--- On Wed, 6 May, 2009 at 6:51 AM Jacob Wegelin wrote:
> > Can anyone point me to a reference for -mkspline- ?
> >
> > I mean, ideally, a published article or book that
> > describes in software-independent terms (so that
> > statisticians who do not know Stata will understand)
> > exactly the algorithm and/or formula that the
> > Stata programmers implemented when they wrote
> > -mkspline-.
> >
> > I apologize if the answer is in [R] Stata Reference
> > Manual. I do not own that book.
--- On Tue, May 5, 2009 at 7:05 PM, Kieran McCaul wrote:
> > Frank Harrell's book "Regression Modelling
> > Strategies" has a chapter on restricted cubic splines.
--- On Wed, 6/5/09, Jacob Wegelin wrote:
> Actually, -mkspline- computes *linear* splines.
Linear splines are the default in -mkspline-, but restricted
cubic spline were newly added as an option in Stata 10, so
you are both right.
As you already guessed the algorithm is discussed in the
Reference manual. A discussion on linear splines can also be
found in:
Lawrence C. Marsh and David R. Cormier (2002) "Spline
regression models" Quantitative Applications in the Social
Sciences, nr 137. Thousand Oaks: Sage.
Though in this book they only discuss the splines you get
when specifying the -marginal- option. This the problem
wen trying to use other documentation than the Stata manual:
there are many parameterizations possible, and only the
Stata manual can tell you with authority which parameterization
is implemented in Stata. The same is true with the reference
to Frank Harrell's book: Stata uses a rescaled version of
the formulae presented in that book...
Hope this helps,
Maarten
-----------------------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany
http://home.fsw.vu.nl/m.buis/
-----------------------------------------
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
