how to delete particular values in this matrix ?

 [a b]=[     0.5708    0.0755         0         0         0         0         0
                  0         0  161.5569         0   84.9907   35.0193   17.0669];

i don't want the values before this

" 0
  0 "

and my answer should be like this..

 [c d]=[ 0         0         0         0
         0   84.9907   35.0193   17.0669];

[Image Analyst]

What do you mean by [a b] and [c d] - that is not MATLAB syntax. Also, do you know about the all() function?

[Matlab111]

Image Analyst- After running my loop i will get "[a b]" and similarly i'm storing this values in "[c d]" that's it..sir

[Stephen23]

But you are giving us code that does not work, and makes no sense in MATLAB:

>> [a b]=[     0.5708    0.0755         0         0         0         0         0
                  0         0  161.5569         0   84.9907   35.0193   17.0669];
Too many output arguments.

Image Analyst is correct: this is not MATLAB syntax. You can assign your vector of values to only one variable (at a time). Why are you trying to allocate these vectors to both a and b , and c and d ?

[Matlab111]

Stephen Cobeldick- just run this code that i attached below, you will get the values.."[allR11;allR21]" this is the matrix that i want to modified.

[Michael Haderlein]

That's quite a lot of code for this question (and most of the code has nothing to do with the question). What is the criterion? Should everything be deleted until a column of zeros? Should the first three columns be deleted? Or is it something else?

case 1: (what Image Analyst was hinting at)

a=[ 0.5708    0.0755    0           0   0         0         0;
    0         0         161.5569    0   84.9907   35.0193   17.0669];
allzeros=find(all(a==0));
b=a(:,allzeros:end);

case 2:

b=a(:,4:end);

If none of these cases are what you're looking for, please give detailed information on what you need and post only relevant code.

[Matlab111]

Michael Haderlein- sir, don't go for code just see below i given some example.

question:

 c=[0.7893    0.8337    0.1479         0         0     0.1479       0.9993];

1.now i should delete the repeated values in that 'c'.

2.i should delete the values that is displayed before the zeros.

3.And finally i should delete zeros also.

Expected output:

 d=[0.9993];

Answer 1

[Stephen23]

To solve your (constantly changing) requirement stated in your last comment to my original answer, try this code:

allR111 = [allR11;allR21];
Y3 = true(1,size(allR111,2));
Y3(1:find(abs(allR111(1,:))<1e-6,1,'last')) = false;
Y3(1:find(abs(allR111(2,:))<1e-6,1,'last')) = false;
allR111(:,Y3)

[Matlab111]

Stephen Cobeldick- really thanks sir, you helped me a lot

Answer 2

[Stephen23]

Try this:

>> c = [0.7893,0.8337,0.1479,0,0,0.1479,0.9993];
>> X = sum(bsxfun(@(a,b)abs(a-b)<1e-6,c(:).',c(:)))<2;
>> X(1:find(c==0,1,'last')) = false;
>> c(X)
ans =  0.9993

As in my answer to your other related question , note that I did not use equality test == or unique with floating point numbers, but instead compared the difference of two values with some tolerance, in this case 1e-6. You can change the tolerance to best suit your problem.

[Matlab111]

Stephen Cobeldick- sir, just see my code that i attached below, just go for last line no-395 that i have highlited in that i modified according to your case, as you said above, in that i'm getting for only 1st case for " allR11 " and i want to get similarly for "allR21" also and i tried but i'm not getting.

Note: sir, just type this in command window and see

1. "allR11" -this is 1st case. (i'm getting for this case)

2. "allR21" -this is 2nd case. (i'm not getting for this case)

sir, if not understanding clearly just ask me once again..

[Stephen23]

You will have to explain that again. I ran your code, and got some vectors for allR11 and allR21. The code I proposed works: the indices removed all repeated values and all values up until and including the last zero. This is the expected behavior of my code, and it is what you requested in your comment to your original question.

What is happening that you do not expect?

Are you expecting to perform this on a matrix with two rows, and remove any column that contains a duplicate value in its own row, or is in front of any zeros?

Currently I have provided code that does exactly what you requested, so you will have to explain your exact situation in more detail if this does not resolve your problem.

[Stephen23]

PS: are you planning on accepting any answer to your last question?:

http://www.mathworks.com/matlabcentral/answers/176641-how-to-delete-an-repeated-values-in-matrix

[Matlab111]

Stephen Cobeldick- sir go for this one " Are you expecting to perform this on a matrix with two rows, and remove any column that contains a duplicate value in its own row, or is in front of any zeros? "

yes sir i'm expecting to perform this on a matrix with two rows,

1. i want to remove duplicate values.

2. i want remove the values that is displayed before the zeros.

3. i want to delete the zeros also.

Note: By taking matrix with two rows, you will be getting like this

 ans=[ 0.6902 0.1634 0.6987   0  124.7462   20.0765   23.4781  194.6180
         0    0         0     0  171.3741  147.7809   46.3979  129.3696]

and i'm expecting output is like this

 ans=[124.7462   20.0765   23.4781  194.6180
      171.3741  147.7809   46.3979  129.3696]

sir,just run this code you will get only "[allR11;allR21]" this that i want to modify

still if you are not clear you can ask me once again...

[Stephen23]

You can just repeat the method that I gave in my original answer for each row, and use the indices together:

>> A = [0.6902, 0.1634, 0.6987, 0, 124.7462, 20.0765, 23.4781, 194.6180; 0, 0, 0, 0, 171.3741, 147.7809, 46.3979, 129.3696];
>> X = sum(bsxfun(@(a,b)abs(a-b)<1e-6,A(1,:).',A(1,:)))<2;
>> Y = sum(bsxfun(@(a,b)abs(a-b)<1e-6,A(2,:).',A(2,:)))<2;
>> X(1:find(A(1,:)==0,1,'last')) = false;
>> Y(1:find(A(2,:)==0,1,'last')) = false;
>> A(:,X&Y)
ans =
       124.75       20.076       23.478       194.62
       171.37       147.78       46.398       129.37

No repeated values and with all terms up until the final zero removed.

I suspect that your code might actually be a lot neater and faster if you do not keep splitting your data up into smaller variables, but keep it together in one array, and just work with indices. In my experience this is often a much faster and more reliable way to manage data.

[Matlab111]

Stephen Cobeldick- sir, i try your logic with my main code ya i'm getting but, actually i want to get like this, just see below... when "r=0" i'm getting correct values that's what i expected but when "r=1" i'm not getting an expected values.

after applying your approach that you given above

and i want the answer should be like this, from column 40 to 71, no problem if it's a repeated values and i want get only that red marked part

Note: i'm getting for correct answer when "r=0", no problem if the values repeated and i have attached my modified code below..

if you not clear just ask me once again.

[Stephen23]

I have replied to this in a new answer.