Extracting the maximum value from each array of different dimensions in a cell?

Question

Niraj Bal Tamang 2021-9-23

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1459629-extracting-the-maximum-value-from-each-array-of-different-dimensions-in-a-cell

评论： Voss 2021-9-27

X.mat

I have a cell (X) with arrays of different dimensions. I want to create a new array containing only the maximum values of each array of the cell. I tried Y=[X{:}]; but it didn't work due to different dimensions of the arrays. Can anyone please help me to exract the maximum value from each array in the cell and create a new variable out of the extracted values.

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

Voss 2021-9-23

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1459629-extracting-the-maximum-value-from-each-array-of-different-dimensions-in-a-cell#answer_793894

在 MATLAB Online 中打开

Look into cellfun. You can likely do this:

X_max = cellfun(@(y)max(y(:)),X);

2 个评论
显示无隐藏无

Niraj Bal Tamang 2021-9-23

Thank you. But it didn't work. The cellfun worked when the dimensions of the array within the cell was same. Previously i tried Y=[X{:}]; which listed the elements of each array into cell columns. Then i applied the cellfun to Y. But in this case, it didn't work as the dimension was different for the arrays.

Voss 2021-9-27

在 MATLAB Online 中打开

Since each element of X is itself a cell array, you have to use cellfun twice. Try something like this:

X_max = cellfun(@(x)max(x(:)),cellfun(@(x)[x{:}],X,'UniformOutput',false),'UniformOutput',false);

Then X_max is a cell array containing the maximum value of each set of arrays forming an element of X, with empty entries where X{i} is a cell array with all empty vectors.

If you want X_max to contain some scalar value, say NaN, where X{i} is a cell array of all empty vectors, you can do something like this next:

X_max(cellfun(@(x)isempty(x),X_max)) = {NaN};

Then you can make X_max a numeric array if you want:

X_max = [X_max{:}];

请先登录，再进行评论。