How to OLS several lines at once?

[phrozenfearz]

Hello PhysicsForums.com!

I've got several sets of data that are all intended to represent the same ideal data set. I need to fit a regression to said sets of data - but have no idea of how to go about it.

All multiple-regression literature I can find reads to the tune of: given y=a*f(x1)+b*f(x2), but I don't have multiple dependence on x. I just have 9 sets of equations that all look like:
Point 1...
y1=f(x1)
y2=f(x2)
y3=f(x3)
Point 2...
y1=f(x1)
y2=f(x2)
y3=f(x3)

If somebody could point me in the right direction, I would be very appreciative!

Thanks!

 

[ivyhawk]

Are you talking about Vector Autoregressive Model?
yt=a*y_t-1+b*x_t+...et
xt=c*x_t-1+d*y_t+...,ut

 

[phrozenfearz]

I don't think so... but please elaborate more on the method.

In return, I'll elaborate more on what I have and what I need:

I have several sets of [x] and [y] data series, taken by experiment, on separate occasions. The experimental setup, in terms of the two variables I need to fit a curve to, only varies in pressure. The variation comes in a slight translation along the x-axis.

My goal is to fit a curve to ALL of these data sets to act as an approximate representation of [y] in terms of [x] across the pressure range from the experiment.

Thanks!