The problem is that 1-1e-10 (for example) is subject to floating point rounding errors, such that it isn't really the same distance from 1 (the upper end of the interval) as 1e-10 is from 0 (the lower end of the interval). # For values of p in, specify lower tail probabilities It's just that, to get highly accurate results in the distribution's upper tail, you'll need to avail yourself of the function's lower.tail argument. It turns out (as noted by Spencer Graves in his response to this same question on the R-devel list-serve) that qnorm() does in fact perform as advertised. I have checked the qt function and it is also not symmetric in the tails. This is quite surprising to me, as it is well known that this distribution is symmetric, and I have seen implementations in other languages that are symmetric. qqnorm (residuals (LM), ylab'Residuals') qqline (residuals (LM)) I can figure out how to get the qqnorm part of the plot, but I can't seem to manage the qqline: ggplot (LM, aes (sample.resid)) + statqq () I suspect I'm missing something pretty basic, but it seems like there ought to be an easy way of doing this. ![]() ![]() What I have noticed is that the qnorm implementation in R is not symmetric when looking at the tails. I am looking for high precision values for the normal distribution in the tail (1e-10 and 1 - 1e-10), as the R package that I am using sets any number which is out of this range to these values and then calls the qnorm and qt function.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |