| forum.alglib.net http://forum.alglib.net/ |
|
| Linear regression http://forum.alglib.net/viewtopic.php?f=2&t=461 |
Page 1 of 1 |
| Author: | patex1987 [ Wed Oct 26, 2011 10:22 am ] |
| Post subject: | Linear regression |
I cant find a way how to calculate the linear regression. The problem is with lrbuild function. Can me someone explain how to interpret the data into double[,] xy . What is Nvar and NPoints. For example i have these points : x : 1 , 2 , 3 ,4 , 5 y : 2 , 4, 6 , 8 , 10 How to put them in to the array xy??? I tried "double[,] x = new double[,]{ {1,2},{2,4},{3,6},{4,5},{5,10}};" but then i get some weird array. Please someone help with this. I am very confused about this |
|
| Author: | Sergey.Bochkanov [ Wed Oct 26, 2011 10:38 am ] |
| Post subject: | Re: Linear regression |
In your case NVars=1 and NPoints=5. And your xy array is correct. ALGLIB Reference Manual has section for linreg subpackage - http://www.alglib.net/translator/man/ma ... nit_linreg It describes several functions which work with linear model structure, one of them is lrprocess() - you should create structure with lrbuild() and use lprocess() to calculate linear model estimate. |
|
| Author: | patex1987 [ Wed Oct 26, 2011 10:53 am ] |
| Post subject: | Re: Linear regression |
One more thing. Iwant to get the coeffitients a,b from the equation y = ax+b . Which function should i use, beacuse i found another one lsfitlinear? |
|
| Author: | Sergey.Bochkanov [ Wed Oct 26, 2011 12:44 pm ] |
| Post subject: | Re: Linear regression |
For historical reasons ALGLIB has two similar sets of functions for linear modelling. You can use linreg functions, and get your coefficients with lrunpack(). |
|
| Author: | joamaca [ Fri Dec 09, 2011 4:58 pm ] |
| Post subject: | Re: Linear regression |
Using lrbuild, is there any way to force that the coefficients not take any value? For example, only inside a range, -20<A<20 |
|
| Page 1 of 1 | All times are UTC |
| Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group http://www.phpbb.com/ |
|