Lognormal distribuiton

[PeterVincent]

Hi,

I would be obliged if the following regarding lognormal distributions can be checked:

X is lognormal if ln(X) = y is distrib Normal

Y ~N(mu, sigma)

Note X ~ LN(mu, sigma) where mu and sigma are mean and standard deviation of Y ~N(mu, sigma)


mu and sigmal are not mean and standardard deviation of X but of Y

Does anyone know of any worksheets on line that I can practice finding probabilities where lognormal distribs are concerned. Like the ones we all practiced on for normal random variables such as:

X~N(13, 6^2) find probability that 4 < x < 19.

Many thanks,

PeterVincent

[vinux]

You can use the normal probability calculation for log normal.

eg: Y ~N and X ~LN
So if you want to calculate P(4<X<19) = P(Log(4) <Y<Log(19) )

[PeterVincent]

Thanks Vinux.

Y ~N(3, 4) and X ~LN
If I want to calculate P(4<X<19)

I will need to standardise, what do I do with the 4 & 19?

= P(Log(4) <Y<Log(19))

P(4<X<19) = P({[Log(4)- 3]/2} < Z < {[Log(19) - 3]/2})

OR

P(4<X<19) = P(Log{[4- 3]/2} < Z < Log{[19 - 3]/2})


where Z ~ N(0, 1)

Thanks again,

PeterVincent

[vinux]

P(4<X<19) = P({[Log(4)- 3]/2} < Z < {[Log(19) - 3]/2})

because Log(X) only follows Normal distribution

[PeterVincent]

Hi Vinux,

I have posted regarding working with log normal distributions at:

http://www.talkstats.com/showthread.php?t=10005

I would really appreciate it if you could look over this, as it is kinda related to this post which you have already helped me greatly with.

Many thanks,

PeterVincent