Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: recursively defined variables

From   Bjorn Persson <>
Subject   st: recursively defined variables
Date   Fri, 16 Sep 2005 14:10:26 -0400

Dear Statalist,

I received a very helpful answer (from Nick Cox) to the problem I posted yesterday re recursive variables, as well as a reference to Roger Newson's useful Stata Tip in Stata Journal 4(4) 2004. However, when I try to replicate the examples in Newson's note I keep getting the error message " obs.nos. out of range" after the command:

replace y = y[_n-1] +... (from Newson's Fibonacci sequence example)

Any ideas on what's going wrong here?

(Sorry if this is a double posting - my first mail seems not to have gone through)


Nick Cox wrote:

This notation I understand to mean
xt -> x at time t xt-1 -> x at time t - 1
x1 -> x at time 1

If so, you initialise
gen x = z in 1
and then you replace
replace x = b * x[_n-1] + z in 2/l
where I guess b is a place-holder for a number you know and note that "l" stands for "l"ast.
The principle is given by the title
of a useful Stata Tip by Roger Newson:
generate and replace use the current sort order. Stata Journal 4(4):484--485 (2004)


I was wondering if anybody could help me to define a variable
recursively in stata? That is, I would like define each observation x
as the sum of a flow variable, z, and the lagged x variable,
discounted by a factor b:

xt = zt + b*xt-1

The three first x variables would then be:

x1 = z1

x2 = z2 + b*z1

x3 = z3 + b*z2 + b^2*z1

and so on).

*   For searches and help try:

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