Hello Statalisters --
There have been a few threads over the last several days re:
distributions, distribution selection, generating distributions etc. I
note particularly the threads entitled "Generating random variables from
the logistic distribution" and what I think is the recent updating of
the SSC package RND by Kit Baum. Anyway, I thought I would try that out
and generate a weibull distribution and then use Nick Cox's packages
-wfit-, -qweibull- and -pweibull- to see how well the RND routines from
SSC did in generating a Weibull distribution. I generated 10000 values
from a Weibull distribution with a shape of 3 and a scale of 2 as per an
example in the rnd package help file. But when I used the following
sequence of commands (see below), it seems like there might be a
mismatch because instead of the scale parameter of 2 (which I used to
generate the data), I get one of 0.5 (its inverse) back. The graphs
generated by Nick's programs seem to do very well (perfect straight
lines) except they didn't provide back the parameters that I thought
they would. Maybe I am mixing up the terminology (b vs. c parameters,
scale vs. shape) or not doing the appropriate backconversion to get the
original parameters. Or is their perhaps a concern?
Notes:
1. (/m# option or -set memory-) 10.00 MB allocated to data
2. (/v# option or -set maxvar-) 5000 maximum variables
. ssc install rnd, replace
checking rnd consistency and verifying not already installed...
all files already exist and are up to date.
. help rnd
. rndwei 10000 3 2
( Generating . )
Variable xw created.
. wbull xw
Fitting Weibull distribution to xw
b 0.49809
c 3.02015
. qweibull xw
Fitting Weibull distribution
b: exp(_b[_cons])
c: exp([ln_p][_cons])
------------------------------------------------------------------------------
xw | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------------
b | .4980915 .0017364 286.86 0.000 .4946883
.5014947
c | 3.02015 .0235553 128.22 0.000 2.973982
3.066317
------------------------------------------------------------------------------
. pweibull xw
Fitting Weibull distribution
b: exp(_b[_cons])
c: exp([ln_p][_cons])
------------------------------------------------------------------------------
xw | Coef. Std. Err. z P>|z| [95% Conf.
Interval]
-------------+----------------------------------------------------------------
b | .4980915 .0017364 286.86 0.000 .4946883
.5014947
c | 3.02015 .0235553 128.22 0.000 2.973982
3.066317
------------------------------------------------------------------------------
. qweibull xw, param( 0.49 3)
. qweibull xw, param( 2 3)
. query born
05 Jul 2005
. which rndwei
c:\ado\plus\r\rndwei.ado
*!version 1.1 1999 Joseph Hilbe
.
thanks for any help that might be offered.
david
David J. Miller, Chief
Chemistry & Exposure Branch
Health Effects Division
Office of Pesticide Programs
U.S. Environmental Protection Agency
703-305-5352 (voice)
703 605-1289 (fax)
visit: http://www.epa.gov/pesticides/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
