Getting "Error using horzcat." when trying to draw displacement graph for a Mechanism

First of, I am new to Matlab and sorry for my bad English. I'll try to explain clearly. I need to draw the displacement graph of a mechanism that has degrees of freedom = 1. That means every movement depends on the "theta21" angle. I tried to code that but I am having that error. What am i missing?
Mechanism;

My code;
A0A = 12.6; AB = 14; AD = 37.1;
EB = 45.5; BD =22.4; alfa = 39;

teta21 = linspace(128.57,488.57);
x1 = [1 -25.2*cosd(teta21-45) -1217.65];
y1 = roots(x1); A0D = y1(y1>=0);
teta31 = asind((A0D*sind(45)-A0A*sind(teta21))/AD);
epsilon = acosd((power(A0A,2)+power(AD,2)-power(A0D,2))/(2*A0A*AD));
A0B = sqrt((power(A0A,2)+power(AB,2))-(2*A0A*AB*cosd(epsilon)));
beta = acosd((A0D*cosd(45)-BD*cosd(teta31))./(A0B));
gamma = 180-(alfa+beta);
x2 = [1 (-2.*A0B.*cosd(gamma)) -1796.68];
y2 = roots(x2); A0E = y2(y2>=0);
A0Ex = A0E*cosd(alfa);
Error that i got;

[Voss]

Do you intend y1 to be the roots of a degree-101 polynomial?

[Mustafa Birkan Kilicaslan]

It is for finding Ao-D distance using cosine theorem. When I solve the equation, I have 2 roots, one of them is negative and the other is positive. I keep both 2 roots in y1 and then pick the positive one with ;

A0D = y1(y1>=0);

[Voss]

teta21 = linspace(128.57,488.57); % size(teta21) is 1-by-100
x1 = [1 -25.2*cosd(teta21-45) -1217.65]; % size(x1) is 1-by-102
y1 = roots(x1); A0D = y1(y1>=0); % size(y1) is 101-by-1. size(A0D) is 51-by-1

[Mustafa Birkan Kilicaslan]

So what is the problem? I don't get it?

[Voss]

When I run it, y1 has 101 elements. You say y1 has 2 elements. I don't know why that would be either.

[Mustafa Birkan Kilicaslan]

What I am trying to do is, define AoD dimension for every teta21 angle. I need to draw the graph for 1 period of the movement. So I picked teta21 angle like linspace(128.78,488.57) . (Starting angle is 128.57 degree, 1 period is 360 degree, then finishing angle is 128.57+360 = 488.57) Is that the right way to do it then?

[Voss]

That seems reasonable. It will give you 100 teta21 angles, spanning one complete period.

The next line calculates the cosine of teta21 and prepends and appends some endpoints. Also reasonable.

But the next line after that is what I'm not sure about. Unless the call to 'roots' here refers to some function in a toolbox I don't have (or a function you wrote, for instance), 'roots' is used to find the roots of a polynomial. In this case the polynomial is of degree 101 and roots returns 101 roots. This seems to be not what is intended, since you say you get 2 roots.

You can find out what roots function gets called by doing this on the command line:

which roots -all

and see if there are any anywhere besides in the MATLAB polyfun directory.

Answer 1

[Meg Noah]

This executes without error, but I'm unclear about whether it solves your problem.

A0A = 12.6; AB = 14; AD = 37.1;
EB = 45.5; BD =22.4; alfa = 39;
teta21 = linspace(128.57,488.57);
x1 = [1 -25.2*cosd(teta21-45) -1217.65];
y1 = roots(x1); A0D = y1(y1>=0);
teta31 = asind((A0D*sind(45)-A0A*sind(teta21))/AD);
epsilon = acosd((power(A0A,2)+power(AD,2)-power(A0D,2))/(2*A0A*AD));
A0B = sqrt((power(A0A,2)+power(AB,2))-(2*A0A*AB*cosd(epsilon)));
beta = acosd((A0D*cosd(45)-BD*cosd(teta31))./(A0B));
gamma = 180-(alfa+beta);
x2 = horzcat(complex(1,0), -2.0*A0B'*cosd(gamma), complex(-1796.68,0));
y2 = roots(x2); A0E = y2(y2>=0);
A0Ex = A0E*cosd(alfa);