sampling from a normal distribution

[Muddy]

Can anyone help me out with an equation (or algorithm) that I can use in a program I am writing to simulate sampling from a Normal distribution? I also need this for a Lognormal distribution.

I basically need a function:

SampledValue = f(RandomNumber)

Thanks for any help!

[DavidHooper]

I personally hate the normal distribution - too many nasty questions on Maths papers!

http://www.itl.nist.gov/div898/handb...n3/eda3661.htm

[Muddy]

Thanks David, but ....

I cannot find an explicit function or equation for the function I need on the page though. Any other references?

If I just had a function that returned the value of a normal cdf, I could probably work out the solution on my own.

Thanks!

[DavidHooper]

This help?:
http://www.itl.nist.gov/div898/handb...n3/eda3669.htm

To be honest, I don't completely understand what you want

[[Digital-X-Treme]]

Originally posted by DavidHooper
I personally hate the normal distribution - too many nasty questions on Maths papers!

*Glad to know it's not just me who thinks this

[Guv]

I found a formula which did not enlighten me much, but it might help you. The following is supposed to be the equation for the normal curve.

y = Exp(.5*x^2)/Sqr(2*pi).

A note says that x is measured in standard deviation units from the mean and the area under the curve is exactly one.

I plotted the above and it looks like a normal curve with a mean of zero. I t seem to me that areas under this curve represent probabilities. Thus integrals of the function would be probabilities. For example, the integral from x = - 1 to +1 would be the probability of a given sample value being between one standard deviation below the mean and one above it.

I have no idea how to use the above information for sampling from a normal distribution. Perhaps you can figure out something.