How do I convert to nan all entries with a particular value in a cell array?

Question

Garima Sharma 2017-6-14

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/344810-how-do-i-convert-to-nan-all-entries-with-a-particular-value-in-a-cell-array

评论： Star Strider 2017-6-15

Each entry of the cell array is a matrix, and I want to convert to nan all entries, say, of value=1. Something like this: a=cellfun(@(x) x(x==1)=nan, a, 'un',0)

I tried to do this by creating a separate function, but I need to delete different values from the matrix assigned to each cell in the cell array, so I'm not sure how that would work. Can I just define a cell function like this?

function out = replaceval(in, val)
    in(in==val) = NaN;
    out = in;
end

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Rik 2017-6-14

This should work as far as I can tell. Keep in mind that you might have to replicate your second input as a cell if it is not the same for the cell you are using as first input.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Star Strider 2017-6-14

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/344810-how-do-i-convert-to-nan-all-entries-with-a-particular-value-in-a-cell-array#answer_270740

在 MATLAB Online 中打开

I could not get this to work with only cellfun calls, and needed to use a loop.

This works:

C = {randi(9,5), randi(9,5), randi(9, 5)};          % Create Data
L = cellfun(@(x) x == 1, C, 'Uni',0);               % Logical Arrays
C1 = C{1}                                           % Display Initial Values
L1 = L{1}                                           % Display Logical Arrays
for k1 = 1:size(C,2)
    C{k1}(L{k1}) = NaN;                             % Replace With ‘NaN’
end
New_C1 = C{1}                                       % Display Replaced Values
C1 =
     2     4     9     3     5
     1     8     1     3     7
     4     5     2     3     4
     5     3     7     6     1
     6     5     7     6     8
L1 =
    5×5 logical array
     0   0   0   0   0
     1   0   1   0   0
     0   0   0   0   0
     0   0   0   0   1
     0   0   0   0   0
New_C1 =
     2     4     9     3     5
   NaN     8   NaN     3     7
     4     5     2     3     4
     5     3     7     6   NaN
     6     5     7     6     8

Remove the ‘Display’ lines. They exist only to demonstrate the results.

2 个评论
显示无隐藏无

Garima Sharma 2017-6-15

在 MATLAB Online 中打开

Actually, this worked for me. I just have to make sure that the 2 cell arrays from which in and val are pulled have the same dimension.

B=cellfun(@replaceval, a, b, 'un',0)  
function out = replaceval(in, val)
      in(in==val) = NaN;
      out = in;
end

Star Strider 2017-6-15

In my code, the logical cell arrays ‘L’ will by definition have the same dimensions as the matrices they are derived from. That is the advantage of calculating them in a separate step.

请先登录，再进行评论。