How can I add line of best fit into my subplots. I have tried to use polyfit but it doesn't work.

subplot(2,2,1)
scatter(WS_TG,mWS1)
polyfit(WS_TG,mWS1)
title('113.5N, 22.25N')
xlabel('time')
ylabel('wind speed (ms^-1)')
subplot(2,2,2)
scatter(WS_TG,mWS2)
title('113.5N, 22.5N')
xlabel('time')
ylabel('wind speed (ms^-1)')
subplot(2,2,3)
scatter(WS_TG,mWS3)
title('113.75N, 22E')
xlabel('time')
ylabel('wind speed (ms^-1)')
subplot(2,2,4)
scatter(WS_TG,mWS4)
title('113.75N,22E')
xlabel('time')
ylabel('wind speed (ms^-1)')

And this is the error that I got if I just run the first part:

Not enough input arguments.

Error in polyfit (line 56)

V(:,n+1) = ones(length(x),1,class(x));

Of curse, I also used the toolbox to draw the regression line for me. But every time when I jump to another plot, the line from the previous plot will disappear.

Answer 1

[the cyclist]

As stated in the polyfit documentation, it requires three inputs. (The third input is the degree of the fitting polynomial.)

You've called it with only two.

[On In Liem]

Thank you so much. So do you know if there is any other ways to create a linear line of the best fit?

[the cyclist]

I don't understand why you would not want to use polyfit. It's just that your calling syntax is wrong. You need

polyfit(WS_TG,mWS1,1)

instead of

polyfit(WS_TG,mWS1)

to fit a straight line to your data.

[On In Liem]

subplot(2,2,1)

scatter(WS_TG, mWS1)

polyfit(WS_TG, mWS1,1)

title('113.5N, 22.25N')

xlabel('time')

ylabel('wind speed (ms^-1)')

subplot(2,2,2)

scatter(WS_TG,mWS2)

polyfit(WS_TG, mWS2,1)

title('113.5N, 22.5N')

xlabel('time')

ylabel('wind speed (ms^-1)')

subplot(2,2,3)

scatter(WS_TG,mWS3)

polyfit(WS_TG, mWS3,1)

title('113.75N, 22E')

xlabel('time')

ylabel('wind speed (ms^-1)')

subplot(2,2,4)

scatter(WS_TG,mWS4)

polyfit(WS_TG, mWS4,1)

title('113.75N,22E')

xlabel('time')

ylabel('wind speed (ms^-1)')

So, I use the coding now. It gave me ans of the polyfit. But its still not plotting any lines on the graph. Do I have to use a y=mx+c equation to plot the line?

Thank you so much

[the cyclist]

Yes, you need to call polyfit with the output arguments, and then use those output arguments to plot the line.

Here is a simple example:

% Simulate some data
N = 20;
x = rand(N,1);
y = 2 + 3*x + 0.5*randn(N,1);

% Fit the polynomial, getting the fit parameters as output
% p(1) is the linear term, and p(2) is the constant term
p = polyfit(x,y,1);

% Pick a couple arbitrary x points, and find the corresponding y points
% from the fit. We'll use these to plot the straight-line fit.
x_fit = [0 1];
y_fit = p(2) + p(1)*x_fit;

% Create the figure window
figure
% Use the "hold" command, for two plots on one axis
hold on

% Plot the data as points
scatter(x,y)

%Plot the fit as a line
plot(x_fit,y_fit)

[On In Liem]

Thank you so much.