Dear all
How does print.htest display the p-value in scientific notation?
(x - cor.test(iris[[1]], iris[[3]]))
Pearson's product-moment correlation
data: iris[[1]] and iris[[3]]
t = 21.65, df = 148, p-value 2.2e-16
alternative hypothesis: true correlation is not equal to 0
95 percent
Isn't it true that 0 2.2e-16?
cheers,
Rolf Turner
On 03/10/11 20:53, Liviu Andronic wrote:
Dear all
How does print.htest display the p-value in scientific notation?
(x- cor.test(iris[[1]], iris[[3]]))
Pearson's product-moment correlation
data: iris[[1]] and iris[[3]]
is(x)
[1] htest
# take a look at stats:::print.htest
format.pval(x$p.value)
[1] 2.22e-16
Does that answer your question?
KK
On Mon, Oct 3, 2011 at 10:53 AM, Liviu Andronic landronim...@gmail.com wrote:
Dear all
How does print.htest display the p-value in scientific notation?
(x -
One point to note, for information, in this discussion is
that cor.test() has apparently returned the P-value as
an exact zero:
x$p.value == 0
# [1] TRUE
identical(x$p.value, 0)
# [1] TRUE
(which, by the way, I was led to after trying log10(x$p.value)
and getting -Inf). Perhaps a more
Thanks all for your pointers. The following does trick:
base::format.pval(x$p.value) ##Hmisc also has such a function
[1] 2e-16
On Mon, Oct 3, 2011 at 10:25 AM, Rolf Turner rolf.tur...@xtra.co.nz wrote:
Isn't it true that 0 2.2e-16?
Yes, but it doesn't mean that the p-value actually hits
format.pval is documented and accessible from outside of base.
So you do not have to qualify it as base::format.pval
On 10/3/2011 11:24 AM, Liviu Andronic wrote:
Thanks all for your pointers. The following does trick:
base::format.pval(x$p.value) ##Hmisc also has such a function
[1] 2e-16
6 matches
Mail list logo