This error keeps popping up and I have no idea why: Error using viscircles Expected input number 1, centers, to be real. Error in viscircles​>validateC​entersAndR​adii (line 283) validateat​tributes(c​enters,{'n​umeric'},{​'nonsparse​','real', ... Erro

I am trying to do an animation for a linkage system and it is not plotting one of the links for some reason.

P1_pos = [56.25;43.1];
P2_pos = [56.25-A*cos(theta);43.1-A*sin(theta)];
P4_pos = C*[0;1];
P5_pos = D*[0*sin(theta);-1+1*sin(theta)]; %P5_pos = D*[0;-1+sin(theta)];

E = sqrt(A^2 + D^2 - 2*A*D*cos(theta));
a = asin(A*sin(theta)./E);
b = acos(E.^2 + C^2 - B^2)./(2*E*C);
P3_pos = [C + B*cos(a+b); B*sin(a+b)];

P5x = P5_pos(1,:);
P5y = P5_pos(:,1);
P5vx = diff(P5x);
P5vy = diff(P5y);
P5v = sqrt(P5vx.^2 + P5vy.^2);
ani = subplot(2,1,1);
P1_pos_cir = viscircles(P1_pos',0.05);
P2_pos_cir = viscircles(P2_pos(:,i)',0.05);
P3_pos_cir = viscircles(P3_pos(:,i)',0.05);
P4_pos_cir = viscircles(P4_pos',0.05);
P5_pos_cir = viscircles(P5_pos(:,i)',0.05);

[KALYAN ACHARJYA]

A?

[dpb]

Don't put message in title...truncated error message.

We can't test as don't have constants but the obvious problem is that one (or more) of your Px_pos values is complex. E is the most likely culprit it would appear.

Use the debugger and see what is going on...

[Matthew Palermo]

Here are the constant values

A = 8.25; %in

B = 48;

C = 42;

D = 20;

t = 0:0.01:20; %sec

omega = 5.0265; %rad/sec

theta = omega*t;

and this is the full error message:

Error using viscircles

Expected input number 1, centers, to be real.

Error in viscircles>validateCentersAndRadii (line 283)

validateattributes(centers,{'numeric'},{'nonsparse','real', ...

Error in viscircles>parseInputs (line 245)

[centers, radii] = validateCentersAndRadii(centers, radii, first_string);

Error in viscircles (line 115)

[ax, centers, radii, options] = parseInputs(varargin{:});

Error in ME450_project_linkage (line 35)

P3_pos_cir = viscircles(P3_pos(:,i)',0.05);

[Jeremy]

P3_pos_cir = viscircles(P3_pos(:,i)',0.05);

Is this inside a loop of the construct for i = 1:n ?

I don't see a for statement here, and i is default to sqrt(-1) in matlab, unless you reassign it.

[dpb]

Yeah, I noticed that too, but that's using i as an array index which would be another error on indexing must be real and positive...

[Matthew Palermo]

Yes, it is inside a for loop

Answer 1

[Walter Roberson]

You did not give us any values to test with for A, B, C, D

E = sqrt(A^2 + D^2 - 2*A*D*cos(theta));

Looks like the area of a triangle with sides A and D and angle theta between them. The area of a triangle is often greater than 1 but depending on the values of the variables could also be small < 1.

a = asin(A*sin(theta)./E);

A*sin(theta)./E can end up greater than 1, leading to asin() of a value greater than 1, which is complex valued. Everything gets messy from that point