is there any problem in my FFT?

AB = C (time domain data)

A and C : known

B : unkown

length of A,B and C is 512

N1 = 512;
t = t1(1:N1-1);
A = A1(1:N1-1);
C =C1(1:N1-1);

T1 = tpp*N1;

fpp             = 1/T1;
a1              = 0.3906;
nr              = N1*a1;
fs              = 1/tpp;
df1             = 0:fs/N1:(N1-1)*fs/N1;
dfp1            = df1(1:nr);

fft1            = fft(A);
fft_Mag_A(1)    = (1/N1)*abs(fft1(1));
fft_Mag_A(2:nr) = (2/N1)*abs(fft1(2:nr));
fft_Phase_A     = angle(fft1(1:nr));
fft2            = fft(C);
fft_Mag_C(1)    = (1/N1)*abs(fft1(1));
fft_Mag_C(2:nr) = (2/N1)*abs(fft1(2:nr));
fft_Phase_C     = angle(fft1(1:nr));

B    = fft_Mag_C/fft_Mag_A*cos(w*t+fft_Phase_C - fft_Phase_A);

fft3 = fft(B);
fft_Mag_B(1)      = (1/N1)*abs(fft3(1));
fft_Mag_B(2:nr)   = (2/N1)*abs(fft3(2:nr));
fft_Phase_B(1:nr) = angle(fft3(1:nr));

C1 = fft_Mag_A*fft_Mag_B*cos(w*t+fft_Phase_A+fft_Phase_B)

here is a problem.

why C1 and C(data) doesn't match perfectly each other when i draw graph?

There's a slight error.

[David Goodmanson]

Hi SK,

could you explain the purpose of doubling the magnitudes between 2 and nr, and also why the magnitude and phase of C are based on fft1 and not fft2?