Regression Learner: One column for x-axis; multiple for y-axis ?

If I have a Matrix like this:

Is it possible, to use the first column for my "x-axis" and ALL of the other columns for my "y-axis"?

I need a linea regression (a plot), that fits all my points at best. Like this:

Now i don't know, how to tell MATLABs Regression Learner to do that. The only option to do seems like taking the last column for the y-axis and all the other as x-Axis. However then only one column at a time as x.

Greetings and Thank you.

[Alan Stevens]

You can save the "y" data as a matrix, Y, (size 11x19 by the looks of it). If you then just plot(x,Y,'*') Matlab will automatically plot the whole lot as in your hand drawn picture.

[dpb]

I don't know the regression learner tool; I can imagine the title would be such that it does expect only single input/single output variables.

You don't need it to just fit a regression model; for the simplest case w/o diagnostics and all that, simply

x=M(:,1);
y=M(:,2:end);
b=polyfit(x,y(:),1);
yh=polyval(b,[min(x) max(x)]);

will give you the coefficients and the two endpoints of the fitted line.

NB HOWEVER: If you include the group of zeros at x=0 and x=2 in the fit, you'll not get a line like you've drawn by hand/eye but one that will come much closer to going through the origin and steeper slope.

Depending upon which toolboxen you have, there are other higher-level fitting routines available.

I see the documentation for the statistics toolbox that contains RegressionLearner doesn't deign to even mention simple linear regression; the "Linear Regression" link just jumps into multivariate and all. That doesn't mean you can't still just use the "LinearModel" tool/object, though.

Sometimes TMW can't see the trees for the forest in the documentation...