Re: About ezfftf

From: Dennis Shea (shea AT XXXXXX)
Date: Wed Mar 12 2003 - 13:23:13 MST

  • Next message: Xuguang Sun: "Is there Kolmogorov-Smirnov Test or Kuiper's Test in NCL?"

    > I am confused by the function of ezfftf. The output is supposed to be
    >the coefficients of the harmonics. For example, cf(0,i) and cf(1,i) are
    >the coefficients of some sin and cos harmonic components. How to express
    >the wave length or the frequency of that harmonics in terms of "i" and the
    >interval and number of the periodic time series? Assume the time series
    >data has interval of T, and the total number is n.

    Based upon an off-line email between Jianjun and I
    it was determined that the NCL function "fourier_info"

      http://ngwww.ucar.edu/ngdoc/ng/ref/ncl/functions/finfo.html
      
    was appropriate for his needs.

    However, to further clarify the frequencies to which
    the individual real [ cf(0,:) ] and imaginary [ cf(1,:) ]
    coefficients refer, we will add the following to the
    ezfftf documentation:

    The frequency spacing is 1./N. The frequencies that
    the 'cf' correspond to are:
        
         cf(0,0) and cf(1,0) corresponds is 1./N
         cf(0,1) and cf(1,1) 2./N
         cf(0,2) and cf(1,2) 3./N
         etc

    ----
    Regards
    Dennis Shea 
    

    _______________________________________________ ncl-talk mailing list ncl-talk@ucar.edu http://mailman.ucar.edu/mailman/listinfo/ncl-talk



    This archive was generated by hypermail 2b29 : Wed Mar 12 2003 - 13:08:45 MST