Why am I not getting the m array?

Hi all,

I have the following function:

function [m,yinter] = radial_lines(X,Y,xi,yi)
if X==Y
    m=inf;
    yinter= X;
   return

end
% Calculate the slope:
m = (Y - yi)./((X) - xi);
% Solve for y-intercept
yinter = yi - m.*xi;
fplot(yinter)
end

%% data
X=[6,7,10,14,13,9];
Y=[12,9,7,8,10,14];
xi=10;
yi=10;

The issue i am running into is that when using the radial_lines function, I get a plot with horizontal lines. However I need the lines to have a slope. I am not quite sure what i am doing wrong when calculating the slopes. I should be getting an arry of m values but the code does not give me that output. Can some one kinldy help me out? Thanks.

[Star Strider]

I am not certain what your code is doing.

An easier way to calculate the parameters for a linear regression is:

%% data
X=[6,7,10,14,13,9];
Y=[12,9,7,8,10,14];

B = [X(:) ones(size(X(:)))] \ Y(:)                                      % Parameters: B(1) = m,  B(2) = b

B = 2×1

-0.3344 13.2885

Yfit = [X(:) ones(size(X(:)))] * B;                                     % Regression Fit (For 'plot’)

figure
plot(X, Y, 'p', 'DisplayName','Data')
hold on
plot(X, Yfit, '-r', 'DisplayName','Regression')
hold off
grid
xlabel('X')
ylabel('Y')
legend('Location','best')
axis([5  15    6  15])

If you want statistics on the fit, use fitlm.

[Mariam Shahab]

Thank you for your resposne. Just to clreaify, i am not calutaling a linear regression. I am trying to get the equations of lines that start from (xi,yi ). The hexagon illustatred is formed by X and Y arrays.

[Torsten]

I am trying to get the equations of lines that start from (xi,yi ).

Then

yinter = Y - m.*X

not

yinter = yi - m.*xi

[Mariam Shahab]

Ok, thank you. I am still getting horizontal lines though (so m=0). Any advice for the slope of theses lines?

[Torsten]

Slope and y-intercept can be obtained from my answer given below.

X=[6,7,10,14,13,9];
Y=[12,9,7,8,10,14];
xi=10;
yi=10;
hold on
for i=1:6
plot([X(i),xi],[Y(i),yi])
end
hold off

[Mariam Shahab]

Thank you for your help. This is the graph I was aimming for.

Answer 1

[Star Strider]

That clears up much.

Plotting them is simply this —

%% data
X=[6,7,10,14,13,9];
Y=[12,9,7,8,10,14];
xi=10;
yi=10;


figure
plot([ones(size(X))*xi; X], [ones(size(Y))*yi; Y], '-k')
grid
axis([5  15    6  15])
axis('equal')
text(X*1.01,Y*1.01,compose('(%2d,%2d)',[X;Y].'))

It is not necessary to plot the individual regressions, although you can certainly do that if you want to. Getting the slopes and intercepts may be enough, and your function does that.

.

Answer 2

[Torsten]

yinter is not a function. So use "plot" instead of "fplot" (whatever abscissa data you might want to choose).

Not that yinter(3) = Inf which is not plotted.

%% data
X=[6,7,10,14,13,9];
Y=[12,9,7,8,10,14];
xi=10;
yi=10;
[m yinter] = radial_lines(X,Y,xi,yi)

m = 1×6

-0.5000 0.3333 Inf -0.5000 0 -4.0000

yinter = 1×6

15.0000 6.6667 Inf 15.0000 10.0000 50.0000

function [m,yinter] = radial_lines(X,Y,xi,yi)
  for i=1:numel(X)
    if X(i)==xi
      m(i)=inf;
      yinter(i)= inf;
    else
      % Calculate the slope:
      m(i) = (Y(i) - yi)/(X(i) - xi);
      % Solve for y-intercept
      yinter(i) = Y(i) - m(i)*X(i);
    end
  end
end