Re: The filter response of filwgts_lancos?

> but the plot of the filter function puzzled me,
> Why the wgt_at_resp (amplitude) converge to the zero when
> wgt_at_freq nearly equal to 0.4? I suggest that wgt_at_resp should
> be zero when wgt_at_freq equal to cutoff freq (0.2).

Conceptually, you are correct. However, you are asking
the impossible. Using 7 wgts for a low pass cut-off at 0.2.
You must use more weights. See

    http://www.cgd.ucar.edu/web/public_html/shea/RESP_LANCOS.gif

The upper right indicates the number of weights used.
Digital filter design is as much an art as it is
a science.

On Fri, 12 Jan 2007, LEO_ARIES wrote:

> Hi,all
>
>
> I've been using "filwgts_lancos" to do a low pass filter,
>
>
> nwt = 7
> fca = 0.2 ; cutoff at 5 yr period
> ihp = 0 ; use low pass
> nsigma = 1. ; "nsigma = 1. is common"
> wgt = filwgts_lancos (nwt, ihp, fca, -999., nsigma)
> wgt = wgt/sum(wgt) ; normalize wgts
> ann_5yr_filter = wgt_runave (ann_pre, wgt, 0) ; compute running ave
> jja_5yr_filter = wgt_runave (jja_pre ,wgt, 0)
>
> I also use some codes as following to do a visual examination of the filter response
> ; test wgt
> wks = gsn_open_wks ("ps","Lancos_low_pass_filter_cutoff_0.2")
> res3 = True
> res3_at_tiMainString = "ihp="+ihp+" fca="+fca+" nwt="+nwt
> plotu = gsn_csm_xy (wks,wgt_at_freq, wgt_at_resp,res3)
>
> but the plot of the filter function puzzled me, Why the wgt_at_resp (amplitude) converge to the zero when
> wgt_at_freq nearly equal to 0.4? I suggest that wgt_at_resp should be zero when wgt_at_freq equal to cutoff freq (0.2).
>
>
> I have no access to the earlier reference ,
> Reference: C. Duchon
> J. Applied Meteorology; August,1979;
> pp 1016-1022
> Lancos Filtering in One and Two Dimensions
>
> Could anyone give me a short explainantion or send me a referece? Thanks !
>
>
> Good Luck
>
>
>
> Lin
>
>
>
> LEO_ARIES
> 2007-01-12
>
_______________________________________________
ncl-talk mailing list
ncl-talk_at_ucar.edu
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Fri Jan 12 2007 - 09:47:21 MST