Re: What distribution is this?
- From: Ray Koopman <koopman@xxxxxx>
- Date: 28 Apr 2007 00:42:07 -0700
On Apr 27, 2:12 pm, Ray Koopman <koop...@xxxxxx> wrote:
On Apr 27, 12:01 pm, Enter The <enter...@xxxxxxxxx> wrote:
Hi,
I have a set of variables (x_i,y_i) where x_i >=0, y_i is a real
number.
I believe it follows a distribution of form
y_i = Y + sqrt(x_i) * stdev * Z
where Y is a real number, Z is a normal random variable, with mean 0
and standard deviation 1
For the given set of variables, how do I determine Y, along with
confidence bounds for Y, and a measure of whether or no the
distribution holds?
the following picture might help to describe what I'm doing.
<a href="http://img339.imageshack.us/my.php?image=whatdistnl7.png"
target="_blank"><img src="http://img339.imageshack.us/img339/6609/
whatdistnl7.th.png" border="0" alt="Free Image Hosting atwww.ImageShack.us"
/></a>
click here if the above doesn't work
http://img339.imageshack.us/img339/6609/whatdistnl7.th.png
The picture looks more like
y_i = Y + (2*W - 1)*sqrt(x_i) + stdev*Z
where W is a Bernoulli (0,1) random variable with p = .5
Actually, the model should include a multiplier, say b:
y_i = Y + b*(2*W - 1)*sqrt(x_i) + stdev*Z
.
- References:
- What distribution is this?
- From: Enter The
- Re: What distribution is this?
- From: Ray Koopman
- What distribution is this?
- Prev by Date: Re: statistical approach question
- Next by Date: Stat problem
- Previous by thread: Re: What distribution is this?
- Next by thread: Re: Some of Reef Fish's behavior MUST NEVER AGAIN BE TOLERATED
- Index(es):
Relevant Pages
|
|