[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
Kit Baum <baum@bc.edu> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
st: Re: spline regression (Kit Baum) |

Date |
Tue, 25 Mar 2008 06:48:22 -0400 |

Mohammed,

If you graph x vs y, and break the line at the knot points, a linear spline allows the line to have kinks, like a dot-to-dot drawing. A quadratic spline has constant first derivatives == slopes at the knot points, so that it will not have any kinks. I don't know how to explain a second derivative in this context except to say that a curved line may have more or less curvature (such as a railroad track on a curve may be a broad curve or a sharp curve) and holding the second derivative constant causes the degree of curvature to be equal before and after the knot point (so that the locomotive will not derail at the knot point).

Kit Baum, Boston College Economics and DIW Berlin

http://ideas.repec.org/e/pba1.html

An Introduction to Modern Econometrics Using Stata:

http://www.stata-press.com/books/imeus.html

On Mar 24, 2008, at 02:33 , Mohammed wrote:

Thank you very much. Pardon me, I am not good in MAth. i will be very grateful if you explain more what you mean by "the derivative (slope) of the function is equal on either side of each knot point, but the curvature on either side may differ" and "The first and second derivatives of the function are equal on either side of each knot point." Thanks again Kit

* * For searches and help try: * http://www.stata.com/support/faqs/res/findit.html * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

- Prev by Date:
**st: correlation among residuals** - Next by Date:
**st: computing average partial effect in nonlinear models using forecasted distribution of x-variables** - Previous by thread:
**st: correlation among residuals** - Next by thread:
**st: computing average partial effect in nonlinear models using forecasted distribution of x-variables** - Index(es):

© Copyright 1996–2014 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |