How to find max and min values from the cell(x*1 matrix) array elements !!

hallo everyone, lets say, cell array have 600*1 matrix in that each cell consists of the X*1 matrix. now i want to find a max and min values. i tried to use min and max functions but these are not useful for cell arrays. so i try to find without inbulit functions. this is my code:-

     d_max = d_main(1,1);
     d_min= d_main(1,1);
     for m = 1:length(mydata)
     d_main = mydata{m,2};
       for q = 1:length(d_main(1:end,1))
            if d_main(q,1) > d_max
                d_max = d_main(q,1);
              elseif d_main(q,1)< d_min
                  d_min = d_main(q,1);
             end
       end
     mydata{m,3} = d_max;
     mydata{m,4} = d_min;
     end

here, the problem is i am getting few rows are correct max numbers and remaining all are wrong. and all min values are wrong. i tried max and min function in command window.

[Jan]

Note: Replace length(d_main(1:end,1)) by size(d_main, 1).

What do you want to get as result? The min and max value for each cell or over the elements of all cells or between the vectors of all cells? A small example would be useful.

What is d_main? The code does not run in the posted form:

d_max = d_main(1,1);
d_min = d_main(1,1);
for m = 1:length(mydata)
   d_main = mydata{m,2};

Here d_main is used before the definition.

[Ram]

@jan thanks for your reply. Yes, d_main is used before in my code. it has the derivative values which has 1000*1 matrix and cells are 600*1. i want to find max and min from those derivative values for each cell. when you observe it has -inf to +inf range. when i replaced to size(d_main,1) also getting incorrect values. for example.-

 max(mydata{415, 1}  )
 ans =
 0.2773

but appearing ans is 0.4364

[Jan]

My note was just a general hint. size(d_main, 1) is nicer and more efficient than length(d_main(1:end,1)), which creates a temporary vector 1:end only to create the temporary vector d_main(1:end,1) only to measure its length. <size< can solve this directly.

Answer 1

[KSSV]

%%make some random data
N = 5 ;
C = cell(N,1) ;
for i = 1:N
    C{i} = rand(randperm(100,1),1) ;
end
% GEt min and max
themin = zeros(N,1) ; themax = zeros(N,1) ;
for i = 1:N
    T = C{i} ;
    ma = T(1); mi = T(1);
    for j = 2:length(T)
        if ma < T(j), ma = T(j);
        elseif mi > T(j), mi = T(j);
        end
    end
    themin(i) = mi ; themax(i) = ma ;
end
%%C0mpare 

[themin cellfun(@min,C)] 
[themax cellfun(@max,C)]

[Ram]

@KSSV thanks alot for your help. i got correct output.

[Ram]

@KSSV, do you have any idea... i am getting error, comment thanks in advance

Answer 2

[Jan]

The problem is, that you do not reset the minimal and maximal values inside the loop. An improved version:

 for m = 1:length(mydata)
   d_main = mydata{m,2};
   % Now ATFER d_main is copied from mydata, not before the loop:
   d_max = d_main(1);    % <-- move this INSIDE the loop
   d_min = d_main(1);    % <-- move this INSIDE the loop
   for q = 2:size(d_main, 1)  % You can start at 2
      if d_main(q) > d_max
         d_max = d_main(q);
      elseif d_main(q)< d_min
         d_min = d_main(q);
      end
    end
    mydata{m,3} = d_max;
    mydata{m,4} = d_min;
  end

But KSSV's method is much better:

mydata(:, 3) = cellfun(@max, mydata(:,2), 'UniformOutput', false);
mydata(:, 4) = cellfun(@min, mydata(:,2), 'UniformOutput', false);

Or:

data2        = cat(2, data{:, 2});
mydata(:, 3) = num2cell(max(data2, [], 2))
mydata(:, 4) = num2cell(min(data2, [], 2))

[Ram]

@jan thanks alot, after keeping those min and max inside loop i got the correct ans. i have this max and min values lets say these are y values. i want to find x values(from my initial cell set, each cell consits 5000*2) w.r.to the y values. should we find w.rto (max and min)_ index value of deri then find intial cell set index value*. because we can't find this max and min values inital cell set. Is there another approach to get where max and min point coordinates._

[Jan]

@Ram: I cannot follow your explanations. Can you provide a small example?

[Ram]

@Jan i want to find(x-value) where max point (is y-value) occurs. here, i have all max points of my cell array 500*1 cell matrix(after derivative cell matrix). with single element(max value) i can't plot it. so i want to find an x-value where max value occurs. for example:- max values are 1,2,3,4,5. now i want to find where 1 occurs on x-axis similarly all other max points. can i find with my before derivative cell matrix which 500*1 and each cell consists has 1000*2. i hope you understand it.

[Jan]

@Ram: Sorry, I still do not understand it. Again: Please provide a small example for the input data and the wanted outputs. Use 2x1 cell and 3x2 arrays to keep it small.

[Ram]

@Jan. actually i want to find index's of each max and min values. After this index values should be compare to before derivative my cell matrix(2*2 matrix in each cell). for example, i have max value is 5 which has index is 2 (from d_main). later, i am finding index 2 value in intial matrix(2*2 =x*y matrix lets say). when i compare this index value i got two values in that matrix(x*y) which has 10,11. but here i extract only x(10, index is 2) values. similarly, for all min values.

my new code is,

    for m = 1:length(mydata)
    [M,I]=max(mydata{m, 2});
    mydata{m,3} = [M,I];
    similarly for min values.
comparing index values but getting error!
    for w =1:length(mydata)
    index1 = mydata{w,3}(:,2)==mydata{w,1}(:,1);
    index2 = mydata{w,4}(:,2)==mydata{w,1}(:,1);
    mydata{w,1}(:,1) = mydata{w,5};
    mydata{w,1}(:,1) = mydata{w,6};
    end

how can i compare these index values to another cell matrix index and extract those first coulmn value only in each cell matrix.