How to compute WRMS of a time series

How to compute WRMS of a time series

[Chris Martin]

I did it

[Jules Ray]

and? I am searching for the same, til now I just computed the RSME

function [rmse,rms] = nanrms(dife,varargin)
%NANRMS  Root mean square.
% For vectors, NANRMS(x) returns the root mean square.
% For matrices, NANRMS(X) is a row vector containing the
% root mean square of each column. All NaN valuez
% are neglected.
  %
  % NANRMS(X,DIM) takes the rms along the dimension DIM of X. 
  %Dwonloaded from Matlab, modified by Julius Jara
  %See also: RMS, nanMEAN, nanMAX, nanMIN, STD

    % $Id$
    % $Date$
    % $Author$
    % $Revision$
    % $HeadURL$
    % $Keywords$

    if isempty(dife) 
    %if isempty(x) 
      %y = NaN;
        rmse = NaN;
      return
    end

    %    rmse = sqrt(nanmean(dife.^2,varargin{:}))/numel(dife);
    %   rmse=sqrt(sum(dife).^2)/numel(dife);
        rms= sqrt (nanmean (dife .^2) );
        rmse=rms/numel(dife);