floor(NaN) problem fixed in massdist.c (PR#291)

About this list Date view Thread view Subject view Author view Other groups

Subject: floor(NaN) problem fixed in massdist.c (PR#291)
ntakebay@bio.indiana.edu
Date: Fri 08 Oct 1999 - 12:57:44 EST


Message-Id: <199910080257.EAA06625@pubhealth.ku.dk>

Full_Name: Naoki Takebayashi
Version: 0.65.0+R-release.diff (Oct 6, 1999)
OS: Linux/Alpha
Submission from: (NULL) (129.79.224.171)

This will fix the "problem 2 (crash in fft)" in Bug ID #277

On Linux/Alpha, make check failed because R could not handle the following
example in base-Ex.R

##___ Examples ___:

# The Old Faithful geyser data
data(faithful)
 :
 :
## Missing values:
x <- xx <- faithful$eruptions
x[i.out <- sample(length(x), 10)] <- NA
doR <- density(x, bw=0.15, na.rm = TRUE)
doN <- density(x, bw=0.15, na.rm = FALSE)
        ### it fails at this statement ###
lines(doR, col="blue")
lines(doN, col="red")
points(xx[i.out], rep(.01,10))

The following statement in density() was not getting the right output:
       
    y <- .C("massdist", x = as.double(x), nx = N, xlo = as.double(lo),
        xhi = as.double(up), y = double(2 * n), ny = as.integer(n),
        NAOK = has.na, PACKAGE = "base")$y

y[1] and y[2] should be 0 and 0 but both of them became NaN after this
statement.

So I looked at src/appl/masdist.c with a debugger. The bug is caused
by floor(). On Alpha/Linux, floot(NaN) returns 0. x[i] in the
following code is the list passed from the above statement in
density(). The array x contains a couple of NaN. When x[i] is NaN,
xpos become NaN. Then ix = floor(xpos) returns an unexpected result.
On linux/i386, ix become -2147483648 when xpos = NaN(0x80000000007a2).
However ix become 0 on Linux/Alpha when xpos=NaN. When xpos = NaN, it
shouldn't go into the following "if" statement. But it goes inside on
Alpha/Linux (because ixmin is 0), and set y[0] and y[1] to NaN.

Following is extracted from massdist.c

    for(i=0 ; i<*nx ; i++) {
            xpos = (x[i] - *xlow) / xdelta;
            ix = floor(xpos);
            fx = xpos - ix;
            if(ixmin <= ix && ix <= ixmax) {
                y[ix] += (1.0 - fx);
                y[ix + 1] += fx;
            }
            else if(ix == -1) {
                y[0] += fx;
            }
            else if(ix == ixmax + 1) {
                y[ixmax + 1] += (1.0 - fx);
            }
    }

So I think the statement

        if(ixmin <= ix && ix <= ixmax) {

should be changed to

            if((! isnan(ix)) && ixmin <= ix && ix <= ixmax) {

or

everything inside of this for loop should be encolsed with

    for(i=0 ; i<*nx ; i++) {
        if ( ! isnan( x[i] ) ) {
                :
                :
       }
    }

A patch to fix this bug is included.

--- R-0.65.0/src/appl/massdist.c.orig Thu Oct 7 21:26:53 1999
+++ R-0.65.0/src/appl/massdist.c Thu Oct 7 21:47:55 1999
@@ -38,6 +38,7 @@
         y[i] = 0;
 
     for(i=0 ; i<*nx ; i++) {
+ if ( ! isnan( x[i] ) ) {
         xpos = (x[i] - *xlow) / xdelta;
         ix = floor(xpos);
         fx = xpos - ix;
@@ -51,6 +52,7 @@
         else if(ix == ixmax + 1) {
             y[ixmax + 1] += (1.0 - fx);
         }
+ }
     }
 
     for(i=0 ; i<*ny ; i++)

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


About this list Date view Thread view Subject view Author view Other groups

This archive was generated by hypermail 2b25 : Tue 04 Jan 2000 - 14:16:09 EST