Page 1 of 1

which rank based INT does rntransform apply

Posted: Fri Jan 27, 2017 4:13 pm
by painoh92
Hello,

I have been reading into rank-based inverse normal transformations (INT) and I was wondering what type of rank-based INT is carried out when using the rntransform function.

According to this paper (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2921808/), most are similar but can be distinguished by the fractional offset (c) used to avoid infinity. From looking at the rntransform function, it appears that the offset value is 0.5. However, the equation in the paper referenced doesn't seem to match with the one in the rntransform function.

Many thanks,

Ollie