On Tue, 12 Dec 2006, Talha Yalta wrote: > I just wanted to inform about a small typo in the gretl webpage. In > the numerical accuracy section there is a citation of Ramanathan's > text (2002) regarding the calculation of R2 when the regression does > not have an intercept. While reading the book (the 5th edition) the > other day, I realized that this discussion takes place on page 151 > instead of pp.163-4.
Thanks, Talha. There's a slightly larger issue that I need to fix in that documentation, namely that we've now moved to the definition of R^2 that is favored by NIST, for least squares models without an intercept. Allin.
