If the '-o' option is not used, then the MEX function is the name of the MATLAB® function with '_mex' appended. The '-o' option is used to name the MEX function that is generated. The output of buildInstrumentedMex is a MEX function with instrumentation inserted, so when the MEX function is run, the simulated minimum and maximum values are recorded for all named variables and intermediate values. The inputs to buildInstrumentedMex are the same as the inputs to fiaccel, but buildInstrumentedMex has no fi object restrictions. To instrument the MATLAB® code, create a MEX function from the MATLAB® function using the buildInstrumentedMex command. Use Min/Max Instrumentation to Identify Overflows The error (bottom plot) is much larger than what you would expect to see for round off error, so it is likely that overflow has occurred. Note that the magnitude plot (center) of the fixed-point FFT does not resemble the plot of the built-in FFT. % %#codegenĮnd end end end Type-aware bit-reversal functionĪdd types table T as an input to the function and use it to cast variables to a particular type, while keeping the body of the algorithm unchanged. % % Reference: % Charles Van Loan, Computational Frameworks for the Fast Fourier % Transform, SIAM, Philadelphia, 1992, Algorithm 1.6.2, p.
% % See also FI_RADIX2FFT_DEMO, FI_M_RADIX2FFT_WITHSCALING.
% % This version of the algorithm has no scaling before the stages. % % T is a types table to cast variables to a particular type, while keeping % the body of the algorithm unchanged. % Twiddle-factors W are computed via % W = fi_radix2twiddles(N) % where N = length(X). % % The length of vector X must be an exact power of two. % Y = FI_M_RADIX2FFT_ALGORITHM1_6_2_TYPED(X, W, T) computes the radix-2 % FFT of input vector X with twiddle-factors W. %FI_M_RADIX2FFT_ORIGINAL_TYPED Radix-2 FFT example. To verify that you correctly implemented the algorithm in MATLAB®, run a known signal through it and compare the results to the results produced by the MATLAB® FFT function.Īs seen in the plot below, the error is within tolerance of the MATLAB® built-in FFT function, verifying that you have correctly implemented the algorithm.įunction x = fi_m_radix2fft_algorithm1_6_2_typed(x, w, T) % % Copyright 2004-2015 The MathWorks, Inc. % Reference: % Charles Van Loan, Computational Frameworks for the Fast Fourier % Transform, SIAM, Philadelphia, 1992, Algorithm 1.6.2, p. % Y = FI_M_RADIX2FFT_ALGORITHM1_6_2(X, W) computes the radix-2 FFT of % input vector X with twiddle-factors W. %FI_M_RADIX2FFT_ALGORITHM1_6_2 Radix-2 FFT example. Function x = fi_m_radix2fft_algorithm1_6_2(x, w)
0 kommentar(er)
