Statalist


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

st: RE: RE: RE: RE: RE: RE: distribution curve with svy


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: RE: RE: RE: RE: RE: distribution curve with svy
Date   Tue, 8 Sep 2009 20:22:46 +0100

In addition to -cumul-, -distplot- from STB/SJ has been supporting
distribution plots since 1999. -search distplot- to see locations. 

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

Martin Weiss

What happens when you run this example? Works for me...


**************
clear*
set obs 10000

gen byte dropout=/*
*/runiform()<0.3
gen pincp=rnormal()
gen pwgpt=1+int(10*runiform())

cumul pincp if dropout==1 /*
*/[fweight=pwgpt], gen(incdrop)

line incdrop pincp/*
*/ if dropout==1, sort
**************

Holly Kosiewicz

I know Stata is a bit finicky at times, but I just eliminated the space
and
Stata is still telling me I am using invalid syntax. The space is not
the
problem.

Martin Weiss [mailto:martin.weiss1@gmx.de] 

No blank after -gen-! Just " gen(incdrop)"

Holly Kosiewicz

Thanks very much for your help. It seems as though this command may
work.
Unfortunately, when I type "cumul pincp if dropout==1 [fweight=pwgpt],
gen
(incdrop)", stata tells me I am using an invalid syntax. It seems as
though
I am following the syntax given in the help window for cumul so I can't
tell
where Stata is finding the error. 

Martin Weiss [mailto:martin.weiss1@gmx.de] 

How about the second example in -help cumul-?

Holly Kosiewicz

I currently have a dataset that contains a sample of individuals; this
dataset also contains a final weight variable that can be used to
generate
population estimates.

I would like to draw a distribution curve of a variable that measures
income, by the educational attainment level of the individual.  The end
goal
is to see two  distributional curves - each curve representing a
population
with a specific educational attainment level - one superimposed on the
other, and compare them. Ideally, I would make these comparisons based
on
the population, not on the sample.

The first thing that came to mind was a histogram command, but I noticed
that the only curve that it can generate is a normal curve. That's not
what
I want. Then I thought about using a graph twoway scatterplot, but that
too
does not produce the distributional curve that I want. Any thoughts?

*
*   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   |   What's new   |   Site index