====== GDL (IDL) for beginners ======

The programming language GDL -- a free variant of IDL -- uses library functions that were specially designed for the analysis of astrophysical 
data and includes routines to read and write FITS files.

The interpreter is started in the console with:
  gdl

Now enter the GDL program code line by line and the interpreter executes it right away.

However, it's more convenient to write the program as a file. Open a text editor of your choice (//kate//, //emacs//, //nedit//, //vi//,...) and write/edit the program file (its file extension should be ''.pro''), then save it and compile it in GDL (in contrast to other languages like Fortran or C this will not create an executable file, because IDL/GDL works differently) using the command
  .compile myprog.pro
The first line of the program code gives the (internal GDL) name of the program after the key word //pro//, e.g. 
  pro prog1
After compiling the program "prog1" can be executed by calling
  prog1
in GDL.

====== Syntax ======

Attention, in contrast to the regular Linux console capitalization does not matter for commands in GDL, i.e.
  print, "Hello World!"
  PRINT, "Hello World!"
and even
  PrInT, "Hello World!"
return the same output. However, for strings and file names it's different, as usual.

==== Basics====

^                         ^  Commands  ^
| Calling the interpreter:|  ''gdl''  |
| Starting GDL help:    |  ''?''    |
| Show all variables:|  ''help''  |
| Compile a script: |  ''.compile myprogram.pro''  |
| Execute the script in gdl: |  ''myprogram''  |
| Leave gdl:         |  ''exit''  |
| First line of a script:             |  ''pro myprogram''  |
| Last line of a script (end of the program): |  ''end''  |

(Note: Replace "myprogram" with useful names that describe the program.)

==== FITS files ====

Read the content of a fits file into the variable //image1//:
  image1 = readfits('filename.fits')
If the image is not in the current directory, give the full (or relative) path (i.e. '../data/filename.fits').

Depending on the internal computer architecture, data consisting of more than one byte are saved in either high or low byte order. In rare cases the default byte order is different between the source system (where the fits file was created) and the reduction system. This can be seen in (seemingly) absurd values and can be easily noticed when viewing the image of the fits file. If that happens, convert the byte order:
  byteorder, image1

Save an array (e.g. //image2//) as fits file:
  writefits, 'newfilename.fits', image2

Note: All FITS commands require the (freely distributed) [[http://idlastro.gsfc.nasa.gov/homepage.html|AstroLib]] for GDL/IDL (which is installed for the praktikum account, see [[en:software:howto_gdl_install|here]] for a how-to).

==== Input/Output ====

If the variable //namevar// has a content (e.g. namevar='world'), create output by writing:

  print, "Hello ", namevar
  
Any number of values/strings/variables can be attached, separated by commas. At the end there will be a line break.

Write into a file:

  openw, handle, 'filename.dat'
  printf,handle,i,variable1,"some text",variable2
  close, handle

//handle// needs to be an arbitrary but unique integer value. Alternatively use:

  openw,out,'filename.dat',/get_lun
  printf,out,i,variable1,"some text",variable2
  close,out
  free_lun,out

In this case GDL will automatically assign a //handle// (via ///get_lun//). The command //free_lun// is needed to unassign that handle again.

==== Variables and arrays ====

Index variables start with the index 0. Thus, an array with the indices 0,1,...,MMAX has MMAX+1 entries.

Dimensioning of a two dimensional array for integer numbers:
  variablename = intarr(MMAX+1,NMAX+1)

In some cases lager values are required than provided by the integer type ($i>2^{15}-1=32767$). In this case use Long or even Long64 (32 bit or 64 bit) arrays:
  variablename = lonarr(MMAX+1,NMAX+1)
  variablename = lon64arr(MMAX+1,NMAX+1)
  
For floating point numbers (Float) use
  variablename = fltarr(MMAX+1,NMAX+1)

For double-precision numbers (Double) use
  variablename = dblarr(MMAX+1,NMAX+1)
  
For strings use
  variablename = strarr(MMAX+1,NMAX+1)
  
The number of entries in an array (e.g. //variablename//) can be checked with the command
  size(variablename,/dimensions)
which returns an array itself (with data type, size, etc.). To obtain the number of entries extract the first value (index 0!) of that array:
  number = (size(variablename,/dimensions))(0)

==== Control structures ====

Loop syntax:

  for j=n1,n2 do begin
    ...
  endfor

Conditional syntax: 

  if a gt b then begin
    ...
  endif

Additional information can be accessed via the help of GDL (type "?"). For more see the [[http://gnudatalanguage.sourceforge.net/documentation.php|Documentation page of GDL]].