CODEWITHSUNDEEP
matlab
varun vyas
varun vyas

Reputation: 1

Using Negative Values in Matlab

In order to find best fit (thru polyfit), i am getting negative p value but matlab is not accepting it (Subscript indices must either be real positive integers or logicals.). Is there any way that I can use it ? I can't think of alternative method. I'll always get negative values.

EDIT:

I am trying to flattening baseline of a curve, for that. I am running for loop to have fit from 1 to 3 order. And then I am using smallest normr s value to to find the best fit and then subtract it from the whole curve to get baseline straight. I tried with few curves it works well but not with all of the data because of above describes issue.

part of the code I am working on:

for i=1:3 
  [p,s]=polyfit(x,y,i); 
  a=s.normr; 
  b(i,1)=p(1); 
  normr(i,1)=a; 
  ind=find(b==min(b)); 
  mn=b(ind,1); 
  Yflat=y-mn(1)*(x-mean(x)); 
  ca{2,2}=Yflat; 
  clear a b normr p s rte ind ind2 Yflat 
end

Upvotes: 0

Views: 3444

Answers (2)

marsrover
marsrover

Reputation: 432

When I translate an image into negative coordinates,

I usually record an offset e.g. offset = [ -5, -8.5 ] and save the intensity values in matrix begin with (1, 1) as usual,

But when comes to calculation, let the coordinates array add up with the offset

e.g. [ actualX, actualY ] = [ x, y ] + offset ;

It may need extra efforts, but it works.

Good Luck!

Upvotes: 1

Pursuit
Pursuit

Reputation: 12345

The code below (your code from the comments + initialization of x, y) executes. What is the problem?

x = 1:50;
y = randn(size(x));

for i=1:3
    [p,s]=polyfit(x,y,i);
    a=s.normr;
    b(i,1)=p(1);
    normr(i,1)=a;
    ind=find(b==min(b));
    mn=b(ind,1);
    Yflat=y-mn(1)*(x-mean(x));
    ca{2,2}=Yflat;
end

Upvotes: 0

Related Questions