Accuracy of Quadratic Regression

[bernhardSA]

Dear HP Prime User Thread

There seems to be a subtlety with the quadratic regression that I cannot figure out: the HP Prime results for a sample problem are subtly different from those from MS Excel, an on line tool (https://ezcalc.me/quadratic-regression-calculator/) and TI-84

x y
2003 813
2004 941
2005 962
2006 1053
2007 1132
2008 1194
2009 1205
2010 1244
2011 1254
2012 1262

-> Excel, and other tools calculate the quadratic fit as:
y = -5.46969697 X^2 + 22010.28788 X -22141315.33
-> HP calculates it is:
y = -5.47571307 X^2 + 22034.44253 X -22165560.54

It is close ... but annoyingly different. I played with HOME/CAS settings and set home number format to standard and floating 11 and CAS format to standard 12 to no avail.

I am not sure if this is related to floating point accuracy (although I thought that HP Prime has better accuracy than TI-84), or I may be missing a special setting.

I am a long time HP28S and HP4G user and HP fan in general and recently purchased my daughter an HP Prime. I would appreciate help/insight into this scenario

Thx

PS: The linear fit matches the other calculators exactly