Error while using kstat2 function: Out of range or non-integer values truncated during conversion to character.

I am using Kolmogorov-Smirnov test to compare 2 datasets. [h,p] = kstest2('Y1','Y2')

Y1 and Y2 are my datasets. They are DataMatrix format from a microarray(generated with matlab's affyrma function) Each contains a single column of data with thousands of rows. Each row has a row heading and the column has its own heading.

The values in each dataset are mostly from -3 to 3, possibly some NaN's in there.

I obtain this error: Out of range or non-integer values truncated during conversion to character. > In kstest2 at 152

To try and figure out what's causing this I've made some mock vectors w/NaN's as well as difference number combinations and the kstest2 function has worked fine in those tests.

Any idea what it means and what is the cause?

[the cyclist]

Could you post a small, self-contained example that we could run, that actually exhibits the error? (Sometimes when trimming the problematic code down to its essence also exposes the issue.)

[Vanessa]

Here is me trying to use this function in the command window.

First I selected 10 rows of my 2 DataMatrix variables that I want to test. Then, I tried 3 alternative ways to enter the ktest2 function. Now that I reconsider it, I should NOT have put the variables into quotes (the first attempt at ktest2). However when I try to run the ktest2 on what I think should be the correct inputs (either #2 or #3) I get a new error (DataMatrix subscripts must be 2 dimensional). I am not sure how to interpret this. I've read over the Matlab page about DataMatrix formats but it doesn't mention this error.

>> mini_set_X1

mini_set_X1 =

                030514JB02
    10700001    -0.0053592
    10700003      -0.12031
    10700004      -0.13406
    10700005      -0.24628
    10700013      -0.06817
    10700014      -0.16459
    10700020     -0.057178
    10700027      -0.61875
    10700029      0.034743
    10700040      -0.15711

>> mini_set_X2

mini_set_X2 =

                022714JB14
    10700001    0.52189   
    10700003    0.87869   
    10700004    0.49216   
    10700005     1.1703   
    10700013    0.71538   
    10700014     1.1207   
    10700020    0.47076   
    10700027     0.9894   
    10700029    0.15174   
    10700040     1.0981

>> [h,p] = kstest2('mini_set_X1', 'mini_set_X2') Warning: Out of range or non-integer values truncated during conversion to character. > In kstest2 at 152 Warning: Out of range or non-integer values truncated during conversion to character. > In kstest2 at 152

h =

     0

p =

    1.0000

>> [h,p] = kstest2(mini_set_X1(:,:), mini_set_X2(:,:)) Error using bioma.data.DataMatrix/subsref (line 45) DataMatrix subscripts must be two-dimensional.

Error in kstest2 (line 101) x1 = x1(~isnan(x1));

>> [h,p] = kstest2(mini_set_X1, mini_set_X2) Error using bioma.data.DataMatrix/subsref (line 45) DataMatrix subscripts must be two-dimensional.

Error in kstest2 (line 101) x1 = x1(~isnan(x1));

[Vanessa]

I found a similar question (unanswered) from Feb 2013 http://www.mathworks.com/matlabcentral/answers/63237-how-do-i-use-the-knnimpute-command-with-a-datamatrix-object

Datasets from microarrays is in the DataMatrix format so how to use various Matlab functions with DataMatrix might be the actual issue. (Or how to convert the data into a useable format?)

[Vanessa]

In addition to kstest2, many of the other functions I was trying to use on my data did not work. They all give the same error: (Error using bioma.data.DataMatrix/subsref (line 45) DataMatrix subscripts must be two-dimensional.)

For the time being, I found I can use the double function to access the data so I can use functions. (See below).

If there is a better way, please advise.

>> mini_set_X2

mini_set_X2 =

                022714JB14
    10700001    0.52189   
    10700003    0.87869   
    10700004    0.49216   
    10700005     1.1703   
    10700013    0.71538   
    10700014     1.1207   
    10700020    0.47076   
    10700027     0.9894   
    10700029    0.15174   
    10700040     1.0981

>> data2=double(mini_set_X2(:,:))

data2 =

    0.5219
    0.8787
    0.4922
    1.1703
    0.7154
    1.1207
    0.4708
    0.9894
    0.1517
    1.0981