Replacing numbers around a specific index(CAN'T FIGURE IT OUT)

I am trying to create a sub-function for a EEG analysis program. Basically the data is a 1x892 matrix of data with numbers ranging from 1 to 99. This portion that i am trying to deal with has two parts:

1. Go through the data and find the value 9, whenever it occurs in the 1x892 matrix. I have already done this part.

2. Replace the numbers surrounding the identified number 9. So if there was data with 9 in it follows, [... a b c d 9 e f g h i j k...]. Basically after the part of identifying the number 9, the letters a through d need to be replaced with 5,6,7,8 respectively. The letters e through k need to be replaced with 10,11,12,13,14,15,16, respectively. I am struck on this part and i would really appreciate any help.

P.s, the way this data is arranged, that we get from an EEG, a 9 will not overlap with another 9.

[Stephen23]

@Sanwal Yousaf: Why did you delete your other identical question, where I had already provided you with a simple working example of how you can achieve this?:

http://www.mathworks.com/matlabcentral/answers/232251-replacing-numbers-around-a-target-number-9-to-range-from-5-to-16

I notice that you paid attention to the questions that I asked, as these are now explained in your new question.

Did my answer not work on your computer?

[Sanwal Yousaf]

Yeah i saw the comment that you posted and looked through the data again to make sure that there were no recurring 9's but that was the only comment that i could see, i didnt see any other comment and i have been stuck on this for almost a week, so i thought it might be worth reposting.

[Sanwal Yousaf]

I checked to see if there was any comment with any hint to an answer. The only thing that i saw was your comment, which did make me look at my data in a new light for which i am endlessly grateful. A new outlook is always welcome! Cheers

[Stephen23]

Question is continued here:

http://de.mathworks.com/matlabcentral/answers/232666-eeg-data-tried-three-different-ways-and-end-up-with-same-subsindex-error

Answer 1

[Star Strider]

This seems to work:

x = randi(99, 1, 200);                                  % Create Data
ix9 = find(x == 9);
for k1 = 1:length(ix9)
    if (ix9(k1)+7) <= length(x) && (ix9(k1)-4) >= 1     % Check For Out-Of-Range Indices
        ixrng = ix9(k1) + [-4:7];                       % Calculate Index Range
        x(ixrng) = 5:16;                                % Substitute
    end
end

[Sanwal Yousaf]

Error using  : 
Too many output arguments.

Error in FormatEEGarry (line 25)
        [x.event(ixrng).epochtype] = 5:16;

I am getting this error for assignment of 5:16 to the data set. Any ideas why??

[Star Strider]

That’s not a line from my code, so I have no idea.

I would use a function form of my code (saving it in your MATLAB path as a separate function file called EEGsubstitute.m and do the substitution as a separate step:

function EEGout = EEGsubstitute(EEGin)
EEGout = EEGin;
ix9 = find(EEGout == 9);
for k1 = 1:length(ix9)
    if (ix9(k1)+7) <= length(EEGout) && (ix9(k1)-4) >= 1     % Check For Out-Of-Range Indices
        ixrng = ix9(k1) + [-4:7];                            % Calculate Index Range
        EEGout(ixrng) = 5:16;                                % Substitute
    end
end
end

Then call it as:

EEGout = EEGsubstitute(EEGin);

You will likely have to assign ‘EEGin’ from your original data or your structure first.

Note that you specified a row vector in your Question, so my code assumes the data will always be a row vector.

Answer 2

[Stephen23]

It can be done very easily by converting the integers to char and using regepxrep:

>> A = repmat(99:-1:1,1,9); % fake data vector
>> Z = +regexprep(char(A),sprintf('.{0,4}%c.{0,7}',9),char(5:16));
>> A(82:102)
ans =
   18   17   16   15   14   13   12   11   10    9    8    7    6    5    4    3    2    1   99   98   97
>> Z(82:102)
ans =
   18   17   16   15   14    5    6    7    8    9   10   11   12   13   14   15   16    1   99   98   97

[Sanwal Yousaf]

I just tried this solution. It changes the numbers surrounding the 9 to 10, 11, 12, the three succeeding numbers. Would you know why that would be. You can check my full code here. Maybe i am making a mistake that i am blind to.

http://www.mathworks.com/matlabcentral/answers/232666-eeg-data-tried-three-different-ways-and-end-up-with-same-subsindex-error

A new outlook on the code would be appreciated.

[Sanwal Yousaf]

So the data becomes something like this,

1 9 10 11 12 1 1 1

[Stephen23]

Most likely this occurs because you have some nines that are too close to each other, so that the replaced values overlap: in this case the output vector will be of a different size to the input vector and some values will be repeated. This is why I asked in your original question (that you deleted) specifically about the overlap of the nines: how close can they be?

Unfortunately you do not provide any sample data for us to try, so it is impossible for me to know why this happens with your data sets. I am happy to invent my own data set, but this is likely going to have quite different properties to the data that you are working with. Indeed you can see that I invented a data vector where the nines are spread very far apart, so there in no problem with overlap this case. If there is some overlap, then you need to define how the output is specified: do the leading or trailing digits have priority, or do we ignore overlapping nines?

I am happy to help you and show you how to make this work, but I will need your data to try out.