How to supplement the matrix?

Hello! I cleaned my signal from emissions, it is in the form of a matrix, but there I used i-1, in consequence of which, instead of 1100, I now have 1099 values, how can I add the last value?\

for i = 1:length(d)-1  % Error 1100    1099
      if g(i)-50>= g(i+1);
      f(i)=g(i);
      end
end

[Stephen23]

Your question is not very clear, but possibly you should preallocate the output array:

https://www.mathworks.com/help/matlab/matlab_prog/preallocating-arrays.html

[Lev Mihailov]

Sorry, I do not quite understand, I need a condition

if c (i) + 10 > size (Matrix, 1);

before that, I simply used 'r' which was 1100, now I’m using 'c' which is 1099, and the matrix itself is 1100

Error Index exceeds matrix dimensions.

[Image Analyst]

How do c and Matrix relate to f and g?

DId you try Stephen's suggestion of putting

f = zeros(size(g));

before your original for loop?

[Jan]

@Lev: I do not see the relation between

if g(i)-50>= g(i+1);

and

if c (i) + 10 > size (Matrix, 1)

Which problem do you want to solve? What about:

f = g(diff(g) < -50)

instead of the loop?

[Lev Mihailov]

Cycle loop has nowhere

f = zeros(size(g)); 
% why do i need to convert to zeros?

The problem removed the outliers from the signal was 1099 values ​​(these are the positions of the maxima), but in the matrix there are 1100 values

[Jan]

@Lev: |zeros()< does not convert anything to zeros. Please read the documentation:

doc zeros

to find out, that this is a pre-allocation of the output. Letting an array grow iteratively is extremely inefficient. Matlab has to allocation a new array in each iteration. So:

r = [];
for k = 1:1e6
    r(k) = rand;
end

needs 8*1e6 bytes finally, but intermediately 8*sum(1:1e6) bytes are allocated: 4 TeraByte!

I do not understand the meaning of your comment. If you detect outliers by subtracting pairs elements, of course there is one element less in the output.

Answer 1

[dpb]

Adding to Stephen's answer Matlab syntax, the crystal ball says

for i = 1:length(d)-1  % Error 1100    1099
  if g(i)-50>= g(i+1);
    f(i)=g(i);
  end
end

becomes

f=zeros(size(g));
ix=diff(g)<-50;
f(ix)=g(ix);

Answer 2

[Lev Mihailov]

Still nothing happened, gives an error

Index exceeds matrix dimensions.

[Lev Mihailov]

I have values ​​in an array, can't I just add a value like 1100?

[dpb]

1. We can't see your terminal -- an error out of context with code that produced it is meaningless...paste the entire red text and the preceding code line that produced the error.
2. You can always augment an array, yes, but you have to do so by a form that is allowable catenation, not by reference outside an existing array bounding.

The code I gave cannot produce an indexing array bounds error...and will have resulting array f of same size as g

[Lev Mihailov]

[d,g]=max(Matrix);
for chek = 1:length(d)-1
  if g(chek )-50>= g(chek +1);
    c(chek )=10;
  else g(chek )+50 <= g(chek +1);
    c(chek)=g(chek );  
  end
end


for i = 1:length(d)
if c(i)+10>size(Matrix,1);
    aan=Matrix(c(i):end,i)  ; 
    aa=Matrix((c(i)-10:c(i)),i) ;
    У{i}=aa(aa ~= 0);
    Х{i}=aan(aan ~= 0);
end

here is the full code

Index exceeds matrix dimensions.

Error in Senteb2019 (line 260)

if c(i)+10>size(Matrix,1);

[dpb]

c can only have length(d)-1 elements in it...as has been said multiple times before, preallocate it to the size wanted or compensate that there will always be one less array element since you're dealing with diff() functionality that produces one less element by its very nature.

[dpb]

In

for chek = 1:length(d)-1
  if g(chek )-50>= g(chek +1);
    c(chek )=10;
  else g(chek )+50 <= g(chek +1);
    c(chek)=g(chek );  
  end
end

the else clause should be elseif it appears.

What is the value for c if abs(diff(g))<50?

Again, use Matlab array syntax; the above loop construct (presuming the "elseif" and zero for then missing "else" clause)

c=zeros(size(g));         % preallocate to same size as g array
ix=diff(g)<-50;
c(ix)=10;
ix=diff(g)>=50;
c(ix)=g(ix);

You need to define what happens in the other cases not satisfied...if they're to remain unchanged, then use

c=g;

for initialization instead of zeros