R devel - Apr 1997 - R-alpha: R <-> S compatibility; more demos -- using S-plus validate(..) ..

Look at the result of
(in S-plus:)
	validate(verbose=TRUE, outfile="......./validate-3.4.out")

which gives (for me, S-plus 3.4 on Solaris 2.5) 
2106 Lines of output.
It's full of tests using "standard data sets".  (--> 
'demos' available!)

Basically, at the end of each example, there are  statements which
should return  TRUE  if the 'validation test' is passed, e.g.,
	all(c(b.test$statistic == 1, abs(b.test$p.value - 0.1562) < tol, 
	      b.test$alternative == "two.sided"))

One bigger job would be to run these through R and analyze the
results... -- any volunteers ?
	
	((I could even send you the file if don't have S-plus
	  available...))

-- Martin
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

D() doesn't handle exp() correctly:

R> D(expression(exp(sin(x))),"x")
sin(x) * cos(x)
R> D(expression(exp(x)),"x")
x


Thomas Lumley
------------------------------------------------------+------
Biostatistics		: "Never attribute to malice what  :
Uni of Washington	:  can be adequately explained by  :
Box 357232		:  incompetence" - Hanlon's Razor  :
Seattle WA 98195-7232	:				   :
------------------------------------------------------------

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

R> D(expression(z * (log(z) /z)),"z")
	Error: invalid form in unary minus check

Yep. this is the result of an implete "if" statement in the C function
"isUminus" in deriv.c.  The function should read

static int isUminus(SEXP s)
{
	if(TYPEOF(s) == LANGSXP && CAR(s) == MinusSymbol) {
		switch(length(s)) { 
		case 2:
			return 1;
		case 3:
			if(CADDR(s) == R_MissingArg)
				return 1;
			else return 0;
		default:
			error("invalid form in unary minus check\n");
		}
	}
	else return 0;
}

This now gives

	> D(expression(z * (log(z) /z)),"z")
	(log(z)/z) + z * (1/z/z - log(z)/z^2)

which I thought was pretty ugly, but

	S-PLUS : Copyright (c) 1988, 1996 MathSoft, Inc.
	S : Copyright AT&T.
	Version 3.4 Release 1 for Sun SPARC, SunOS 4.1.3_U1 : 1996
	Working data will be in .Data
	> D(expression(z * (log(z) /z)),"z")
	(log(z)/z) + z * (1/z/z - log(z)/z^2)

So we are no worse than S.

And shame on you for trying to fool such a dumb bit of software :-).
	Ross
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To:
r-devel-request@stat.math.ethz.ch
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-