R help - Jan 2004 - ecdf function

I am trying to use the ecdf function to find p-values (using a vector of numbers
to represent my new distribution and a test specific t-statistic value).  I am
using :
1-ecdf(vector)(t-stat)
 
vector<-c(5.386, 3.701717, 3.8289, 3.602, 4.469, 5.2087, 6.1613, 4.71181,
5.07716, 2.3517)
ecdf(vector)(4.6604)
[1] 0.5

 
R will only give me 1 significant digit after the decimal point and I am
interested in more significant digits.  How would I be able to do this?
 
I made up some data and it did give me many significant digits.
 
> g<-c(5,1,4.5,8.,6.75,9)
> ecdf(g)(7)
[1] 0.6666667

I was confused why this worked and my original example did not.

 
Thank you,
Maya 


---------------------------------


	[[alternative HTML version deleted]]

On Fri, 16 Jan 2004, Maya Sanders wrote:
> I am trying to use the ecdf function to find p-values (using a vector of
numbers to represent my new distribution and a test specific t-statistic value).
I am using :
> 1-ecdf(vector)(t-stat)
>
> vector<-c(5.386, 3.701717, 3.8289, 3.602, 4.469, 5.2087, 6.1613,
4.71181, 5.07716, 2.3517)
> ecdf(vector)(4.6604)
> [1] 0.5
>
>  R will only give me 1 significant digit after the decimal point and I
> am interested in more significant digits.  How would I be able to do
> this?
Add zeros: 0.5000000000000000000000000000000000000000000000

Your empirical CDF value at that point is 5/10, R just doesn't print
trailing zeroes.

	-thomas