Chris@82: Next: Memory Allocation, Previous: Complex numbers, Up: Data Types and Files [Contents][Index]
Chris@82:You can install single and long-double precision versions of FFTW,
Chris@82: which replace double with float and long double,
Chris@82: respectively (see Installation and Customization). To use these
Chris@82: interfaces, you:
Chris@82:
-lfftw3f or
Chris@82: -lfftw3l instead of (or in addition to) -lfftw3. (You
Chris@82: can link to the different-precision libraries simultaneously.)
Chris@82:
Chris@82: <fftw3.h> header file.
Chris@82:
Chris@82: fftw_complex becomes fftwf_complex, fftw_execute
Chris@82: becomes fftwf_execute, etcetera.)
Chris@82:
Chris@82: double with float or long double for
Chris@82: subroutine parameters.
Chris@82:
Chris@82: Depending upon your compiler and/or hardware, long double may not
Chris@82: be any more precise than double (or may not be supported at all,
Chris@82: although it is standard in C99).
Chris@82:
Chris@82:
We also support using the nonstandard __float128
Chris@82: quadruple-precision type provided by recent versions of gcc on
Chris@82: 32- and 64-bit x86 hardware (see Installation and Customization).
Chris@82: To use this type, link with -lfftw3q -lquadmath -lm (the
Chris@82: libquadmath library provided by gcc is needed for
Chris@82: quadruple-precision trigonometric functions) and use ‘fftwq_’
Chris@82: identifiers.
Chris@82: