HP Forums / HP Museum Forums / Old HP Forum Archives v / WP34s STATs Prediction Question

Linear Mode
WP34s STATs Prediction Question


▼
Post: #5
06-11-2011, 09:14 PM
Norman Dziedzic Offline
Unregistered
Posts: 203
Threads: 29
Joined: Nov 2009

I either don't understand how this function works or there is a bug with it...

I enter the following X & y values with Sigma+

Y,X

36,6

25,5

81,9

100,10

With each entry the wp34s shows the number of entries as expected.

I do a STAT BestF then I key h-6(L.R.) and the top area says Power. X register is -6.09818323034e-14 and y register is 2. So far so good.

Now I key in 4 and press f-6 for a predicted y but instead of 16 I get 8.

Edited: 11 June 2011, 9:15 p.m.

Find
▼
Post: #6
06-11-2011, 09:33 PM
Paul Dale Offline
Unregistered
Posts: 3,229
Threads: 42
Joined: Jul 2006

Looks like a bug. I know I had this working at one stage :-(
Fix will come.


- Pauli

Find
▼
Post: #7
06-11-2011, 09:49 PM
Norman Dziedzic Offline
Unregistered
Posts: 203
Threads: 29
Joined: Nov 2009

I believe it is a general fitting error as Exp mode doesn't work either. It seems the linear model is used no matter what.

Find
▼
Post: #8
06-11-2011, 10:06 PM
Paul Dale Offline
Unregistered
Posts: 3,229
Threads: 42
Joined: Jul 2006

The parameters were being calculated correctly using the appropriate model but a linear model was used after that. Must have lost the correct code somewhere along the line :-(

Updated version is in subversion now.


- Pauli

Find

Forum Jump: