Matrix regression- how to make?

Hi . I am completely new in Matlab and programming in general. I am trying to find a polynomial which would convert values from my image data to the mean value of this array. right now I have something like that:

   A= importdata( '1slide.tif');
  % Rewrite the image in matrix
  B= im2uint16(A);
  % Find the mean value
  M=mean2(B);
  E= ones (2048,2048,'uint16');
  Mat = M *(E);
  Poly= polyfit(B,Mat,6);
  A= importdata( '1slide.tif');
  % Rewrite the image in matrix
  B= im2double(A);
  % Find the mean value
  M=mean2(B);
  E= ones (2048,2048,'double');
  Mat = M *(E);
  Poly= polyfit(B,Mat,6)

But I got an error:

Undefined function 'qr' for input arguments of type 'uint16'.Error in polyfit (line 66)[Q,R] = qr(V,0);

Why do I need these QR? What should I use for them?

Answer 1

[Matt J]

The input arguments to polyfit are not supposed to be matrices and they are not supposed to be of integer type. Polyfit will perform a single 1D polynomial fit to a vector of x,y data (singles or doubles) that lie on the polynomial.

To better advise, we'd have to know what you're expecting by passing matrices to polyfit instead of vectors.

[Nata Salt]

Thank you, I did not know that I can not use matrix... If there is another function to find the polynom which converts one matrix value to another one?

[Nata Salt]

May be a loop? for each matrix raw?

[Matt J]

which converts one matrix value to another one

I think you'll need to define for us what that means.