Talk:Statistics/Curve fitting

I would suggest perhaps mentioning that modern spreadsheet programs typically allow one to do such regression analysis. Microsoft Excel for example offers six basic trend fit choices: linear, logarithmic, polynomial, exponential, power, and moving average. --Billymac00 04:06, 10 December 2006 (UTC)

Examples
(ok I'm back) here is a simple example using the free Pari-GP program: given a data set  (2,5),(3,10) and (4,19)

to solve to a presumed quadratic fit,


 * a=[1,2,4;1,3,9;1,4,16]  : b=[5;10;19]

c=matsolve(a,b) returns  [7   -5    2 ]


 * 2x^2-5x+7

One can then evaluate this in a spreadsheet/plot for adequacy of fit

Now, say you change the presumed fit to the form Y=C2*EXP(x)+C1* x+ C0

we modify a to a=[1,2,7.39;1,3,20.01;1,4,54.6]    (where 54.6=EXP(4))

for this limited dataset, we get an equally great fit ::(C2,C1,C0)=(0.1821,2.702,-1.750)

...--Billymac00 (talk) 16:40, 7 March 2008 (UTC)

I hope I'm not intruding
I just finished a lesson on curve/line fitting in calculus today, I hope you don't mind it if I change some things that I believe are helpful. Just change it back if it doesn't reflect the statistics way of doing this, or if what I'm doing doesn't make sense. --Danthemango (talk) 09:17, 7 March 2009 (UTC)

Wolfram Alpha
This site is very powerful and can do a host of regression fits. For example, from the article, the box inputs are: linear fit {{2.99,388.7}, {3.29,404.67}, {3.49,349}} or quadratic fit {{2.99,388.7}, {3.29,404.67},  {3.49,349}}

Other choices are cubic fit, exponential fit, log fit for example It also provides a plot of the points and fit--Billymac00 (talk) 03:44, 3 December 2009 (UTC)