Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: qnorm


From   Nick Cox <n.j.cox@durham.ac.uk>
To   "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu>
Subject   RE: st: qnorm
Date   Mon, 5 Mar 2012 15:23:52 +0000

The approach in Maarten's program is to generate a number of random samples and show the lot as replicates. 

Here as an alternative is some example code for individual 95% confidence intervals for each plotted point. -qplot- used at the end is from SJ. The code isn't smart about missing values, but it could easily be made smarter. I also guess the code could be shortened in the middle. 

sysuse auto, clear

mata 
y = sort(st_data(., "mpg"), 1) 
mean = mean(y)
sd = sqrt(variance(y)) 
n = rows(y) 

compare = J(n, 0, .)

for (j = 1; j <= 100; j++) { 
	compare = compare, sort(rnormal(n, 1, mean, sd), 1) 
}

envelope = J(n, 2, .) 
for (i = 1; i <= n; i++) { 	
	x = sort(compare[i,]', 1) 
	envelope[i,] = ((x[2] + x[3])/2, (x[97] + x[98])/2)  
} 

names = tokens("_lower _upper") 
(void) st_addvar("float", names) 
st_store(., names, envelope) 

end 

qplot mpg _lower _upper, ms(O i i) c(.  J J) legend(off) ytitle("`: var label mpg'") 

Nick 
n.j.cox@durham.ac.uk 

Maarten Buis

On Mon, Mar 5, 2012 at 10:03 AM, Nick Cox wrote:
> 1. Simulate several samples from a distribution with the same mean and
> standard deviation (or more generally an appropriate mean and standard
> deviation) and use the resulting portfolio of plots in assessing what
> kind of variability is to be expected.

An easy way to do so is to use the -margdistfit- package, which you
can install by typing in Stata -ssc install margdistfit-. The default
is actually to first sample the mean and the standard deviation from
its sampling distribution and than sample a new variable with those
sampled means and standard deviation. I suspect that this makes sense
in most cases, though I also suspect that it won't matter much. If you
want to do exactly what Nick proposes you can add the -noparsamp-
option.

Here is an example of what such a graph would look like:

*-------- begin example -----------
sysuse auto, clear
reg mpg
margdistfit, qq name(qq)
margdistfit, pp name(pp)
margdistfit, hangroot name(hangr)
margdistfit, cumul name(cumul)
*--------- end example ------------
(For more on examples I sent to the Statalist see:
http://www.maartenbuis.nl/example_faq )

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany


http://www.maartenbuis.nl
--------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index