HP 41 advantage module curve fitting giving confusing results.

[halbitton]

Hi Folks.
I'm poking around the CFIT program on my 41cx with the advantage module installed. To verify I'm using the program correctly I'm putting in the following 3 sets of coordinates: (3,3) (4,4) (5,5), which should return what amounts to Y=X using the LIN (linear) fit option. Instead, I,m getting a=2.0577 b=.7877 r^2=.4129, which I think translates to y=2.0577 + .7877x. What am I doing wrong?
best regards, Hal

[trojdor]

(12-29-2022 04:39 AM)halbitton Wrote:  Hi Folks.
I'm poking around the CFIT program on my 41cx with the advantage module installed. To verify I'm using the program correctly I'm putting in the following 3 sets of coordinates: (3,3) (4,4) (5,5), which should return what amounts to Y=X using the LIN (linear) fit option. Instead, I,m getting a=2.0577 b=.7877 r^2=.4129, which I think translates to y=2.0577 + .7877x. What am I doing wrong?
best regards, Hal

Hal,
Looks to me like you're probably just entering the data wrong. Those three sets of data points should give you a Lin r^2 value of 1.00. (IOW, an exact fit)
An r^2 of .4129 is a horrible fit.
If I were to hazard a guess, I suspect you're not entering each of the pairs of data as X,Y pairs. IOW, you first clear the summation registers, then type your first Y (3, for your example), then press Enter to place it on the Y reg, then type your first X (3, for your example), then press the Summation + key (as labeled in the upper left). Follow the same steps for the next data pair, etc. When you've entered all the data, then do the curve FIT, and choose your curve preference. (in your example, LIN)
Follow with the R/S key to get your A, B, and R^2 values.
If you enter everything correctly, you'll get an a=0, b=1 and r^2=1.00

mike