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

st: RE: AW: Generating random numbers

From   "Barth Riley" <>
To   <>
Subject   st: RE: AW: Generating random numbers
Date   Wed, 25 Nov 2009 10:09:22 -0600

Hi Martin

I took your advice and placed my set seed command at the top of my do file
(which contains several nested loops). And I used the c(current_time)
function, stripping the ":" out of the string and converting the resulting
string into a number for my seed value. This did the trick! Thanks


-----Original Message-----
[] On Behalf Of Martin Weiss
Sent: Wednesday, November 25, 2009 9:07 AM
Subject: st: AW: Generating random numbers


Your "seed generating process" for "sd" also uses -uniform()-, i.e. your
first line is itself dependent on a -seed-. 

Some Statalisters also use the current date, such as in -set seed
`=date("`c(current_date)'", "DMY")'- to -set seed-...


-----Ursprüngliche Nachricht-----
[] Im Auftrag von Barth Riley
Gesendet: Mittwoch, 25. November 2009 15:35
Betreff: st: Generating random numbers


I have written a program in Stata to create several datasets containing
simulated item response values, which are based on randomly generated item
response theory parameters. I am finding that the randomly generated
parameters for each dataset are following a predictable pattern (one set of
parameter values for odd-numbered datasets; another set of values for
even-numbered datasets). I have been using the set seed command to try to
avoid this repetitive pattern, for example:

local sd = `j' * int(uniform() * 1000)
set seed `sd'
local b = invnormal(uniform())

Where `j' is a forvalues counter variable. Is there a better way to do this?



*   For searches and help try:

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index