R devel - May 2001 - deriv (PR#953)

------- start of forwarded message -------
From: Martin Maechler <maechler@stat.math.ethz.ch>
To: R-core@stat.math.ethz.ch
Subject: PROTECT() bugs in deriv(*, *, function.arg = )
Date: Mon, 16 Apr 2001 21:02:10 +0200

In R versions 0.50 and 0.64.2 ,
the following worked

      > deriv(expression(sin(cos(x) * y)), c("x","y"),
function(x,y){})
      function (x, y)
      {
          .expr1 <- cos(x)
          .expr2 <- .expr1 * y
          .expr4 <- cos(.expr2)
          .value <- sin(.expr2)
          .grad <- array(0, c(length(.value), 2), list(NULL, c("x",
              "y")))
          .grad[, "x"] <- -.expr4 * (sin(x) * y)
          .grad[, "y"] <- .expr4 * .expr1
          attr(.value, "gradient") <- .grad
          .value
      }



The second bug (gctorture-provoked) is a missing PROTECT(), which I will
commit when I've run make check.

The first bug is very strange and I suspect there's something going on in
the memory management.  Supplying a function as the third argument should
cause the body of the function to be replaced by the derivative. This
sometimes works, but doesn't depend on the function in any simple way.

It seems to depend on how new the function is. With a new anonymous
function it never works, but with a base function like lm or log it always
works:

 ## works
 deriv(expression(sin(cos(x) * y)), c("x","y"), log)
 ## identical function
 LOG<-function (x, base = exp(1))
    if (missing(base)) .Internal(log(x)) else .Internal(log(x, base))
 ## doesn't work
 deriv(expression(sin(cos(x) * y)), c("x","y"), LOG)

Creating a new function somewhere else in the search path doesn't help
 assign("a",function(x,y){},pos=3)
 # doesn't work
 deriv(expression(sin(cos(x) * y)), c("x","y"), a)

but copying a loaded function does

  library(tcltk)
  a<-tkbind
  #works
  deriv(expression(sin(cos(x) * y)), c("x","y"), a)


????

	-thomas


-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

tlumley@u.washington.edu writes:
> The first bug is very strange and I suspect there's something going on
in
> the memory management.  Supplying a function as the third argument should
> cause the body of the function to be replaced by the derivative. This
> sometimes works, but doesn't depend on the function in any simple way.
> 
> It seems to depend on how new the function is. With a new anonymous
> function it never works, but with a base function like lm or log it always
> works:
> 
>  ## works
>  deriv(expression(sin(cos(x) * y)), c("x","y"), log)
>  ## identical function
>  LOG<-function (x, base = exp(1))
>     if (missing(base)) .Internal(log(x)) else .Internal(log(x, base))
>  ## doesn't work
>  deriv(expression(sin(cos(x) * y)), c("x","y"), LOG)
> 
> Creating a new function somewhere else in the search path doesn't help
>  assign("a",function(x,y){},pos=3)
>  # doesn't work
>  deriv(expression(sin(cos(x) * y)), c("x","y"), a)
> 
> but copying a loaded function does
> 
>   library(tcltk)
>   a<-tkbind
>   #works
>   deriv(expression(sin(cos(x) * y)), c("x","y"), a)
<grin> 
You had me going there too.

Try stripping the "source" attribute from the template functions....
(or from the output of deriv() for that matter). I'll give it a think
how to do that inside the deriv function so as to preserve user
sanity.

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk)             FAX: (+45) 35327907
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._