(50g) Normal Distribution

[John Keith]

Thanks to Albert and Dieter's help I have a working program for the normal PDF with mean and variance. Dieter's method for dealing with the mean and variance greatly simplified the program.

On the other hand, the idea of using 1-ln(sqrt(2*pi)) proved to be less accurate on the HP 50 so I stuck with dividing by sqrt(2*pi) instead.

The resulting program is still big and ugly but reasonably fast, around 43 milliseconds. It is accurate to +/- 1 ULP for all test cases in the above posts.

Code:


\<< ROT - SWAP \v/ SWAP OVER / DUP SQ IP .5 * UNROT DUP IP SWAP FP \-> s i f
  \<< DUP NEG EXP SWAP i SQ .5 * - i f * - f SQ .5 * - EXP * 2.50662827463 s * /
  \>>
\>>