Hello,
Sorry, my mistake, I grossly misunderstood the question.
qnorm(1e-300)
#[1] -37.0471
Anyway, you cannot go much smaller.
p <- 10^seq(-300, -400, by = -10)
z <- qnorm(p/2)
exp(z*z/2)
Hope this helps,
Rui Barradas
Às 16:11 de 21/06/19, jing hua zhao escreveu:
Dear Rui,
Thanks for
Hello,
Well, try it:
p <- .Machine$double.eps^seq(0.5, 1, by = 0.05)
z <- qnorm(p/2)
pnorm(z)
# [1] 7.450581e-09 1.22e-09 2.026908e-10 3.343152e-11 5.514145e-12
# [6] 9.094947e-13 1.500107e-13 2.474254e-14 4.080996e-15 6.731134e-16
#[11] 1.110223e-16
p/2
# [1] 7.450581e-09 1.22e-09
Hi Peter, Rui, Chrstophe and Gabriel,
Thanks for your inputs -- the use of qnorm(., log=TRUE) is a good point in
line with pnorm with which we devised log(p) as
log(2) + pnorm(-abs(z), lower.tail = TRUE, log.p = TRUE)
that could do really really well for large z compared to Rmpfr. Maybe I am
Hi Jing,
Peter pointed out how you can, more or less, get numbers for this, and he's
absolutely right. At the risk of giving unsolicited advice, though, Im
don't think you *should* in this case.
Someone on this list with more applied Statistics or Statistical Genetics
experience can correct me
> On 6/21/19 10:56 AM, Tierney, Luke wrote:
> [...]
> Something that would be useful and is being considered is having a
> mechanism for registering default condition handlers. This would allow
> the condition to be re-signaled with a custom class and then having
> a custom conditionMessage method
You may want to look into using the log option to qnorm
e.g., in round figures:
> log(1e-300)
[1] -690.7755
> qnorm(-691, log=TRUE)
[1] -37.05315
> exp(37^2/2)
[1] 1.881797e+297
> exp(-37^2/2)
[1] 5.314068e-298
Notice that floating point representation cuts out at 1e+/-308 or so. If you
want
You should take a look at https://CRAN.R-project.org/package=Rmpfr
Regards, Christophe
Christophe Dutang
CEREMADE, Univ. Paris Dauphine, PSL Univ., France
Web : http://dutangc.free.fr
Hi Luke,
Thank you for your response.
On 6/21/19 10:56 AM, Tierney, Luke wrote:
Thanks for the suggestion. However I don't think it is the right way
to go. I also don't care for what install.packages() does. Signaling a
warning and then an error means someone has to catch both the error
and the
Dear Rui,
Thanks for your quick reply -- this allows me to see the bottom of this. I was
hoping we could have a handle of those p in genmoics such as 1e-300 or smaller.
Best wishes,
Jing Hua
From: Rui Barradas
Sent: 21 June 2019 15:03
To: jing hua zhao;
In specific cases fligner.test() can produce a small p-value even when both
groups have constant variance.
Here is an illustration:
fligner.test(c(1,1,2,2), c("a","a","b","b"))
# p-value = NA
But:
fligner.test(c(1,1,1,2,2,2), c("a","a","a","b","b","b"))
# p-value < 2.2e-16
Thanks for the suggestion. However I don't think it is the right way
to go. I also don't care for what install.packages() does. Signaling a
warning and then an error means someone has to catch both the error
and the warning, or suppress the warning, in order to handle the error
programmatically.
Dear R-developers,
I am keen to calculate exp(z*z/2) with z=qnorm(p/2) and p is very small. I
wonder if anyone has experience with this?
Thanks very much in advance,
Jing Hua
[[alternative HTML version deleted]]
__
R-devel@r-project.org
Dear all,
`head()` returns a problematic output when a character is fed to its `n`
parameter.
doubles and logicals are converted to integer as if `as.integer` was used,
which I think is intuitive enough :
```
head(1:10, 4.1) # [1] 1 2 3 4
head(1:10, 4.9) # [1] 1 2 3 4
head(1:10, TRUE) # 1
Dear All,
Thanks for the input - it was sufficient for me to figure out a
solution, which I'll briefly describe here for future reference. This
draws primarily on the section of Writing R Extensions quoted by Martin
Morgan below, giving rise to the idea of combining environmental
variables, and
14 matches
Mail list logo
