On Tuesday, April 23, 2013 6:25:10 PM UTC+12, ghasem wrote: > > Interval in frequency is: > > > df=1/(2*N*dt); > > > where N is no of data and dt is the interval between data. > > > f1 corresponds to index no: > > > indx=round(f1/df); > > > So, for dct Y: > > > Y(indx:end)=0; > > > y=idct(Y); > > ============================== > > in fact,I have a one dimensional numerical simulation in spatial domain,that my answer become divergent.I think that divergence is due to frequencies higher than cutoff frequency.So,I want to remove these frequencies,such that my result become convergent and don't change. > > I applied your code,but my divergence became worse... > > why ? > > ghasem
Yes, you're right, I'm totally to blame for the problems with your data. I've been in your gunsight ever since I woke up this morning. First of all, I had my breakfast before replying to your post and you rightly chastised me for that. Then, I failed to realise that you need things to be explained in words of one syllable, otherwise you do not understand. And now after I told how to do what you wanted to do, I've caused your problem to get worse rather than better. Oh dear, I'm hopeless. I'm off to bed, hoping to have a better day tomorrow.