapping consists of turning each array column entry from `sideways' to `vertical', so that an array of 4096 values becomes a column with 4096 rows. Scalar columns map to keywords in the output file.

Each of the rows in the type 2 file specified by either the "rows" parameter (whose use is currently being 'deprecated') or the datamodel filter syntax (i.e. infile="filname.fits[#row=N:M,P:Q,..]"), is copied to a type 1 file. The rows are specified in a comma- and colon-delimited list. Each output file has the basic header keywords and subspace copied from the input. If the input file is a pha file, keywords required by the OGIP format are also written; otherwise, generic keywords are written. For each scalar column in the input, the coordinate keyword information is written, and the datatype, value, and unit are written as a keyword. For each input column that is an array, a column is created from the datatype, value, and unit; the row of input data is read and written to the output column.

The rows can be specified in the comma- and colon-delimited list in any order (e.g., "1:4,6,9:12" or "6,1:4,9:12" or "1:4,9:12,6", etc ). A colon-delimited interval is inclusive (e.g., "2:5" means rows 2, 3, 4, 5). If the number of rows specified does not equal the number of output files specified, the program exits with an error message. If no rows are specified, rows 1,2,...,n are written, where n is the number of output files in the outfile stack.