Replicate R's pnorm function

[Marc Peterson]

Hi Statalist,

I'm porting a couple R scripts to Stata and have hit a roadblock. I have been unable to replicate R's pnorm() function in Stata. I've searched this forum, Stata documentation, and the Web to no avail.

R has a dnorm() function that “returns the value of the probability density function (pdf) of the normal distribution” and appears to be the same as Stata’s normalden()
Example:
R: dnorm(-0.2231436, -0.1682927, 0.1931885) returns 1.983462
Stata: normalden(-0.2231436, -0.1682927, 0.1931885) returns 1.9834621

R’s pnorm() function “returns the value of the cumulative density function (cdf) of the normal distribution” and I haven't found a way to replicate it in Stata.
Using the same value, mean, and SD as above…
R: pnorm(-0.2231436, -0.1682927, 0.1931885) returns 0.3882342
Stata: I’ve had no luck.

Do any Statalisters know how to find the value of the CDF of the normal distribution?

Thank you,
Marc Peterson

[Clyde Schechter]

Stata's cumulative distribution function for the normal distribution is -normal()-. But you cannot give it three arguments the way you can with -normalden()-. You have to convert your x, m, and s into a z-score, z = (x-m)/s, and that becomes the argument.

. display normal((-0.2231436-(-0.1682927))/0.1931885)
.38823422

[Marc Peterson]

Fantastic! Thank you, Clyde!