indices to column vectros of a matrix using “crossvalind”

i got a 3x50 matrix. i would like to obtain the 50 3x1 column vectors and assign to them indices using the crossvalind command seen above

 Indices = crossvalind('Kfold',N, 10);

so lets hypotheticaly say that i had a 3x3 matrix named "input" :

input=[1 2 3;4 5 6;7 8 9]

how can I imply an index generation for the vector columns of my "input" matrix using crossvalind?

 A=[1;4;7]
 B=[2;5;9]
 C=[3;6;9]

**the reason i wanna do it is so i can after apply cross validation while having as train/test set these column vectors

Answer 1

[Azzi Abdelmalek]

A=input(:,1)

[Azzi Abdelmalek]

laplace laplace commented

thanx for answering but i got a feeling that you missunderstood my question :))

[Azzi Abdelmalek]

What are your feeling about this?

A=[1;4;7]
Indices = arrayfun(@(x) crossvalind('Kfold',x, 10),A,'un',0);

[laplace laplace]

can you explain me the command you wrote :D maybe its correct and i am doing smth wrong :(

Answer 2

[laplace laplace]

inputs is the 3x50 matrix

    C = num2cell(inputs,1);
    Indices = cellfun(@(C) crossvalind('Kfold',C, 10),C,'un',0);

Indices =

Columns 1 through 9

    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]

Columns 10 through 18

    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]

Columns 19 through 27

    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]

Columns 28 through 36

    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]

Columns 37 through 45

    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]

Columns 46 through 50

    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]    [3x1 double]

what i wanted to get was just indices for each column vector... any thoughts?

Answer 3

[laplace laplace]

noone?:D