Fill blank spaces from lagged data with NaN

Question

Robert 2015-12-15

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/260346-fill-blank-spaces-from-lagged-data-with-nan

编辑： dpb 2015-12-16

I have a code which lags one data set in relation to another. The lag is determined based on r^2 values previously figured out.

The problem I have is that when the data is lagged, some data is removed. For example where there is a lag of +1 a vector of 96 rows becomes a vector of 95 rows. I want to fill that missing row is NaN so that it remains at 96. See code below. Thank you in advance for any help.

clear all;
clc
lag = load ('lagtime.txt'); %File contains the time by which each column will be lagged
      A1 = load ('data\Anomalies\TWSA1anomalies.mat');
      Approx1x = A1.TWSA1anomalies;
      Approx1y = reshape(Approx1x,[1505,96]);
      Approx1z = Approx1y';
      NDVIdata = load ('data\Anomalies\NDVIanomalies.mat');
      NDVIx = NDVIdata.NDVIanomalies;
      NDVIy = reshape(NDVIx,[1505,96]);
      NDVIz = NDVIy';
      for j = 1:1505
         m = lag(1,j)
         GRACEcrop = Approx1z(1:(96-m),j); 
         NDVIcrop = NDVIz((1+m):96,j);
            dlmwrite ('GRACECROPVEC.txt',GRACEcrop,'-append');
            dlmwrite ('NDVICROPVEC.txt',NDVIcrop,'-append');
      end
GRACEVEC = load ('GRACECROPVEC.txt');
NDVIVEC = load ('NDVICROPVEC.txt');
NEWGRACELAG = reshape(GRACEVEC,[96,1505]);

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

dpb 2015-12-15

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/260346-fill-blank-spaces-from-lagged-data-with-nan#answer_203327

编辑：dpb 2015-12-16

在 MATLAB Online 中打开

Instead of selecting the first N points,

GRACEcrop = Approx1z(1:(96-m),j);

just set the last m to NaN...

Approx1z(end-m+1:end,j)=nan;

and then write the modified array. No need for a new array at all (unless you have reason to also keep the unmodified in memory as well, of course).

2 个评论
显示无隐藏无

Robert 2015-12-15

Thanks for your answer, however, it didn't seem to work, it made all of the output NaN. I'm also not sure that I understand your reference to changing the last m to NaN, as m is only 1 number per loop (generally 0-6).