MATLAB Answers

# Bug in matrix indexing?

1 view (last 30 days)
Stewart Tan on 4 Sep 2019
Commented: Walter Roberson on 4 Sep 2019
So i have the matrix below, which are the DCT coefficients for an image.
test =
1.0e+03 *
1.1506 -0.0094 -0.0043 -0.0012 -0.0001 -0.0007 0.0003 0.0004
-0.0082 0.0004 -0.0012 0.0007 0.0017 -0.0003 -0.0011 0.0005
0.0008 -0.0002 -0.0015 0.0001 0.0004 0.0010 -0.0002 0.0014
0.0002 -0.0000 -0.0018 0.0010 -0.0018 -0.0005 0.0009 -0.0012
0.0001 0.0001 -0.0024 -0.0008 0.0004 0.0010 0.0011 -0.0001
-0.0012 0.0004 -0.0001 -0.0007 -0.0004 -0.0002 -0.0001 0.0009
0.0003 0.0002 0.0011 -0.0002 -0.0014 -0.0004 0.0000 -0.0004
0.0005 -0.0002 -0.0021 -0.0001 0.0007 0.0005 0.0008 -0.0007
and I'm splitting the matrix above into 4 equal parts since it's an 8x8 matrix. The first part would contain
1.1506 -0.0094 -0.0043 -0.0012
-0.0082 0.0004 -0.0012 0.0007
0.0008 -0.0002 -0.0015 0.0001
0.0002 -0.0000 -0.0018 0.0010
second part would be
-0.0001 -0.0007 0.0003 0.0004
0.0017 -0.0003 -0.0011 0.0005
0.0004 0.0010 -0.0002 0.0014
-0.0018 -0.0005 0.0009 -0.0012
etc..
So i managed to index out the first part, with.
N = 8;
test(1:N/2, 1:N/2)
which gave me the first part above. But suddenly, when i want to get the second part and the rest, I'm getting weird outputs as shown.
sec_part = test(1:N/2, N/2+1:N) %2nd part
sec_part =
-0.1250 -0.7400 0.3082 0.3531
1.6533 -0.3160 -1.0713 0.5453
0.3779 0.9798 -0.1831 1.3921
-1.8311 -0.4653 0.8852 -1.2236
I was supposed to get:
-0.0001 -0.0007 0.0003 0.0004
0.0017 -0.0003 -0.0011 0.0005
0.0004 0.0010 -0.0002 0.0014
-0.0018 -0.0005 0.0009 -0.0012
but I'm getting the random numbers as shown in sec_part. The same happened for the third part and fourth part. Why is it that only the first part is extracted correctly from the matrix?

#### 2 Comments

Stephen Cobeldick on 4 Sep 2019
"Why is it that only the first part is extracted correctly from the matrix?"
It isn't, all of them are extracted correctly. You need to pay attention to the display format of your original matrix, which starts with this:
1.0e+03 *
...
Stewart Tan on 4 Sep 2019
@Stephen Cobeldick thanks, i have overlooked that display format.

Sign in to comment.

### Accepted Answer

Walter Roberson on 4 Sep 2019
You missed part of the output. Notice you have
test =
1.0e+03 *
That 1.0e+03 means that the part displayed below all has to be multiplied by 1000 to get the actual numbers. So for the -0.0001 entry displayed, the real value stored is roughly -0.1 -- with -0.1250 being entirely consistent with that possibility.
I recommend that you stop using "format short" and start using "format short g" or "format long g"
There is a Preference you can set to change the default output format.

#### 2 Comments

Stewart Tan on 4 Sep 2019
Is Matlab’s default short?
Walter Roberson on 4 Sep 2019
Yes, the default is short.
Preferences -> Command Window -> Text Display -> Numeric format
This preference controls what you get when you start up MATLAB; you can change the active format at any time using a format command.

Sign in to comment.

### More Answers (1)

KSSV on 4 Sep 2019
USe this:
test = 1.0e+03 *[1.1506 -0.0094 -0.0043 -0.0012 -0.0001 -0.0007 0.0003 0.0004
-0.0082 0.0004 -0.0012 0.0007 0.0017 -0.0003 -0.0011 0.0005
0.0008 -0.0002 -0.0015 0.0001 0.0004 0.0010 -0.0002 0.0014
0.0002 -0.0000 -0.0018 0.0010 -0.0018 -0.0005 0.0009 -0.0012
0.0001 0.0001 -0.0024 -0.0008 0.0004 0.0010 0.0011 -0.0001
-0.0012 0.0004 -0.0001 -0.0007 -0.0004 -0.0002 -0.0001 0.0009
0.0003 0.0002 0.0011 -0.0002 -0.0014 -0.0004 0.0000 -0.0004
0.0005 -0.0002 -0.0021 -0.0001 0.0007 0.0005 0.0008 -0.0007] ;
[m,n] = size(test);
k = [4 4]; % you want 4*4 matrix
C = mat2cell(test,k(1)*ones(m/k(1),1),k(2)*ones(n/k(2),1));

#### 0 Comments

Sign in to comment.

Sign in to answer this question.