如何用上面的行替换连续的nans元素

Question

我有以下矩阵

A = [ 0    0  
      0    0 
      1   -1 
    NaN  NaN 
    NaN  NaN 
      0    0 
    NaN  NaN 
    NaN  NaN]

我想用它上面的行替换所有NaN行。对于上面的矩阵，这将是

A = [0  0 
     0  0 
     1  -1 
     1  -1 
     1  -1 
     0   0 
     0   0 
     0   0] 

Answer 1

一个想法是首先创建一个压缩矩阵B，它不包含NaN的行，然后再次将此矩阵展开为相同的长度作为原始矩阵：

mask = any(isnan(A), 2);
B = A(~mask, :);
result = B(cumsum(~mask), :);

Answer 2

我不确定这是否可以被矢量化......更简单的解决方案可能是一个循环：

newRow = [];
nans = isnan(A(:,1));
for ii = 1:size(A,1)

    if nans(ii)

        %# first row might be NaN -- skip it
        if ii==1
            continue; end

        %# for all other rows:
        if isempty(newRow) 
            newRow = A(ii-1,:); end

        A(ii,:) = newRow;

    else
        newRow = [];
    end

end