Re: question about linmsg interpolating missing value in a region that result has not changed as before ?

From: dyjbean <dyjbean_at_nyahnyahspammersnyahnyah>
Date: Tue Nov 26 2013 - 08:50:30 MST

 if those missing values exist over lakes, but i want to fill them with arounding values, what you said they stand
at the same location is true, but the linmsg interpolation has left them uninterpolated .




dyjbean

From: Maria Gehne
Date: 2013-11-26 23:44
To: dyjbean; ncl-talk
Subject: Re: [ncl-talk] question about linmsg interpolating missing value in a region that result has not changed as before ?
Do you know what parts of your data are missing values? If the missing values are always in the same location (like missing values over land, ocean or lakes) then interpolating in time is not going to give you a different result than what you started with.

Maria

On 11/25/2013 07:16 PM, dyjbean wrote:

hi,
  when i processed the missing value with linmsg interpolation method ,there met a question as follows:
the part of codes is as follows:

______________________________________________________________
    lat!0 = "lat"
    lat@long_name="lat"
    lat@units="degrees-north"
    lat&lat=lat

    lon!0 = "lon"
    lon@long_name="lon"
    lon@units="degrees-east"
    lon&lon=lon

    slrad!0="time"
    slrad!1="lat"
    slrad!2="lon"
    slrad&lat=lat
    slrad&lon=lon
    slrad&lat@units="degrees-north"
    slrad&lon@units="degrees-east"

    slrad=where(ismissing(slrad),2.e+20,slrad)
    slrad@_FillValue=2.e+20
    slrad@missing_value=2.e+20
    print(num(ismissing(slrad))+" before")
;; linear interpolation
     sslrad=linmsg(slrad(lat|:,lon|:,time|:),-1)
     sslrad!0="lat"
     sslrad!1="lon"
     sslrad!2="time"
     slrad=(/sslrad(time|:,lat|:,lon|:)/)
     print(num(ismissing(sslrad))+" temp")
     print(num(ismissing(slrad))+" after")
     delete(sslrad)
________________________________________________________________

the output result in screen is as follows

(0) 65472 before
(0) 65472 temp
(0) 65472 after


it seemed the linear interpolation didn't work.
it's strange appearance




dyjbean

 

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk




--
Maria Gehne
Postdoctoral Fellow
NCAR/CGD/CAS
P.O. Box 3000
Boulder, CO 80307

_______________________________________________
ncl-talk mailing list
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk
Received on Tue Nov 26 08:50:41 2013

This archive was generated by hypermail 2.1.8 : Wed Dec 04 2013 - 20:42:38 MST