Why use the CFITSIO Library?

There are many reasons why CFITSIO is the most widely used FITS interface library.

• CFITSIO is the result of many man-years of development with major contributions from external projects such as the data centers for INTEGRAL (ISDC), XMM, and Chandra space missions.
• Provides 100% complete support for the FITS Standard, including:
  • all FITS datatypes, including bit, byte, logical, and complex
  • both signed and unsigned integer support
  • vector columns
  • variable length vector columns
  • null values in images and table columns
• Supports many proposed FITS conventions that are widely used but not yet standard, including:
  • CHECKSUM keyword convention
  • CONTINUED long string value keywords (longer than 68 characters)
  • ESO HIERARCH convention for keyword names longer than 8 characters
  • Hierarchical Grouping convention (used by INTEGRAL)
  • Keyword units convention (in square brackets in the comment field)
  • 64-bit long-integer datatypes in FITS images and tables.
• C++, Fortran, Perl, and Tcl wrappers are available for calling CFITSIO from other languages
• The subroutine interface is based on an abstract data model concept which hides the low level FITS format details from users.
• The CFITSIO data access routines are simple and intuitive. The basic interface is defined in only 15 pages of the CFITSIO Quick Start Guide.
• A dozen simple yet powerful example programs are provided to illustrate how to use CFITSIO.
• Supports creating new files based on ASCII template file (from the ISDC)
• Supports implicit datatype conversion when reading or writing FITS data
• CFITSIO uses plug-in I/O drivers (developed at the ISDC) which makes it easy to to support a variety of physical file formats, including:
  • IRAF images (code provided by Doug Mink)
  • raw binary data files
  • FTP, HTTP, and ROOT file access (code provided by ISDC)
  • externally compressed files (.gz, .Z, and .zip)
  • files piped on 'stdin' or 'stdout'
  • 'Tile Compressed' images which are actually stored in binary tables
  • files in shared memory segments (code provide by ISDC)
  • files is core memory (widely used in real-time data acquisition)
• Has a simple set of self-contained WCS routines, plus has routines to cleanly interface with Doug Mink's more extensive WCS library (but has no linkage dependencies on the WCS library)
• The high performance lexical parser (designed by Uwe Lammers) allows completely general table filtering and is as fast as compiled C code.
• The parser supports filtering of tables by region files and GTI files.
• CFITSIO's Virtual File syntax supports many features, including:
  • extracting subimages from larger images
  • modifying or creating table columns on the fly
  • creating images by binning table columns
• The I/O speed has been highly optimized and is largely limited only by the speed of the underlying physical device; Sustained I/O data rates of more than 40MB/s are possible on current generation magnetic disks.
• Has modest memory requirements; FITS files sizes are not limited by the amount of available memory.
• Supports the 'large file' specification and can read and write FITS files up to 6 terabytes in size.
• Extensive error reporting is built into the interface, which aids both developers and end users in debugging problems.
• CFITSIO can be built on nearly every Unix platform with the usual './configure' and 'make' procedure. It supports 64-bit architectures.
• A binary .dll library is available for Windows platforms.