The RATS Software Forum

Thanks, Tom.
I was not aware of prj or sstats. They simplify my code a lot.
And, yes, the denominator should be squared in the calculation of PRESS.
Best wishes,
Ken.

Is there a simple procedure for validating a linear regression with the predicted regression error sum of squares (PRESS) statistic other than writing code like make x # x1 x2 compute h =x*inv(tr(x)*x)*tr(x) compute press =0.0 do i=1,N compute press =press+(%resids(i)/(1-h(i,i)))^2 end do i display ...

Thanks Tom,
I have found that axis=none in gcontour still produces a horizontal line at y=0.
How can that line be eliminated?
I want to produce, with vgrid and hgrid, only the axis values corresponding to a max or min value of f.
Thanks,
Ken.

What if array A is packed/lower triangular?
Thanks,
Ken.

I have an array A(625,625) with minimum value returned by %minvalue(A).
Is there a simple way of returning the indices for this value?
%minindex(A) does not work even though its argument is said to be an array.

gcontour(x=a,y=b,f=c) does not work if a, b, and c contain NA values. Can anyone suggest a solution? Sample code stores c(i,j)=i*j only for j>i: declare vector a(5) declare vector b(5) declare rectangular c(5,5) do i=1,4 do j=i+1,5 compute a(i)=i compute b(j)=j compute c(i,j)=i*j end do j end do i g...

I know there is a neural nets option in RATS, but are there any procedures for Multivariate Adaptive Regression Splines (MARS) or Hinging Hyperplanes (HHP) which have become reasonable alternatives? MARS was introduced by Friedman (1991) Annals of Statistics, vol 19, pp 1-141. HHP was introduced by ...

I can't find hshade in any search of the docs.
For a simple y vs. x scatterplot, I would liketo shade values x>3.5 (say).
How do that?

Bai-Perron fits multiple structural changes to a multiple regression. In the simple linear regression case, this is x<= or x> some value equal to one of the x values, if two linear pieces are estimated. Almost always, the intersection of the two pieces will not be at one of the x values, producing a...

NLLS has a METHOD=GENETIC option. I assume this is a genetic algorithm, and it works well on a small problem (4 declared NONLIN variables with 6 observations) where the partial derivative of the squared errors is discontinuous in one of the four variables but all partial derivatives exist for the ot...

Is it possible to use integer or binary restrictions on variables in the solution to an LP or QP program such as in RREG or LQPROG? If not possible for either: could I "trick" a QP to refelect a binary restriction on x1 by adding M*x1*(1-x1) to the objective function with a large value of ...

Thank you. However, I may not need to use ewise: In OLS, the normal regression equations are X'X*Beta=X'y In one of my test cases, X'X (4 x 4) is perfectly singular, but Beta is returned by linreg with no reported error. (|X'X| is reported as -3.3E-10, which is consistent with singularity) The SVD o...

%svdecomp(A) decomposes matrix A into A=U*D(w)*V'.
Is there a simple way in RATS to replace the elements of D by their
reciprocals, but by 0.0 if an element w is close to zero?
Or is the only way a do loop?
Thanks,
Ken.