Least squares regression


[Prev][Next][Index][Thread]

Least squares regression



I happened to be working with least squares regression modeling
and in utilizing the TI-86 to do this for me I noticed the
equation form it uses is a+bx (linear regression). For least
squares I need ax+b. I noticed in documentation for earlier
models that the ax+b and a+bx were both available. Yet in the
TI-86 docs there is only mention of a+bx. Does anyone know
anymore on this in relation to the TI-86? I know that I can just
reverse the values(bx+a) and the answer is correct, but I am just
wondering about the ax+b possibility...

Regards,

Christopher Zukowski
czukowski@qualcomm.com
PGP Fingerprint: 4471 22ED F5F7 FF67 520E OE97 5F84 427B
PGP DSS-fprint: DDCB 7A5F FE21 1364 F54D E016 8399 1E7F F079 D856
"Poor soul was to high strung."