(50g) Normal Distribution

[Albert Chan]

Hi, John Keith

Now that you have a good PDF function, here is the bad news.
The extra precision is basically useless.

z-score = (x–µ)/σ, with all 3 variables non-exact.

Even if you have accurate z-score, say 20.33333 33333 (12 digits precision)

eps = ±5e-11, thus PDF = 6.64644886446e-91 to 6.64644887799e-91

PDF value is good only for 9 digits, even if the function can generate more.
That was why I mentioned direct method is good enough, 6.64644887e-91