Description

S-Lang allows you to define variables that hold scalars, arrays, structures, or user-defined data types. Variable names in S-Lang are case sensitive, and the data type of a variable is determined upon assignment:

  variable foo, bar, baz;
  foo = 5;
  bar = [0, 5, 10, 15, 20];
  baz = "This is a string.";

After these assignments, foo is an integer (Integer_Type), bar an array (Array_Type) of integers, and baz a string (String_Type). It is also possible for a variable to be a structure, with fields that store data of different types:

  variable fileinfo = struct { pathname, filename, nrows };
  variable foo = @fileinfo;
  foo.pathname = "/data/ciao/";
  foo.filename = "evt2.fits";
  foo.nrows = 100;

The above defines a variable fileinfo to be a structure, and then populates the elements of this structure. Structures are used to store data returned by Varmm routines, where the data is stored in arrays, and the metadata - such as the number of rows in a table - are stored in fields beginning with a single underscore (ie '_') character. Note that the "@fileinfo" command uses the deference operator (@) to create an instance of the fileinfo structure.

The Varmm print() function can be used to view the content of a structure, and S-Lang contains a number of intrinsic functions, such as typeof(), for manipulating and querying variable types:

  chips> print(foo)
  pathname         =  /data/ciao/
  filename         =  evt2.fits
  nrows            =  100
  chips> print(typeof(foo))
  Struct_Type
  chips> print(typeof(foo.filename))
  String_Type
  chips> print(typeof(foo.nrows))   
  Integer_Type

Structures are used to store data read in by a Varmm function such as readfile(), or if you wish to create a FITS file using writefits(). In the following, we read in an ASCII file containing two columns into a structure, and then use the print() function to view its contents.

  sherpa> AGauss = readascii("phas.dat");
  sherpa> print(AGauss)
  _filename        =  phas.dat
  _path            =  /data/analysis/
  _filter          =  NULL
  _filetype        =  1
  _header          =  NULL
  _ncols           =  2
  _nrows           =  128
  col1             =  Float_Type[128]
  col2             =  Float_Type[128]

Here we use another Varmm function, readfile(), to read in selected columns from an event list. Note that the filename can contain DM filters - here we restrict access to the first ten rows and select only the time and status columns:

  chips> evt = readfile("evt2.fits[#row=1:10][cols time,status]")
  chips> print(evt)
  _filename         =  evt2.fits
  _path             =  /data/ciao/
  _filter           =  [#row=1:10][cols time,status]
  _filetype         =  4
  _ncols            =  2
  _nrows            =  10
  time              =  Double_Type[10]
  status            =  UChar_Type[10,4]