need help with function and lsqcurvefit please

Hi there! :) I'm still learning how to program, and since i dont have any teacher i need some help guys, please.. I've done a few interesting programs with matlab, but still, i run away from functions, because i still can't understand it... In this assignment I need to find a fit for some data. I need a funtion to calculate a sort of things with input parameters being my ydata (time2) and a lot of constants (T, P, e, K, Y, w) that must be found for the curve to fit the real data. The function must be something like this:

   function [RV] = RVfun(time2, T, P, e, K, Y, w)

   M1 = 2*pi*(time2 - T)/P;

   E(:,1) = M1 + e*sin(M1) + (e^2/2)*sin(2*M1);
   M(:,1) = E(:,1) - e*sin(E(:,1));

   for i = 2:1000
       E(:,i) = E(:,i-1) + ((M1 - M(:,i-1))./(1 - e*cos(E(:,i-1))));
       M(:,i) = E(:,i) - e*sin(E(:,i));

       if max(abs(M(:,i) - M1)) < 10^-13          
         Mean = M(:,i);
           Ee = E(:,i); 
           break
       end
   end

   ta = 2*atan(sqrt((1 + e)/(1 - e)).*tan(Ee./2)); % True anomaly    
   RV = Y + K.*(e*cos(w) + cos(w + ta)); % Radial Velocity
   end

I think the problem is here, because i have an error that says:

   ??? Error using ==> minus
   Matrix dimensions must agree.

   Error in ==> RVfun at 3
   M1 = 2*pi*(time2 - T)/P;

I don't understand this error because if i run this code out of the function, it works perfectly.

Anyway, the code is like this:

   % Estimates:
   T0 = min(time2);
   P0 = 1/f(Per == max(Per));
   e0 = 0.4;
   K0 = (max(rad_v2) - min(rad_v2))/2;
   Y0 = median(rad_v2);
   w0 = 0;

   x0 = [T0 P0 e0 K0 Y0 w0];
   minl = [min(time2) 700 0 0.02 min(rad_v2) 0]; % Lower limit 
   maxl = [Inf 900 1 0.06 max(rad_v2) 2*pi]; % Upper limit

   % Fit data:
   options = optimset('Algorithm', 'levenberg-marquardt', 'MaxIter', 1000,    'TolFun', 1e-10, 'TolX', 1e-10);
   [x, resnorm] = lsqcurvefit(@RVfun, x0, time2, rad_v2, minl, maxl, options);

Can you see what am I doing wrong?

Thank you! :)