I have extended the Matlab interface for NFFT to support also single precision. It currently accepts its inputs of f and f_hat both in single and double precision. The output is always the same as the precision with which the NFFT was built.
Since Matlab does not support extended precision, it would not make much sense to support the long double precision with NFFT.
I have extended the Matlab interface for NFFT to support also single precision. It currently accepts its inputs of
fandf_hatboth in single and double precision. The output is always the same as the precision with which the NFFT was built.Since Matlab does not support extended precision, it would not make much sense to support the long double precision with NFFT.