following is the error while calculating euclidean distance using pdist2, what should i do ..

testdata =

3×1 cell array

    [54×6 double]
    [44×6 double]
    [44×6 double]

>> traindata

traindata =

8×1 cell array

    []
    []
    []
    [50×6 double]
    [47×6 double]
    [38×6 double]
    [31×6 double]
    [43×6 double]

>> pdist2(testdata,traindata); Warning: Converting input data to double. > In pdist2 (line 232) Error using cast Conversion to double from cell is not possible.

Error in pdist2 (line 234) X = cast(X,outClass);

Answer 1

[KSSV]

You need to convert your cells into matrix and then call _pdist2).

testdata = cell2mat(testdata) ;
traindata = cell2mat(traindata) ;

pdist2(testdata,traindata);

[Balaji M. Sontakke]

Thank you

[Balaji M. Sontakke]

Sir, in cell array dimension of row is vari i.e. somewhere

    [50×6 double]
    [47×6 double]
    [38×6 double]
    [31×6 double]
    [43×6 double]

can I fix the rowsize by function? like

[20X6]
[20X6]

[Walter Roberson]

Consider the first block, 50 x 6. The first 20 rows would be processed, then the second 20 rows would be processed. That leaves 10 unprocessed rows from the first block. Should those 10 rows be dropped, or should the first 10 rows from the 47 x 6 be taken to finish the group of 20?

[Balaji M. Sontakke]

Sir, actually the block[50 X 6] contains 'minutiae' information which contains the following data in columns [X, Y, CN, Theta, Flag, 1]and rows contains number of minutiae information on the dorsalhandvein. if i go like this the matrix size increase, so I want to reduce minutiae information which is in row. sir is there any function to reduce row size.

[Walter Roberson]

reduced_testdata = cellfun(@(M) M(1:min(end,20), :), testdata, 'uniform', 0);

[Balaji M. Sontakke]

Thank you sir...