C.2 Intrinsic Functions

The tables within this section summarize Fortran intrinsic (or library) functions. These intrinsic functions are categorized as follows:

Category	Description and Functions
Mathematical	Functions that evaluate mathematical functions: ACOS, ACOSD, ASIN, ASIND, ATAN, ATAND, ATAN2, ATAN2D, COS, COSD, COSH, EXP, LOG, LOG10, SIN, SIND, SINH, SQRT, TAN, TAND, TANH	(Section C.2.1)
Numeric	Functions that perform type conversions and numeric operations: ABS (and IABS), CMPLX, CONJG, DBLE, DCMPLX, DFLOAT, DIM (and IDIM), DPROD, FLOAT, IFIX, IMAG, INT (AINT, IDINT, and IQINT), MAX (MAX0, MAX1, and AMAX0), MIN (MIN0, MIN1, and AMIN0), MOD, NINT (ANINT, IDNINT, and IQNINT), QEXT, QFLOAT, REAL (and SNGL), REAL (and DREAL), SIGN (and ISIGN), ZEXT	(Section C.2.2)
Bit Manipulation	Functions that manipulate bits in the binary patterns that represent integer data types: BTEST, IAND, IBCLR, IBITS, IBSET, IEOR, IOR, ISHFT, ISHFTC, NOT	(Section C.2.3)
Character	Functions that perform character operations: CHAR, ICHAR, INDEX, LEN, LGE, LGT, LLE, LLT	(Section C.2.4)
Miscellaneous	Functions that return information on processes: NWORKERS; bytes of storage: SIZEOF; whether a number is a NaN value: ISNAN (Alpha only); and the class of an IEEE real argument: FP_CLASS (Alpha only).	(Section C.2.5)

Each table shows the following information:

Function: A short description of the function; some functions show mathematical definitions.
No. of Arguments: The number of arguments that each function requires.
Generic Name: The generic names you can use to refer to the function. If you use a generic name, the data type of the argument determines which specific function the processor calls.
Specific Name: The specific names you can use to explicitly choose a combination of argument data type and result data type.
Type of Argument: The data type that the function argument must be.
Type of Result: The data type that the function returns.

In a table, numbers in brackets refer to footnotes following the table.

For example, the following is the mathematical table entry for the ACOSD intrinsic function:

Function	No. of Arguments	Generic Name	Specific Name	Type of Argument	Type of Result
Arc Cosine [1,2](degree)	1	ACOSD	ACOSD	REAL*4	REAL*4
Arc Cos a			DACOSD	REAL*8	REAL*8
			QACOSD[3]	REAL*16	REAL*16

The following explains the table information for ACOSD:

Function	This function returns the arc cosine of the argument in degrees. The short description is "Arc Cosine (degrees)". This description also shows a mathematical definition "Arc Cos a" (the numbers in brackets refer to table footnotes 1 and 2).
No. of Arguments	This function requires one argument.
Generic Name	ACOSD is the generic name.
Specific Name	ACOSD has three specific names: ACOSD, DACOSD, and QACOSD (the superscript refers to table footnote 3).
Type of Argument	For ACOSD, the argument must be REAL4; for DACOSD, the argument must be REAL8; for QACOSD (VMS, UX), the argument must be REAL16.
Type of Result	For a REAL4 argument (ACOSD), the result returned is REAL4; for a REAL8 argument (DACOSD), the result returned is REAL8; for a REAL16 (VMS, UX) argument (QACOSD), the result returned is REAL*16.

On intrinsic functions, in general, see Section 6.3.
On bit manipulation intrinsic functions, in general, see Section D.2.
On character and lexical intrinsic functions, in general, see Section 6.3.2.
On intrinsic functions that allow compatibility with older versions of Fortran, see Section A.8.
On system subprograms, see Section D.1.
On specific data types, see Sections 2.1 and 2.2.
On BLAS routines available on VAX processors, see Section C.3.