Replicating R's qnorm() function in Stata

[Ariel Linden]

Hi Fellow Statalisters,

I am trying, unsuccessfully, to replicate the results when using R's qnorm() in Stata. qnorm() is R's equivalent to Stata's invnormal(), but it takes 3 parameters qnorm(p, mean = 0, sd = 0), where p is the signficance level.

In Stata, invnormal(p) just takes the p value.

Here is the code in R and the resulting values:

 
 qnorm(0.025, 0.0224111282843895, sqrt(0.0043818)) 

[1] -0.107329

 qnorm(0.001, 0.0224111282843895, sqrt(0.0043818)) 

[1] -0.1821473

 qnorm(1-0.025, 0.0224111282843895, sqrt(0.0043818)) 

[1] 0.1521513

 qnorm(1-0.001, 0.0224111282843895, sqrt(0.0043818)) 

[1] 0.2269696
I would be grateful for some help on this!

Thanks!

Ariel

[Scott Merryman]

. disp 0.0224111282843895 +invnormal(.025)*sqrt(0.0043818)
-.10732901

. disp 0.0224111282843895 +invnormal(.001)*sqrt(0.0043818)
-.18214731

. disp 0.0224111282843895 +invnormal(1- .025)*sqrt(0.0043818)
.15215127

. disp 0.0224111282843895 +invnormal(1- .001)*sqrt(0.0043818)
.22696957

[Ariel Linden]

Wow, thank you, Scott! That certainly works like a charm!