I've found some funny behavior in the dlnorm() function that I believe to be a
bug. For example, the following command:
> dlnorm( c(5e-323, 5e-324, 5e-325), 0, 0.3 )
Produces the output below, including the NaN value that should (ideally would
be) zero.
[1] 0 NaN 0
Warning message:
In dlnorm(c(4.94065645841247e-323, 4.94065645841247e-324, 0), 0, :
NaNs produced
Calling the log() function directly seems to produce appropriate results, e.g.:
> log(c(5e-323,5e-324,5e-524))
[1] -742.1375 -744.4401 -Inf
I realize that I am working in the regime of very small numbers, and that these
values are smaller than the value reported by .Machine$double.xmin ==
2.225074e-308, but it would be better (for me, at least!) if this corner
case(s?) was handled at a lower level.
I have found this behavior in the 64-bit versions (but NOT the 32-bit versions)
of R that I have access to. These include:
R version 3.2.5 (2016-04-14) -- "Very, Very Secure Dishes"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)
And
R version 3.3.0 (2016-05-03) -- "Supposedly Educational"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-w64-mingw32/x64 (64-bit)
And
R version 3.0.0 (2013-04-03) -- "Masked Marvel"
Copyright (C) 2013 The R Foundation for Statistical Computing
Platform: x86_64-w64-mingw32/x64 (64-bit)
And
R version 3.0.1 (2013-05-16) -- "Good Sport"
Copyright (C) 2013 The R Foundation for Statistical Computing
Platform: x86_64-unknown-linux-gnu (64-bit)
And
R version 3.0.1 (2013-05-16) -- "Good Sport"
Copyright (C) 2013 The R Foundation for Statistical Computing
Platform: x86_64-unknown-linux-gnu (64-bit)
But NOT in any of the 32-bit versions I have access to, including:
R version 2.11.1 (2010-05-31)
Copyright (C) 2010 The R Foundation for Statistical Computing
ISBN 3-900051-07-0
And
R version 3.0.0 (2013-04-03) -- "Masked Marvel"
Copyright (C) 2013 The R Foundation for Statistical Computing
Platform: i386-w64-mingw32/i386 (32-bit)
Thanks,
Ben
[[alternative HTML version deleted]]
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
